2017 Book TheBusinessOfBanking

PA LG R AV E M AC M I L L A N S T U D I E S I N
BANKING AND FINANCIAL INSTITUTIONS

S E R I E S E D I TO R : P H I L I P M O LY N E U X
The Business
of Banking
Models, Risk and Regulation
Edited by Giusy Chesini,

Elisa Giaretta and Andrea Paltrinieri
Palgrave Macmillan Studies in Banking and
Financial Institutions
Series editor
Philip Molyneux
University of Sharjah
Sharjah, UAE
The Palgrave Macmillan Studies in Banking and Financial Institutions
series is international in orientation and includes studies of banking
systems in particular countries or regions as well as contemporary themes
such as Islamic Banking, Financial Exclusion, Mergers and Acquisitions,
Risk Management, and IT in Banking. The books focus on research and
practice and include up to date and innovative studies that cover issues
which impact banking systems globally.
More information about this series at

http://www.springer.com/series/14678
Giusy Chesini Elisa Giaretta
•
Andrea Paltrinieri
Editors
The Business
of Banking
Models, Risk and Regulation
Editors
Giusy Chesini Andrea Paltrinieri
Department of Economics Department of Economic and Statistical
University of Verona Sciences
Verona, Italy University of Udine
Udine, Italy
Elisa Giaretta
Department of Business Administration
University of Verona
Verona, Italy
Palgrave Macmillan Studies in Banking and Financial Institutions

ISBN 978-3-319-54893-7 ISBN 978-3-319-54894-4 (eBook)
DOI 10.1007/978-3-319-54894-4
Library of Congress Control Number: 2017943467
© The Editor(s) (if applicable) and The Author(s) 2017

This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made. The publisher remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
Cover image: © Maciej Bledowski/Alamy Stock Photo

Cover design: Tom Howey
Printed on acid-free paper
This Palgrave Macmillan imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Acknowledgements
First and foremost, we would like to thank all our contributors, whose
biographies are provided in this volume, without which the edited book
would not have been possible.
Also, we want to express our gratitude to all participants of the 2016
Wolpertinger Conference—organized by the European Association of
University Teachers of Banking and Finance in September 2016 at the
University of Verona—for their insightful comments about all the papers
included in this volume.
We would also like to show our gratitude to Professor Philip
Molyneux (Professor of Banking and Finance and Dean of the College of
Business, Law, Education and Social Sciences), editor in chief of the
Palgrave Macmillan Studies in Banking and Financial Institution Series,
for approving our book proposal and for his support during the process.
Also, many thanks to the Palgrave Macmillan team, Aimee Dibbens
and Natasha Denby, for their support during the publishing process.
Special thanks to the institutions which kindly supported the
Wolpertinger Conference in Verona and so contributed to make it a
pleasant and fruitful event: Banco Popolare, Unicredit (Verona), and
S&P Global Market Intelligence.
v
vi Acknowledgements
Finally, as conference organizers, we would like to thank Steven

Ongena, Professor at the University of Zurich, for giving a plenary
speech at the conference on “Relationship Lending Reloaded” and the
other speakers at the Jack Revell Session on “The Financial System in the
Macroeconomy” (Cesare Bisoni, University of Modena; Maurizio
Faroni, General Director at Banco Popolare; Phil Molyneux, Bangor
Business School; and Greg Udell, Indiana University).
Contents
1 Introduction 1
Giusy Chesini, Elisa Giaretta and Andrea Paltrinieri
2 Interest Rates and Net Interest Margins: The Impact

of Monetary Policy 5
Paula Cruz-García, Juan Fernández de Guevara
and Joaquín Maudos
3 The Swedish Mortgage Market: Bank Funding, Margins,

and Risk Shifting 35
Viktor Elliot and Ted Lindblom
4 Incapability or Bad Luck? Testing the “Bad

Management” Hypothesis in the Italian Banking
System 55
Fabrizio Crespi and Mauro Aliano
vii
viii Contents
5 Why Do US Banks React Differently to Short Selling

Bans? 79
Daniele Angelo Previati, Giuseppe Galloppo, Mauro Aliano
and Viktoriia Paimanova
6 Reputational Risk in Banking: Important to Whom? 109

Ewa Miklaszewska and Krzysztof Kil
7 The Business Model of Banks: A Review of the

Theoretical and Empirical Literature 131
Stefano Cosma, Riccardo Ferretti, Elisabetta Gualandri,
Andrea Landi and Valeria Venturelli
8 On European Deposit Protection Scheme(s) 169

Milena Migliavacca
9 A Technical Approach to Deposit Guarantee Schemes 203

Francesca Arnaboldi
Index 235
Notes on Contributors
Mauro Aliano is an Assistant Professor of Banking and Finance at the

University of Cagliari, Faculty of Economics, Law and Political Sciences.
He is a specialist in applying statistics techniques for analyzing financial
markets, in methods for analyzing financial instruments and in portfolio
models. In 2012, he was a research fellow at the University of Rome Tor
Vergata.
Francesca Arnaboldi is Associate Professor of Banking and Finance at
the University of Milan, Italy. She holds an MSc in Financial
Management from the University of London, CeFIMS, and a Ph.D. in
Finance from the University of Bologna. She has been a visiting scholar
at Stern School of Business, New York University. She is a member
of the CEFIN (Centre for Research in Banking and Finance), University
of Modena and Reggio Emilia and of the Centre for Banking Research,
Cass Business School, City University. Her main areas of interest are
international banking, bank competition and performance, financial
innovation, bank regulation, and supervision.
ix
x Notes on Contributors
Giusy Chesini is Associate Professor in Banking and Finance at the

University of Verona, Italy. She holds a Ph.D. in financial markets and
intermediaries from the University of Bergamo, Italy. Giusy is also the
author of several papers and books related to the evolution of financial
intermediation. Her research topics include financial markets, stock
exchanges, banking and corporate finance.
Stefano Cosma is Professor of Banking at the University di Modena and
Reggio Emilia, Italy. He holds a Ph.D. in Business Administration from
University Cà Foscari, and a Master’s degree in Organization and
Management of Human Resources. He is a member of CEFIN (Centre
for Research in Banking and Finance).
Fabrizio Crespi is researcher of Banking and Finance at the University
of Cagliari and Contract professor at ALTIS—Post Graduate School
Business and Society—at the Catholic University of Milan. His main
research fields are as follows: organizational models of credit institutions
and other financial intermediaries, internationalization and financial
intermediaries, asset allocation, and portfolios optimization.
Paula Cruz-García is a Ph.D. student in Economics and Quantitative
Finance at the University of Valencia. She graduated in Economics at the
University of Valencia, and she holds a Master’s degree in Quantitative
Banking and Finance from the University of Valencia, Complutense
University of Madrid, University of Castilla La Mancha, and the
University of the Basque Country. Her main research interests are
banking economics, financial economics, and econometrics.
Viktor Elliot is Assistant Professor in Banking, Finance and Accounting
at the School of Business, Economics & Law, University of Gothenburg.
His interests include performance management, risk and regulatory
implications in banking, funds transfer pricing, and savings banks.
Current research is conducted within the areas of financial exclusion,
supply chain finance, and regulatory implications in banking.
Riccardo Ferretti is Professor of Banking and Finance at the University
di Modena and Reggio Emilia, Italy. He holds a Ph.D. in Financial
Markets and Portfolio Management from University of Bergamo. He has
been a visiting Ph.D. student at Graduate School of Business
Notes on Contributors xi
Administration, New York University. He is a member of CEFIN

(Centre for Research in Banking and Finance).
Giuseppe Galloppo is Assistant Professor of Financial Markets and
Institutions at Tuscia University, Viterbo, and a research fellow at the
School of Economics, Tor Vergata University of Rome. He teaches
banking and finance, with a particular focus on financial markets and
institutions and risk methods. He is a specialist in applying statistical
techniques and methods for analyzing financial instruments and portfolio
models and for assessing risk profiles of securities and financial assets
portfolios.
Juan Fernández de Guevara is Assistant Professor of Economics at the
University of Valencia, where he graduated and obtained his Ph.D. (with
special honors) in Economics. Since 2011, he is associate researcher of the
Ivie. His specialized fields are banking, productivity analysis, and social
capital. At present, he is a researcher in the SPINTAN Project of the
European Union’s Seventh Framework Programme and has been a
consultant for the United Nations and the European Investment Bank
(EIB). He has published fifteen books and chapters and more than twenty
articles in specialized journals such as Journal of Banking and Finance,
The Manchester School, The European Journal of Finance, Journal of
International Money and Finance, Revista de Economía Aplicada,
Regional Studies, Applied Economics Letters, among others, and has
taken part in more than twenty-five national and international congresses.
Elisa Giaretta is a Research Fellow at the University of Verona, Italy,
where she received a Ph.D. in Business Administration and
Management. She works in the ‘Polo Scientifico e Didattico di Studi
sull’Impresa’, an academic centre focused on the analysis of Italian
enterprises. Her research topics include private equity, companies’ net-
works and bank risks.
Elisabetta Gualandri is Full Professor of Banking and Finance in the
“Marco Biagi” Department of Economics of the University of Modena
and Reggio Emilia. She is a director of the European Association of
University Teachers in Banking and Finance. She served as an auditor of
xii Notes on Contributors
Banca d’Italia from 2007 to 2012, when she was appointed to the Board
of BPER banca. Her recent research topics include regulation and
supervision of financial intermediaries and markets, financial crisis, the
financing of innovative SMEs, and public intervention programs.
Krzysztof Kil is Assistant Professor of Banking and Finance at the
Cracow University of Economics, Faculty of Finance and Law, Poland.
His research concentrates on the issues of bank stability and bank effi-
ciency in Central and Eastern Europe.
Andrea Landi is Full Professor of Banking and Finance in the “Marco
Biagi” Department of Economics of the University of Modena and
Reggio Emilia, and the President of Fondazione Cassa di Risparmio di
Modena from 2005 to 2015. He served as an auditor of Cassa Depositi e
Prestiti from 2014 to 2016. His recent research topics include bank
strategies, efficiency and performance, financial crisis, financing of
innovative SMEs and public intervention programs, asset management.
Ted Lindblom is Professor of Business Administration at the University
of Gothenburg, Sweden. His current research interests mainly concern
corporate finance and banking. In the corporate finance area, he par-
ticularly focuses on corporate governance, capital budgeting, and finan-
cial structure decisions. In the banking area, his emphasis is on banking
strategies, pricing, profitability, and risk management under different
market condition and regulatory frameworks. He has authored and
co-authored several articles and books regarding these issues.
Joaquín Maudos is Professor of Economics at the University of
Valencia, Research Deputy Director at the Ivie and collaborator at the
CUNEF. He has been visiting researcher at the Florida State University
Finance Department, at the College of Business at Bangor University,
and at the School of Business of the University of Glasgow. He has also
been a consultant to the European Commission, the European
Investment Bank, and the United Nations. He has jointly published
seventeen books and nearly ninety articles in specialized journals
(European Journal of Finance, Journal of Banking and Finance, Journal
of Comparative Economics, Journal of Financial Services Research,
Notes on Contributors xiii
Journal of International Financial Markets, Institutions and Money,

Journal of International Money and Finance, Regional Studies, Review
of Income and Wealth, Journal of Business Economics and
Management, etc.). He is a member of the Editorial Board of the journal
Inversión & Finanzas, as well as director of competitive projects.
Milena Migliavacca is a Ph.D. candidate and Contract Professor of
Financial Intermediaries at Catholic University in Milan and visiting Ph.
D. student at Essex Business School, University of Essex. Her current
research interests concern financial literacy, social impact investment, and
deposit insurance schemes design.
Ewa Miklaszewska is a Professor in Banking and Finance at the Cracow
University of Economics (CUE), Poland, where she chairs the Banking
Division. She is an Associate Professor of Economics at the Jagiellonian
University in Cracow, Department of Management and Public
Communication. She has held several visiting positions at both foreign
universities and Polish financial regulatory institutions. Her research
interests focus on bank regulation and bank strategies.
Andrea Paltrinieri is Assistant Professor in Banking and Finance at the
University of Udine, Italy. He holds a Ph.D. in Business Administration
from the University of Verona, Italy. Research topics include asset man-
agement and institutional investors, with a particular focus on sovereign
wealth funds, Islamic finance and the relative financial instruments such as
sukuk, commodity markets and exchange traded products.
Viktoriia Paimanova is a researcher of Financial Markets and
Institutions, School of Economics and Management, at V.N.Karazin
Kharkiv National University (Ukraine). She focuses her interests on
statistics and analyzing of financial markets, assessment of risk profiles
and financial assets. In 2014–2016, she was a research fellow at Tuscia
University (Italy).
Daniele Angelo Previati is Full Professor of Financial Markets and
Institutions in the Department of Management of the University of
Rome III and Professor at the SDA Business School, Bocconi University,
Milan. He has been teaching banking and finance for more than 30
xiv Notes on Contributors
years, with particular focus on bank management, strategy, and organi-

zation in the financial services industry and e-finance. His main research
interests relate to various perspectives on bank management: human
resources management, intellectual capital, organizational change,
stakeholder management, and finance for SMEs. He has published
widely in academic journals and books. He has also acted as a consultant
for banks and the Italian Central Bank on organization design and
human resources management.
Valeria Venturelli is Associate Professor in Banking and Finance at the
“Marco Biagi” Department of Economics of the University of Modena and
Reggio Emilia, where she teaches Financial Markets and Institutions at
both undergraduate level and graduate level. She graduated in Economics
from the University of Modena and Reggio Emilia and received a Ph.D. in
Financial Markets and Institutions from the Catholic University of Milan.
Her main research interests are the economics of banking and other
financial institutions and valuation methods. She is the author of several
articles in leading academic journals. She has acted as a consultant to
various public institutions and consulting firms. She is a member of
CEFIN—Center for Studies in Banking and Finance and Softech-ICT.
List of Figures
Fig. 2.1 Intervention interest rates by the main Central Banks 6

Fig. 2.2 Three-month interbank rates evolution 23
Fig. 2.3 Net interest income evolution (% total assets) 24
Fig. 2.4 Economic impact of the net interest margin determinants
(bp). The graph shows the effect on net interest income
of a variation of 25–75 percentile of the distribution in each
of the explanatory variables. The bars that have a more
subdued colour correspond to variables whose effect is not
statistically significant. The variables are sorted from
highest to lowest impact on net interest income. The
equation [2] of the Table 2.2 was used for the analysis 27
Fig. 2.5 Observed changes in interest rates and predicted changes
in the net interest margin (bp) 29
Fig. 3.1 Illustration of overcapitalization of CB issues 38
Fig. 3.2 Growth (in mEuro) of outstanding CBs on the Swedish
market since 2006 40
Fig. 3.3 Indirect issuing of CBs through a bank-owned
“building society” 40
Fig. 3.4 Direct issuing of CBs through a bank-owned
“building society” 41
Fig. 3.5 Illustration of the roles of actors involved in CB issues 41
xv
xvi List of Figures
Fig. 3.6 Average discounts on the banks’ officially offered

mortgage rates the past year 45
Fig. 3.7 Aggregate bank funding 1996–2015 46
Fig. 3.8 Pre- and post-CB claims by different bank funds
providers on bank assets in case of bankruptcy 46
Fig. 3.9 Average market rates and risk premiums on 2-
and 5-year covered bonds (CB/MB) relative government
bonds (GB) for 2000–2016 47
Fig. 3.10 Average official mortgage rates of the five Swedish
“mortgage” banks from 2000 to 2016 47
Fig. 3.11 Deviations of the banks’ officially offered mortgage rates
from 2000 to 2016 48
Fig. 3.12 Marginal mortgage lending margins of the average bank
from 2000 to 2016 49
Fig. 3.13 Illustration of average interest rate margins if borrowing
short and lending long 49
Fig. 7.1 BM definition: the different approaches in the management
literature 139
Fig. 7.2 Strategic components of banking business model (BBM) 141
Fig. 8.1 EU28 Member States DPSs’ design 197
Fig. 9.1 Banks (%) and coefficients 212
List of Tables
Table 2.1 Descriptive statistics (2003–2014 averages) 18

Table 2.2 Determinants of net interest income: 2003–2014 25
Table 2.3 Observed changes in interest rate and yield slope curve
and predicted changes in net interest margin (bp) 28
Table 3.1 Credit ratings of CBs issued by Swedish institutions 42
Table 3.2 Key figures (in SEKm) of selected banks as of December
31, 2015 43
Table 4.1 Dataset description 66
Table 4.2 Mean ratio of impaired loans on credit to clients for the
banks in our dataset 67
Table 4.3 Descriptive statistics of variables used 69
Table 4.4 Panel regression. Bank-specific determinants on credit
quality 72
Table 4.5 VIF test results 74
Table 4.6 Redundant Fixed Effects Tests 74
Table 4.7 Panel regression. Bank-specific determinants on credit
quality. GMM 74
Table 5.1 Fundamental variables description 87
Table 5.2 Short selling ban interventions in US financial sector 91
Table 5.3 Descriptive statistics 92
Table 5.4 Fundamental firms’ analysis 95
xvii
xviii List of Tables
Table 5.5 Stock price change 97

Table 5.6 Total and systematic risk analysis 100
Table 6.1 Scoring scale used in the model 123
Table 6.2 Description of explanatory variables 124
Table 6.3 Panel data estimations for MLPS, CEE-11, 2009–2014 125
Table 6.4 Panel data estimations for ROE, CEE 2009–2014 126
Table 8.1 Deposit protection scheme (DPS) features 173
Table 8.2 Design features distribution across European countries 179
Table 8.3 Deposit protection schemes (DPSs) in the EU28 182
Table 8.4 Funding and management design 196
Table 9.1 Balance sheet ratio computed from Bankscope 210
Table 9.2 Thresholds, classes and coefficients 211
Table 9.3 Statutory position, aggregate indicator and scaled
aggregate indicator 212
Table 9.4 Year-to-year statutory position 213
Table 9.5 Weighted average aggregate indicator 215
Table 9.6 Changes in statutory position after risk
adjustment—all years 216
Table 9.7 Descriptive statistics—EBA core and additional indicators 217
Table 9.8 Buckets, boundaries and individual risk score 222
Table 9.9 Buckets, relative boundaries and individual risk score 222
Table 9.10 Aggregate risk weight 225
Table 9.11 Number of banks, risk classes, ARWcore and
ARWcore+additional (2013 and 2014) 225
Table 9.12 Risk classes (FITD versus EBA) 227
Table 9.13 Changes in risk classes 228
Table A.1 Test for difference in means—ROE 231
Table A.2 Top and bottom quartile—ROE 232
1
Introduction
Giusy Chesini, Elisa Giaretta and Andrea Paltrinieri
This text comprises a selection of papers that offers new insights into
banking business models, risks and regulation proposals in the aftermath
of European financial crises. It investigates the main issues affecting the
business of banking nowadays such as low interest rates and
non-performing loans. The combined effect of low to negative interest
rates and weak economic growth has encouraged banks to shift their
business towards new areas, less related to interest rates, that financial
markets and institutional investors are evaluating. Contributions also
G. Chesini (&) E. Giaretta

Department of Business Administration,
University of Verona, Verona, Italy
e-mail: giusy.chesini@univr.it
E. Giaretta
e-mail: elisa.giaretta@univr.it
A. Paltrinieri
Department of Economic and Statistical Sciences,
University of Udine, Udine, Italy
e-mail: andrea.paltrinieri@univr.it
© The Author(s) 2017 1
G. Chesini et al. (eds.), The Business of Banking, Palgrave Macmillan Studies
in Banking and Financial Institutions, DOI 10.1007/978-3-319-54894-4_1
2 G. Chesini et al.
offer new insights into topics not yet fully investigated by the literature
such as banks’ short-selling bans after Brexit, the European Deposit
Guarantee Scheme and banks’ risk appetite framework.
These chapters were originally presented as papers at the annual
conference of the European Association of University Teachers of
Banking and Finance (otherwise known as the Wolpertinger Conference)
which was held at the University of Verona, Italy, at the beginning of
September 2016.
In particular, the second chapter, “Interest Rates and Net Interest
Margins: The Impact of Monetary policy,” by Paula Cruz-García, Juan
Fernández de Guevara, and Joaquín Maudos, examines the determinants
of bank’s net interest margin, focusing on the effect of interest rates, and
thus monetary policy decisions. The analysis is carried with a panel of
banks from 32 OECD countries over the period 2003–2014. The results
show a quadratic relationship between net interest margins and interest
rates, implying that the variation of the latter (and therefore monetary
policy) has a greater effect when interest rates are low. An important
implication of economic policy regarding the results obtained is that
there is a trade-off between economic growth and financial stability
associated with the impact of expansionary monetary policy when the
level of interest rates is very low. As a result, if the current scenario of low
and even negative interest rates persists for much longer in certain
countries (such as in the Eurozone), it will have a negative effect on bank
profitability and consequently on financial stability.
Chapter 3, “The Swedish Mortgage Market—Bank Funding, Margins
and Risk Shifting”, by Viktor Elliot and Ted Lindblom analyzes the
Swedish mortgage market, especially focusing on bank funding, margins,
and risk shifting. It discusses the move from mortgage-backed bonds to
covered bonds regime in Sweden and its implications for bank’s prof-
itability and risk-taking. This chapter concludes by offering a discussion
about the risk of a new financial crisis in Sweden.
In Chap. 4, “Incapability or Bad Luck? Testing the “Bad
Management” Hypothesis in the Italian Banking System” by Fabrizio
Crespi and Mauro Aliano, by using specific evidences from the Italian
banking sector and following a microeconomic approach, the authors test
the “bad management” hypothesis first introduced by Berger and
1 Introduction 3
Deyoung (1997), which suggests that poor managerial practice causes an

increase in problem loans after a lag. This chapter gives a contribution to
the existing literature in this field; in that, it investigates nonperforming
loans (NPLs) and other soured loans jointly. Their results confirm the
“bad management” hypothesis, in that they discover a positive (lagged)
relation between the value of past due/overdrawn loans and NPLs which,
in a management perspective, indicates the incapability of the credit
manager to anticipate or to recover (at least partially) problematic credits.
The fifth chapter, “Why Do US Banks React Differently to Short
Selling Bans?” by Daniele Angelo Previati, Giuseppe Galloppo, Mauro
Aliano and Viktoriia Paimanova, is about short-selling ban which caught
high attention of policy modeling in different countries. This work is one
of the first to explain the evidence of different bank price reactions in
terms of country and stock market conditions, and to consider both stock
price reaction and risk side. All in all, their findings suggest that the
impact of the ban on the overall market efficiency is heterogeneous and,
in most cases, modest for the countries analyzed. Indeed, you either do
not observe any improvements or they are only short-lived. This chapter
checks the short selling response of US banks, listed in the SP500 in
2008. For the first time, we document that banks react to ban restrictions
differently, mostly because of their variety in terms of fundamental fac-
tors (balance sheet indicators). Considering that, US banks show dif-
ferent reactions to the ban on short selling. Policy makers should decide
which of firm characteristics are better to choose and whether interven-
tions are effective on the market.
Chapter 6, “Reputational Risk in Banking: Important to Whom?” by
Ewa Miklaszewska and Krzysztof Kil, aims to examine the relevance of
reputational risk for banks and the incentives to manage it. The efforts to
manage reputational risk as a self-standing type of risk, and not within an
operational risk framework, are quite recent. Consequently, in the
empirical part of this chapter, the authors propose a methodology to
measure reputational risk, based on the bank stakeholders’ perspective.
Chapter 7, “The Business Model of Banks: A Review of the
Theoretical and Empirical Literature” by Stefano Cosma, Riccardo
Ferretti, Elisabetta Gualandri, Andrea Landi, and Valeria Venturelli
considers that the business model (BM) has become a key concept in
4 G. Chesini et al.
banking literature. The topic’s relevance is due to the impact of the crisis
on bank profitability and risk levels, leading to new challenges for bank
managers, analysts, and regulators. This chapter deals first of all with the
definition of BM in the management literature; afterward the focus is on
bank business model (BBM) and the business model analysis
(BMA) literature, also considering the nexus with bank diversification.
The point of view of supervisory authorities is critically analyzed with
specific regard to BMA embedded in the Supervisory Review and
Evaluation Process (SREP).
Chapter 8, “On European Deposit Guarantee Schemes” by Milena
Migliavacca, aims to provide a dynamic overview of the Deposit
Protection Schemes (DPSs) across the EU28. Using data gathered by the
World Bank’s Bank Regulation and Supervision Surveys, the analysis
critically systematizes the different features that shape the national DPSs’
design. Finally, this study highlights the area where legislative interven-
tion is most needed in order to reach a full-fledged European Deposit
Insurance Scheme (EDIS).
Finally, Chap. 9, “A Technical Approach to Deposit Guarantee
Schemes” by Francesca Arnaboldi, fits within the debate on deposit
guarantee schemes in the European Union, currently under revision,
investigating the changes proposed by directive 2014/49/EU of the
European Parliament and the Council and regulated by the European
Banking Authority. For Italian banks, new rules introduce risk-based
contributions to be paid ex ante to the national deposit guarantee
scheme. The framework proposed by the European Banking Authority
results in a better classification for Italian banks, which requires lower
payments to the scheme. Concerns are raised about the effectiveness of
the European Banking Authority guidelines.
2
Interest Rates and Net Interest Margins:
The Impact of Monetary Policy
Paula Cruz-García, Juan Fernández de Guevara
and Joaquín Maudos
2.1 Introduction
Several central banks have adopted an expansionary monetary policy in
recent years so as to combat the impact of the last financial crisis on the
economy. In addition to the low monetary policy interest rates (Fig. 2.1)
resulting from the measures adopted (both conventional and unconven-
tional), there is also a fall in the long-run natural rate of interest1: This
derives from an excess of savings in relation to investment due to
demographic factors (such as the ageing of the population and the
P. Cruz-García (&) J. Fernández de Guevara J. Maudos

University of Valencia, Valencia, Spain
e-mail: paula.cruz@uv.es
J. Fernández de Guevara
e-mail: radoselo@uv.es
J. Maudos
e-mail: joaquin.maudos@uv.es
J. Fernández de Guevara J. Maudos
Instituto Valenciano de Investigaciones Económicas, Valencia, Spain
6 P. Cruz-García et al.
7.00%
6.00%
5.00%
European Central Bank
(ECB)
4.00%
Bank of Japan (BoJ)
3.00% Bank of England (BoE)
2.00% Federal Reserve (Fed)
1.00%
0.00%
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
Fig. 2.1 Intervention interest rates by the main Central Banks. Source: Bank of
Spain.
depression in consumption which this entails), a lower rate of techno-

logical progress (with the consequent secular stagnation of economies),
low prices of raw materials, particularly oil, and increased demand for safe
assets (pushing prices upward and decreasing yields), etc. What all of this
leads to is a scenario of very low (or even negative) interest rates. In fact, a
high percentage of debt in many countries has negative interest rates.
According to a recent analysis by the International Monetary Fund
(2016), the expansionary monetary policies adopted by some central
banks have eased the access to finance (by reducing the cost of funding
and increasing the availability of credit), thus stimulating aggregate
demand. However, a prolonged period of reduced interest rates can
impair bank intermediation margins (and therefore profitability), given
the existence of a floor in deposit interest rates, since it is difficult for
banks to pass on the drop in interest rates to the deposits interest rates, at
least in the case of households.2 For this reason, the net interest margin is
seen to be most affected in those banks with a greater proportion of
financing via deposits. Likewise, the greater the proportion of variable
interest rate loans in a bank, the greater the deterioration of its
2 Interest Rates and Net Interest Margins: The Impact … 7
profitability, as a result of the fall in financial revenues due to the

reduction of the money market interest rates.
The European Central Bank (2016), on the other hand, highlights in
its Annual Report that the expansionary measures adopted have a positive
impact. The positive impact is driven by the fact that the drop in interest
rates has led to an improvement in the quality of bank assets (since less
risky projects are financed), an increase in lending activity and a drop in
non-performing loans as a result of economic recovery.
Other papers, such as Rostagno et al. (2016), also provide empirical
evidence of the increase in credit growth due to the policy of negative
rates, showing that loans to companies have increased in the Eurozone
with the current expansionary policy.
Taking the above mentioned into account, it is important to differ-
entiate between the impact that falling interest rates have had up until
now and the impact of these extremely low, or even negative, rates
persisting for a prolonged period. To date, the effect has not been neg-
ative, as stated by the IMF and the ECB. However, the IMF warns that if
this scenario persists for much longer, it will have an adverse effect on the
net interest margin and therefore on bank profitability, primarily due to
the floor in interest rates on deposits, as well as a flattening of the yield
curve which has taken place with the falling interest rates.
In this context, the objective of this study is to analyse the impact of the
variation of interest rates on the net interest margin and the possible
existence of a non-linear relationship which would explain why the impact
of monetary policy differs depending on the level of interest rates. Thus, if
the relationship is quadratic, a fall in interest rates could be harmful for
bank margins if the level of rates is low, while the same drop in rates can
have beneficial effects with high rates (as a result of the reactivation of the
demand for credit, reducing non-performing loans, etc.).
Since there are very few works to date which have empirically analysed
the effect of a prolonged period of low-interest rates on banking net
interest margins (and therefore on profitability), further evidence is
needed on this subject. But given the current context of such low rates,
this issue has attracted attention as shown by the recent works by Borio
et al. (2015) and by Claessens et al. (2016). Using samples of banks from
various countries, both papers provide evidence demonstrating the
existence of a non-linear relationship between interest rates and the net

interest margin. In addition to these two works, are those by Genay and
Podjasek (2014) which analyse the effects of expansionary monetary
policy on the bank margin in the USA, and Busch and Memmel (2015)
for German banks.
Our work provides further empirical evidence for a sample of
32 countries from around the world for the period 2003–2014, a period
that includes years of expansion in which accommodative monetary poli-
cies were adopted and the subsequent years of crisis in which expansionary
monetary policy measures were implemented, both conventional (such as a
decrease in intervention rates), as well as unconventional (QE, negative
rates which penalise excess bank reserves, etc.). The work is focused on
quantifying the impact of short-term interest rates on bank interest mar-
gins, testing the hypothesis of whether the relationship between interest
rates and the margin is indeed quadratic. However, we also consider the
impact of other variables as determinants of net interest margins, which
capture the characteristics of each bank (market power, credit risk, risk
aversion, operating costs, etc.), along with other control variables (market
risk, etc.). We have taken variables used as determinants in the model by
Ho and Saunders (1981) and some of their additions together with the
reference framework by Borio et al. (2015).
The results obtained indicate that the impact of interest rates on the
intermediation margin is quadratic rather than linear. Accordingly, tak-
ing into account this concave relationship and the current low rates, a
normalisation in monetary policy would have a significant effect on
margin recovery. Similarly, this result also shows that if this situation
persists for much longer (and even worse, if the negative rates which
penalise excess bank reserves in some countries are increased), it could
have a negative impact on financial stability as a result of the fall in bank
profitability, which is already at an extremely low level (and below the
cost of raising capital) at least in European banking.
In addition to this introduction, our paper is structured as follows.
Section 2.2 examines the theoretical framework on the determinants of
bank intermediation margins and presents the testable hypothesis.
Section 2.3 describes the sample used, defines the variables of the model
and the empirical approach, and explains the methodology used.
Section 2.4 shows the results obtained and Sect. 2.5 checks the
robustness of the results. Finally, Sect. 2.6 presents the conclusions and
the economic policy implications.
2.2 Theoretical Framework and Testable

Hypothesis
2.2.1 Theoretical Framework
There are various theoretical frameworks in which the behaviour of net

interest margins is modelled (see, for example, Zarruk 1989; or Wong
1997). However, most of the works in the literature take the model
developed by Ho and Saunders (1981) as a starting point. Allen (1988)
extended this model by incorporating different types of loans and
deposits. In this extension, the author showed that the margins can be
reduced when one considers the cross elasticity of demand between
banking products. Angbazo (1997), on the other hand, expanded the
original model by taking into account credit risk as well as interest rate
risk. Maudos and Fernández de Guevara (2004) extended the model to
include operating costs. In addition, their analysis of net interest margins
in the main sectors of European banking uses a direct measure of the
degree of market power, such as the Lerner index. Carbó and Rodríguez
(2007) included non-interest income as a determinant of the margin.
In all these models, the bank is considered as an (risk averse) inter-
mediary, maximising the expected utility of its wealth EU (W), between
suppliers and demanders of loans in a static framework over a single
period. In the model, the banks set interest rates (rL and rD) on their
loans (L) and deposits (D), setting markups a and b on the money market
interest rate (r). Banking activity is subject to two types of risks: (1) the
uncertain profitability of their loans associated with default risk; and
(2) the risk that banks take because of their position in the money market
to which they call on when they need to grant new loans or to place
excess liquidity. Both risks are introduced by assuming that interest rates
on loans and the money market have a probability function with variance
a2L y a2C , respectively. In addition, both risks are related (with covariance
rLC). For each additional loan or deposit, banks must assume operating
costs Exp(QL) or Exp(QD), respectively. Finally, the loans and deposits
reach banks according to Poisson processes which depend on the spreads
that banks set on the interbank interest rate. These processes include the
parameters that determine the market power (a/b) of banks in their
markets.
In an application for the case of German banking, Entrop et al. (2015)
include the cost of the maturity transformation, defining the equation
that describes the determinants of the intermediation margin (s) in the
following way:

1 a 1 ExpðQL Þ ExpðQD Þ 1 rL rD
s¼ þ þ
2 b 2 QL ð 1 þ r Þ Q D ð 1 þ r Þ 2 ð1 þ r Þ

00 ðQ þ Þðr þ þ
1 U ðW Þ C Þ 2ðrLD þ rCD ÞðD0 þ L0 Þ þ rD ð2D0 þ QD Þ
2 2 2
L 2L0 L 2r LC r
þ
4 U ðW Þ
0 ð1 þ rÞ
With these additions to the original model by Ho and Saunders

(1981), the determinants of the net interest margin are the level of
interest rates (r, rLand rD), the degree of competition (a/b), risk (credit
risk, as well as market risk, and their interaction- r2L , r2C and rLC -), bank

risk aversion, 1=2U00 ðWÞ=U 0
ðWÞ, the overheads, the volume of the
initial credit portfolio L0 and of deposits D0, and the average size of
operations QL and QD.
Other group of papers (see Gerali et al. 2010) use a dynamic stochastic
general equilibrium model with an imperfect competition. These authors
postulate a linear relationship between bank margin and the level of
interest rates. Alesandri and Nelson (2015) consider a simple version of
former model in partial equilibrium with the same conclusion.
More recently, Borio et al. (2015) used the Monti-Klein model for the
case where oligopolistic competition exists between N banks, incorpo-
rating the cost of maturity transformation, the capital requirements
coefficient and an equation for provisions for possible loans losses. The
determinants of the net interest margin included in the empirical
application are the three-month interbank interest rate, the slope of the
yield curve and the interest rate risk, in addition to macroeconomic
indicators and variables that approximate the characteristics of each bank

(bank size, risk aversion, liquidity and efficiency). This paper focuses on
the influence of monetary policy on the intermediation margin both
through the impact of the short-term interest rates and the slope of the
yield curve. These authors find that the level of interest rates, which is the
key variable in our work, has a positive non-linear relationship with the
net interest margin, depending on the curvature of the value of elasticity
of demand for loans and deposits and on capital requirements.
In the same vein, Claessens et al. (2016) provide empirical evidence on
the negative effect of the drop in interest rates on net interest margin,
with the impact being greater when interest rates started at a low level,
obtaining a quadratic relationship between the money market interest
rates and the net interest margin.
2.2.2 Testable Hypothesis
In this context, our work takes into account all previous contributions in
so far as we analyse the determinants of the net interest margin by
including the various explanatory variables put forward, but with
emphasis on the effect of interest rates and hence the impact of monetary
policy.
Our testable hypothesis is the following: controlling for bank char-
acteristics and macroeconomic variables, an increase in interest rates has a
positive effect on net interest margin, the impact being greater when
interest rates are low. In other words, we expect a positive and concave
relationship between net interest income and the level of interest rates.
2.3 Data, Definition of Variables

and Methodology
2.3.1 Data
The data used for the empirical analysis come from the BankScope
database (Bureau Van Dijk), which contains information on the balance
and the income statement of a representative sample of banks from

around the world. To control the influence of other macroeconomic
variables which affect the intermediation margin, the World Bank
database is used, while the money market interest rates come from the
OECD database.
The sample used includes financial institutions (banks, savings banks,
credit unions and other types of banks) from 32 OECD3 countries.
The period examined is from 2003 to 2014. Excluded from the
sample are those banks that do not provide the necessary data with which
to calculate any of the variables required for econometric specification
and those whose input prices, necessary for estimating the Lerner index
of market power, are outside the range of the 2.5 standard deviations on
either side of the mean calculated for each year. With these filters, the
panel of data finally used is made up of 54,540 observations.
Variables
In order to carry out the empirical contrast, we used variables put for-
ward by Ho and Saunders (1981) and their subsequent extensions,
adding the level of interest rate and its square, as do Borio et al. (2015).
Therefore, the following variables are needed for econometric specifica-
tion: the level of short-term interest rates, market power, the degree of
bank risk aversion, money market volatility (interest rate risk), credit risk,
the interaction between both types of risk, the volume of credit, liquidity
reserves and average production costs. Each of these variables is
approximated as indicated below:
Level of Interest Rates
We use the three-month interbank market interest rate (Short-term

interest rate) to approximate the level of short-term interest rates. The
expected sign of this variable on the net interest margin is positive.
To capture a possible non-linear relationship between the level of
interest rates and the intermediation margin, the square of the level of
interest rates is included as an explanatory variable.
Market Power
As an approximation of market power, two alternative measures are used.

The first is the Lerner index of market power, which is estimated at bank
level using the approach commonly taken in other works, such as Berg
and Kim (1994) or Maudos and Fernández de Guevara (2004).
The Lerner index measures the ability of companies to set a price
above the marginal cost and is defined as the price-cost margin in relation
to the price:
Pi MCi
Lerner indexi ¼
Pi
where Pi is the average price of banking products, which is approximated

by the total assets and is measured as a ratio between total income and
total assets, and MCi is the marginal cost of production, which is cal-
culated based on the following translog cost function:
1 2
X3
ln Ci ¼ a0 þ a1 ln TAi þ ak ðln TAi Þ þ bj ln wji
2 j¼1
1X 3 X 3
1X 3
þ bjk ln wji ln wki þ c ln TAi ln wji þ l1 Trend
2 j¼1 k¼1 2 j¼1 j
1 X3
þ l2 Trend2 þ l3 Trend lnTAi þ dj Trend ln wji þ ln ui
2 j¼1
where Ci is the total costs of the bank (financial and operating costs) and
TAi is total assets. The definition of the price of production factors is the
following:
w1: Price of labour = Staff costs/total assets4.
w2: Price of capital = Operating costs (except staff costs)/fixed assets.
w3: Price of deposits = Financial costs/deposits.
The cost function estimate is carried out by using a data panel con-
sisting of all the banks in the analysis. So as to capture the influence of
specific variables for each bank, fixed effects are introduced in the cost
function estimate. Finally, a trend variable was also introduced (Trend) to
show the effect of technological change, resulting in displacement of the
cost function over time. As is a common practice, the estimate was made
by imposing the restrictions of symmetry and grade one homogeneity in
input prices.
The second indicator of market power is the Herfindahl index which
approximates the structure or concentration of the market. Although it is
common to use market concentration measures as indicators of com-
petition, such measures have significant limitations for two reasons.
Firstly, the theory shows that when judging competition, it is not always
the number of competitors (or the concentration) that is relevant, but the
rivalry that exists between them. And secondly, indicators of concen-
tration do not show variations between banks in the same country.
Therefore, since the Lerner index is a measure of market power that is
theoretically better grounded than the Herfindahl index, as well as pre-
senting variations at bank level, it will be the preference in the estimate.
However, the sensitivity of the results will be analysed using the
Herfindahl index.
The expected sign of the variables (both the Lerner index and
Herfindahl index) is positive, since banks with greater market power can
set higher margins.
Bank Size
The logarithm of loan volumes (log-loans) is included as a proxy for bank

size, since for a given credit risk, the potential losses will be proportional
to the loan volume, and consequently the risk premium applicable to the
margin. Alternatively, as in Borio et al. (2015), the logarithm for total
assets (log-assets) is also included to verify the robustness of the estimate.
In both cases, the expected sign is positive.
Risk Aversion
The degree of bank risk aversion (Risk aversion) follows the approach
used by McShane and Sharpe (1985) and is approximated by the fol-
lowing ratio:
Equity
RISKAVER =
Total Assets
The expected sign of this variable is positive, since banks with greater
risk aversion will set a higher margin.5
Credit Risk
Given the possibility of non-payment or default on loans, banks include

a risk premium, which is implicit in the interest rates charged on such
transactions. Credit risk is approximated by the ratio between the pro-
vision for insolvencies and the volume of credit granted (Prov/loans),
since the greater the likelihood of insolvency and non-performing loans,
the more provisions banks will provide. The expected sign of this variable
is positive.
Interest Rate Risk
Money market uncertainty is approximated by using the coefficient of

variation calculated with monthly data on the three-month interbank
interest rate (Interest rate risk). The expected sign is positive since, ceteris
paribus, greater volatility means higher risk and thus a greater interme-
diation margin is needed to offset this risk.
Interaction Between Credit Risk and Market Risk (Risk

Covariance)
Interaction between credit risk and market risk (Risk covariance) is

proxied by the product of the measurement of credit risk and the interest
rate risk. The expected sign of this variable is positive, since given a
higher correlation between both types of risk, banks require a greater
intermediation margin.
Average Cost of Transactions (Average Cost)
This is defined as the ratio between total operating costs divided by total
assets. As demonstrated by Maudos and Fernández de Guevara (2004),
the expected sign is positive, since the intermediation margin should
cover at least the operating costs.
Liquid Reserves (Reserves)
A high volume of liquid reserves has a positive effect on the bank

intermediation margin to the extent that they mean an opportunity cost
by banks forgoing investment of these reserves in profitable assets. As a
result, banks have to set a higher intermediation margin to offset lower
interest income. This variable is approximated using the ratio between
liquid reserves and total assets.
It is common practice in some studies to add other control variables.
In particular, also included are implicit interest payments and an indi-
cator of management quality. In addition, GDP growth is included to
capture the possible influence of the economic cycle in determining the
net interest margin.
Implicit Interest Payments
Following Ho and Saunders (1981), Angbazo (1997) and Saunders and

Schumacher (2000), among others, an indicator of implicit interest
payments is included. As an approximation to these payments, we use the

variable operating expenses net of non-interest revenues as a percentage
of total assets (Implicit interest rates). The expected sign of this variable is
positive since higher implicit payments mean increased transaction costs
which demand wider margins to compensate banks for the costs this
entails (instead of fees being charged explicitly, they are implicit in the
form of a greater margin).
Efficiency
Efficient management involves choosing the most profitable assets and

the lowest cost deposits. Management quality is therefore approximated
by the ratio between operating costs and the operating income (cost to
income ratio, Efficiency). The expected sign of this variable is negative,
since the higher the ratio, the greater the operating inefficiency and thus
the smaller the margin.
GDP Growth
As is common practice in studies which analyse banking margins, the

estimate of the annual GDP growth rate (GDP growth) is included to
control for the possible influence of the economic cycle on the net
interest margin.
Net Interest Margin
Finally, the dependent variable to account for, i.e. the net interest margin
per unit of assets (NII), is defined as the difference between revenue and
financial costs in relation to total assets.
Table 2.1 shows the weighted average of each of the variables con-
cerned in our study for the countries analysed.
Table 2.1 Descriptive statistics (2003–2014 averages)
18
Net interest Short-term Implicit Efficiency Lerner Herfindahl Volatility of Credit risk
margin/ total interest interest index index market (prov/loans)
assets (%) rate (%) payments interest rates (%)
(%) (%)
Australia 1.83 4.70 0.87 45.95 0.26 0.08 7.11 0.18
Austria 2.04 2.15 0.99 68.17 0.27 0.06 17.64 1.09
Belgium 1.63 2.01 0.70 69.40 0.27 0.11 17.85 0.09
Canada 1.88 2.52 0.91 73.06 0.28 0.15 8.31 0.22
Chile 3.67 4.33 1.28 58.55 0.36 0.07 20.52 1.01
P. Cruz-García et al.
Colombia 4.79 5.88 1.29 63.48 0.39 0.06 8.23 7.41

Czech 1.88 1.87 1.12 72.94 0.33 0.10 11.37 0.46
Republic
Denmark 3.22 2.27 1.57 67.50 0.36 0.18 14.30 1.45
Finland 1.42 1.30 0.63 68.85 0.37 0.25 25.78 0.17
France 2.06 2.09 0.77 78.80 0.30 0.05 18.31 0.02
Germany 2.39 2.22 1.38 70.28 0.25 0.03 16.50 3.01
Greece 2.21 2.21 0.71 60.96 0.32 0.11 17.34 1.74
Hungary 3.70 6.36 2.18 97.47 0.31 0.09 14.68 2.72
Iceland 3.31 8.49 −0.01 39.08 0.30 0.92 10.10 2.23
Ireland 0.84 2.02 −0.22 31.56 0.38 0.26 18.38 0.27
Israel 2.14 3.07 1.23 73.30 0.27 0.20 19.10 0.49
Italy 2.58 2.52 1.38 71.71 0.30 0.06 15.82 0.79
Japan 1.69 0.30 1.18 100.02 0.29 0.18 15.21 0.64
Korea, Rep. 2.15 3.66 0.59 61.91 0.32 0.05 6.12 2.01
Latvia 2.22 4.53 0.63 71.66 0.40 0.07 26.43 1.73
Luxembourg 0.96 1.56 −0.22 58.39 0.38 0.04 18.48 −0.19
Mexico 8.31 4.91 4.91 67.89 0.34 47.10 5.31 4.71
(continued)
Table 2.1 (continued)
Net interest Short-term Implicit Efficiency Lerner Herfindahl Volatility of Credit risk
margin/ total interest interest index index market (prov/loans)
assets (%) rate (%) payments interest rates (%)
(%) (%)
2
Netherlands 1.42 1.87 −0.08 48.78 0.39 0.23 20.06 0.64

New 2.01 4.68 0.73 54.10 0.31 0.12 6.22 0.22
Zealand
Norway 2.13 3.00 1.01 61.76 0.32 0.13 11.30 0.22
Poland 3.04 4.53 1.13 63.23 0.34 0.06 8.22 1.11
Portugal 2.28 1.11 1.21 68.50 0.29 0.08 28.60 1.01
Russian 5.31 7.71 2.66 80.44 0.37 0.06 20.57 −1.45
Federati
Slovak 2.83 2.19 1.25 65.12 0.36 0.10 22.47 1.13
Republic
Slovenia 2.16 2.34 0.77 53.67 0.28 0.11 22.14 3.05
South Africa 3.82 7.00 1.19 68.57 0.36 0.08 5.91 1.60
Spain 1.87 2.30 0.75 67.58 0.30 0.07 16.74 2.61
Sweden 3.03 1.86 1.47 62.24 0.4 0.1344 27.15 0.3
Switzerland 1.38 0.78 0.48 67.97 0.32 0.1101 49.25 1.32
UK 1.61 2.79 0.85 65.41 0.72 0.0804 12.38 0.86
USA 3.12 2.01 1.59 69.87 0.35 0.0387 20.52 0.58
Loans Total assets Risk Operating costs (% Reserves (% total GDP Number
(log) (log) aversion total assets) assets) growth of obs.
(%) (%)
Australia 14.98 15.53 7.27 1.83 4.12 2.86 322
Interest Rates and Net Interest Margins: The Impact …
Austria 12.48 13.17 8.36 2.58 1.55 1.68 1983

Belgium 13.23 14.74 9.04 2.06 1.19 1.63 449
(continued)
19
20
(%) (%)
Canada 14.29 14.91 8.46 2.14 2.37 2.41 490
Chile 13.99 14.56 14.35 3.02 6.53 4.77 79
Colombia 12.71 13.67 18.08 5.45 6.12 4.85 401
Czech 13.76 14.53 10.02 3.21 2.75 2.76 232
Republic
Denmark 12.57 13.18 13.27 3.14 3.82 0.52 842
Finland 13.55 14.23 8.79 2.02 2.35 0.49 195

France 14.10 14.95 9.87 2.65 1.7 1.65 2474
Germany 12.99 13.63 7.05 2.49 2.24 1.34 12923
Greece 14.67 15.23 11.17 1.97 2.79 0.05 120
Hungary 12.97 13.89 10.88 4.72 6.15 2.16 70
Iceland 11.45 11.98 15.14 3.48 5.81 3.58 105
Ireland 14.13 15.76 14.12 1.62 2.08 3.18 129
Israel 15.38 15.86 6.41 2.45 11.32 3.36 80
Italy 13.05 13.61 11.05 2.55 1.24 0.81 2969
Japan 14.32 14.97 5.41 1.33 2.27 1.64 3265
Korea, Rep. 15.18 16.28 10.59 2.46 5.08 4.21 242
Latvia 12.02 13.24 10.35 2.64 7.59 3.54 202
Luxembourg 12.64 14.68 9.50 1.45 2.95 3.05 585
Mexico 13.87 14.89 14.03 7.70 4.96 3.24 27
Netherlands 13.80 15.17 11.20 1.35 7.26 1.48 167
New 14.52 14.91 8.98 1.42 4.3 2.65 95
Zealand
Norway 12.83 13.04 9.86 1.54 2.66 1.7 1211
(continued)
(%) (%)
2
Poland 14.02 14.62 12.02 2.77 4.18 4.24 277

Portugal 12.32 13.06 11.22 2.13 1.34 −0.19 500
Russian 10.51 11.25 20.43 19.40 6.4 3.62 7012
Federati
Slovak 13.51 14.16 15.57 2.53 5.97 3.74 132
Republic
Slovenia 13.71 14.33 8.81 1.92 3.61 1.44 158
South Africa 13.50 14.52 11.64 4.90 15.26 3.2 166
Spain 14.29 15.02 8.58 1.60 1.39 2.03 778
Sweden 12.26 12.68 13.75 3.08 0.71 1.98 993
Switzerland 12.55 13.07 8.23 2.09 4.21 2.08 4473
UK 12.66 13.79 11.07 1.99 4.64 1.62 1369
USA 13.98 14.5 10.67 2.88 3.8 2.66 9786
Source: BankScope and authors’ calculations
Interest Rates and Net Interest Margins: The Impact …
21
2.3.2 Methodology
With all the variables described, the following equation is estimated:
NIIt ¼ f ðNIIt1 ; Short-term interest ratet ; Short-term interest rate2t ;

Implicit interest paymentst ; Efficiencyt ; Lerner indext ; Interest rate riskt ;
Credit riskt ; Risk covariancet ; Sizet ; Risk aversiont ;
Average costt ; Reservest ; GDP growtht Þ
The analysis of the net interest margin determinants is based on an

estimation of a dynamic panel data model using the Generalized Method
of Moments based on Arellano and Bond (1991) and Blundell and Bond
(1998). In addition to including the net interest margin with its time lag
as an explanatory variable to capture the inertia in its evolution, possible
endogeneity problems are corrected by estimating the model in differ-
ences and using the lagged variables as instruments. Time effects are
included in the estimation to show the impact of specific variables in each
year.
2.4 Results
2.4.1 Base Scenario
Before commenting on the results obtained from the econometric esti-

mation, it is important to analyse how the main variable in our study has
evolved: short-term interest rates. As shown in Fig. 2.2, short-term
interest rates (approximated by the three-month interbank interest rate)
suffered a sharp increase during the years prior to the recent financial
crisis, due to the accommodative monetary policy adopted by the main
central banks. When the crisis hit in 2007, interest rates dropped sharply
as a result of the expansionary monetary policies implemented to combat
the effects of the crisis and have generally remained at levels close to zero
since 2010.
7.00%
6.00%
Eurozone
5.00%
Japan
4.00%
United
Kingdom
3.00%
United States
2.00%
Other
countries in
1.00% the sample
0.00%
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Fig. 2.2 Three-month interbank rates evolution. Source OECD and authors’
calculations
Furthermore, it is also worth observing the evolution of the net

interest margin, as it is the dependent variable in our study. As can be
seen in Fig. 2.3, there are significant differences in the level of net interest
margins between countries/geographical areas throughout the period
analysed. The UK, Japan and the Eurozone have lower margins, while
they are much higher in the USA and the group called “other countries”.
We can also observe that the margin has fallen in the USA, the
Eurozone and Japan, but increased in the group “other countries” and
remained more or less stable in the UK.
Table 2.2 presents the results of the estimation of the equation which
explains the net interest margin. The first column estimates the deter-
minants of the intermediation margin, assuming a linear relationship
between the margin and short-term interest rates. As can be seen, the
effect of the level of interest rates is not statistically significant, thus
discarding a linear relationship between the intermediation margin and
the level of interest rates. The second column also includes the square of
short-term interest rates, obtaining a positive and significant impact on
3.50%
3.00%
2.50%
Eurozone
2.00% Japan
United Kingdom
1.50%
United States
1.00% Other countries
0.50%
0.00%
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Fig. 2.3 Net interest income evolution (% total assets). Source: BankScope and
authors’ calculations
the level but negative for the square, which shows a quadratic rather than
linear relationship. Consequently, a change in interest rates has a greater
impact on the net interest margin the lower the level of interest rates.
Table 2.2 also shows that the maximum in the relationship between
interest rates and the margin is observed at 0.085 (8.5%).
Of the remaining variables, i.e. implicit interest payments, operating
efficiency, bank size, risk aversion and GDP growth, they are significant
and have the expected sign. Thus, higher implicit payments, lower effi-
ciency, larger banks, greater risk aversion, and a positive GDP growth
increase net interest margins.
Robustness of the Results
The third and fourth column analyses the robustness of the results to changes
in the empirical approach to some of the determinants of the net interest
margin. As shown in column 3, the results are maintained when the size is
approximated by the total asset logarithm. Likewise, the results do not vary
when market power is approximated by the Herfindahl index (column 4).
Table 2.2 Determinants of net interest income: 2003–2014

[1] [2] [3] [4]
NIM-1 0.278*** 0.295*** 0.281*** 0.225***
(0.056) (0.052) (0.052) (0.081)
Short-term interest rate 0.090 0.451** 0.408** 1.350**
(0.080) (0.183) (0.181) (0.568)
Short-term interest rate2 −2.663** −2.510** −9.775**
(1.240) (1.236) (3.827)
Implicit interest payments 0.463*** 0.426*** 0.476*** 0.501**
(0.144) (0.134) (0.131) (0.232)
Efficiency −0.008** −0.008** −0.008** 0.006
(0.004) (0.004) (0.004) (0.010)
Lerner index 0.035 0.360 0.282
(0.746) (0.704) (0.702)
Herfindahl index 0.042
(0.038)
Interest rate risk 0.004 0.012 0.013 0.019
(0.007) (0.008) (0.008) (0.012)
Credit risk (provisions/loans) 0.000 0.005 0.001 0.051
(0.010) (0.009) (0.009) (0.045)
Risk covariance −0.023 −0.020 −0.022 −0.063*
(0.026) (0.024) (0.024) (0.034)
Log (loans) 0.278*** 0.313*** 1.392**
(0.099) (0.093) (0.567)
Log (total assets) 0.325***
(0.112)
Risk aversión 0.092*** 0.091*** 0.096*** 0.234**
(0.031) (0.029) (0.030) (0.108)
Average cost −0.004 −0.022 −0.021 −0.064**
(0.013) (0.014) (0.014) (0.031)
Reserves 0.055 0.036 0.035 0.232**
(0.041) (0.038) (0.038) (0.096)
GDP growth 0.191*** 0.204*** 0.218*** 0.445***
(0.056) (0.052) (0.055) (0.128)
Constant −0.038** −0.053*** −0.056*** −0.255**
(0.016) (0.016) (0.020) (0.109)
Max. short-term interest rate 0.085 0.081 0.069
Number observations 38,835 38,835 38,835 38,835
Arellano-Bond test for AR(1) in −2.37 −2.34 −2.35 −0.76
first differences [p-valour] [0.018] [0.019] [0.019] [0.450]
(continued)

[1] [2] [3] [4]
Arellano-Bond test for AR(2) in −0.32 −0.58 −0.64 −0.90
first differences [p-valour] [0.748] [0.559] [0.524] [0.370]
Sargan test of overid. 23.26 22.77 25.84 4.09
Restrictions [p-valour] [0.445] [0.415] [0.259] [0.664]
* p < 0.10, ** p < 0.05, *** p < 0.01
Estimations are done using the generalised method of moments (GMM) based on
Arellano and Bond (1991) and Blundell and Bond (1998), where Lerner index is
instrumented with Herfindahl index, and NIM and other endogenous variables are
instrumented with their own first and second differences. All estimations include
fixed and time effects. Format of the data in the table: Coef. (Robust Std. Error)
Source: Authors’ calculations
2.5 Economic Impact of the Determinants

of Net Interest Margin
To be able to assess how the variation of each explanatory variable affects
the net interest margin it is not enough to simply compare the magnitude
of the estimated coefficient, but rather, the intra-sample variation of each
variable must be taken into account in order to know the economic
impact. Figure 2.4 therefore quantifies the impact of an interquartile
variation in each of the explanatory variables (a change from percentile
25 to 75 of the distribution), taking the estimated parameters in column
2 as references. The variables are ordered from highest to lowest impact,
and the bars in the figure in a more subdued colour represent variables
which are not statistically significant.
As can be seen in the figure, the most important determinants of the
net interest margin for the period analysed are the level of interest rates
(due to the large increase caused by accommodative monetary policy
during the years before the crisis, as well as the sharp fall in rates as a
result of aggressive monetary policy followed by the major central banks
to combat the financial crisis), bank size, the degree of risk aversion, the
economic cycle and operating efficiency. Thus, a variation in short-term
interest rates which means going from percentile 25 to 75 of the dis-
tribution entails an increase in the intermediation margin of 119 basis
points. In the case of bank size, growth in net interest income would be
140
120
100
80
60
40
20
-20
-40
Lo
Ri
Im
Ri
In
In
Re
Le
Cr
Ef
D
ve
sk
sk
te
te
g
pl
fic
ed
rn
se
P
re
re
ra
(lo
ic
er
av
co
rv
it
ie
gr
st
st
it
ge
nc
ris
es
an
er
va
in
ow
in
ra
ra
sio
co
y
de
s)
ria
te
te
te
th
st
x
re
n
nc
le
ris
st
ve
e
k
pa
l
ym
en
ts
Fig. 2.4 Economic impact of the net interest margin determinants (bp). The graph
shows the effect on net interest income of a variation of 25–75 percentile of the
distribution in each of the explanatory variables. The bars that have a more
subdued colour correspond to variables whose effect is not statistically significant.
The variables are sorted from highest to lowest impact on net interest income. The
equation [2] of the Table 2.2 was used for the analysis. Source: Authors’ calculations
83 pb to an equivalent variation of the variable. This variation in the case

of banks’ risk aversion implies an increase in the intermediation margin
of 56 pb; being 51 pb in the case of GDP growth. Finally, a variation in
the operating efficiency of percentile 25 and 75 entails a drop of 15 pb in
the intermediation margin.
Focusing on the impact of interest rates, if instead of using the
interquartile variation range we use the variation which has taken place in
the period analysed, as seen in Table 2.3 and Fig. 2.5 from 2003 to 2007
(subperiod of expansion), the increase in the intermediation margin
explained by the increase in interest rates is 98 bp in the Eurozone,
231 bp in the USA, 117 bp in the UK, 31 bp in Japan and 61 bp in the
group “other countries”. During the subperiod of the crisis 2008–2014,
interest rates fell primarily as a result of the expansionary monetary policy
28
Table 2.3 Observed changes in interest rate and yield slope curve and predicted changes in net interest margin (bp)
Change in Predicted Change in Predicted Change in Predicted
three-month change in net three-month change in net three-month change in net
interest rate interest margin interest rate interest margin interest rate interest margin
2003–2007 2003–2007 2008–2014 2008–2014 2003–2014 2003–2014
Eurozone 194 98 −442 −147 −212 −84
USA 412 231 −284 −107 −103 −43
UK 229 117 −495 −158 −313 −115
Japan 66 31 −64 −28 12 5
Other 126 61 −222 −87 −56 −24
countries
in the
sample
Source: Authors’ calculation
Other countries in the sample
Japan
United Kingdom
United States
Eurozone
-600 -500 -400 -300 -200 -100 0 100 200 300 400 500
Change in Predicted Change in Predicted

3-month change in 3-month change in
interest net interest interest net interest
rate margin rate margin
2008-2014 2008-2014 2003-2007 2003-2007
Fig. 2.5 Observed changes in interest rates and predicted changes in the net
interest margin (bp). Source: Authors’ calculations
measures taken, which led to a fall in the net interest margin of 147 bp in
the Eurozone, 107 bp in the USA, 158 bp in the UK, 28 bp in Japan
and 87 bp in the group “other countries”. For the entire period analysed,
the total effect of the variation in interest rates on the intermediation
margin was a fall of 84 bp in the Eurozone, 43 bp in the USA, 115 bp in
the UK, 24 bp in the group “other countries”, and an increase of 5 bp in
Japan.
2.6 Conclusions
A cause for concern today is the impact that unconventional monetary
policy measures adopted by several central banks to combat the crisis
could have on bank interest margins and thus on the profitability.
Although the effect has been positive so far, the prolonged low level of
interest rates in some countries (as is the case with those belonging to the
Eurozone) might end up negatively affecting the intermediation margin,
given the existence of a floor in the level of interest rates on bank
deposits. The quadratic, rather than linear, relationship between net
interest margin and interest rates mean that a further drop in rates will
damage profitability.
In this context, the results obtained in this study for a large sample of
banks in OECD countries for the period 2003–2014 confirm that the
above-mentioned quadratic relationship does indeed exist. This indicates
that the impact of a variation in interest rates is higher for low levels than
for high values. Consequently, if this current scenario of very low-interest
rates persists over time (and even worse, if there is a further drop),
banking margins could be adversely affected and therefore, profitability.
This result is in line with the evidence obtained recently by Borio et al.
(2015) and Claessens et al. (2016), who also obtained a positive quad-
ratic relationship between net interest margin and the level of short-term
interest rates.
An important implication of economic policy regarding the results
obtained is that there is a trade-off between economic growth and
financial stability associated with the impact of expansionary monetary
policy when the level of interest rates is very low. Thus, while on the one
hand expansionary measures are adopted to combat the crisis (increasing
the rate of inflation and encouraging economic growth), the negative
impact on the net interest margin also negatively affects the profitability
of banks, thus increasing the likelihood of a systemic crisis.
In this context, of particular concern is the case of the banks in the
Eurozone, which currently have a problem with low profitability as a
consequence of the regulatory pressure and the high amount of
non-performing assets. The fact that the inflation rate is well below the
ECB target of 2% justifies the expansionary measures taken (such as the
expanded asset purchase programme (APP) and the penalty of up to
−0.4% of excess of reserve requirements and deposit facility). But taking
into account the results obtained in this paper, these same measures can
have a negative impact on bank profitability. This explains the IMF’s
recent warning (2016) not to further increase the negative interest rates
on marginal deposit facility and excess reserves. Until now the expan-
sionary monetary policy has stimulated the volume and quality of bank
lending and, by this way, profitability. But now that interest rates are so
low (even negative), monetary policy is holding back banks’ profitability.
Notes
1. See Laubach and Williams (2015).
2. In the same vein, the recent study by Borio and Zabbai (2016) analyses
both the negative and the positive effects of unconventional monetary
policy measures that are being adopted. The authors conclude that
although there is evidence that these measures are successful in improving
financial conditions, over time they could have a negative impact on bank
profitability.
3. Australia, Austria, Belgium, Canada, Colombia, Czech Republic,
Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Israel,
Italy, Japan, Rep. Korea, Latvia, Netherlands, New Zealand, Norway,
Poland, Portugal, Russian Federation, Slovak Republic, Slovenia, South
Africa, Spain, Sweden, Switzerland, UK and USA.
4. The price of labour is approximated by the ratio of Staff costs/total assets.
5. The ratio own resources/assets is a capitalisation measurement with lim-
itations, due to the influence of regulation on own resources, as a measure
of risk aversion. Therefore, the results should be interpreted with caution.
Acknowledgments Authors gratefully acknowledge comments of an anony-

mous referee, the useful comments of Alin Marius Andrieş (discussant of the
paper in the Wolpertinger Conference 2016) and the financial support of the
Spanish Ministry of Science and Innovation (research project
ECO2013-43959-R). Joaquin Maudos also acknowledges financial support of
Generalitat Valenciana (research project PROMETEOII/2014/046). Paula
Cruz-García also acknowledges financial support of Spanish Ministry of
Education (FPU2014/00936).
References
Alessandri, P., and B. Nelson. 2015. Simple banking: Profitability and the yield
curve. Journal of Money, Credit and Banking 47 (1): 143–175.
Allen, L. 1988. The determinants of bank interest margins: A note. Journal of
Financial and Quantitative Analysis 23 (2): 231–235.
Angbazo, L. 1997. Commercial bank net interest margins, default risk,
interest-rate risk and off-balance sheet banking. Journal of Banking & Finance
21: 55–87.
Arellano, M., and S. Bond. 1991. Some tests of specification for panel data:
Monte Carlo evidence and an application to employment equations. Review
of Economic Studies 58: 277–297.
Berg, S.A., and M. Kim. 1994. Oligopolistic Interdependence and the Structure
of Production in Banking: An Empirical Evaluation. Journal of Money, Credit
and Banking, 26 (2): 309–322.
Blundell, R., and S. Bond. 1998. Initial conditions and moment restrictions in
dynamic panel data models. Journal of Econometrics 87 (1): 115–143.
Borio, C. E. and A. Zabai. 2016. Unconventional monetary policies: A
re-appraisal. BIS Working Papers, 570.
Borio, C. E., L. Gambacorta, and B. Hofmann. 2015. The influence of
monetary policy on bank profitability. BIS Working Papers, 514.
Busch, R. and C. Memmel. 2015. Banks’ net interest margin and the level of
interest rates, Discussion Papers 16/2015, Deutsche Bundesbank, Research
Centre.
Carbó, S., and F. Rodriguez. 2007. The determinants of banks’ margins in
European banking. Journal of Banking & Finance 31: 2043–2063.
Claessens, S., N. Coleman, and M. Donnelly. 2016. Low-for-long interest rates
and net interest margins of banks in advanced foreign economies. IFDP
Notes.
Entrop, O., C. Memmel, B. Ruprecht, and M. Wilkens. 2015. Determinants of
bank interest margins: Impact of maturity transformation. Journal of Banking
& Finance 54: 1–19.
European Central Bank. 2016. Annual Report.
Genay, H., and R. Podjasek. 2014. What is the impact of a low interest rate
environment on bank profitability? Chicago Fed Letter, (Jul).
Gerali, A., S. Neri, L. Sessa, and F. Signoretti. 2010. Credit and banking in a
DSGE model of the euro area. Journal of Money, Credit and Banking 42:
107–141.
Ho, T., and A. Saunders. 1981. The determinants of banks interest margins:
Theory and empirical evidence. Journal of Financial and Quantitative Analysis
16 (4): 581–600.
International Money Fund. 2016. Negative interest rate policy (NIRP):
Implications for monetary transmission and bank profitability in the Euro
Area. In Euro Area Policies Selected Issues, IMF Country Report No. 16/220.
Laubach T. and J. Williams. 2015. Measuring the natural rate of interest redux,
Federal Reserve Bank of San Francisco. Working Paper, No. 2015–16,
October.
Maudos, J., and J. Fernández de Guevara. 2004. Factors explaining the interest
margin in the banking sectors of the European Union. Journal of Banking &
Finance 28 (9): 2259–2281.
McShane and Sharpe. 1985. A time series/cross section analysis of the
determinants of Australian Trading bank loan/deposit interest margins:
1962–1981. Journal of Banking & Finance, 9: 115–136.
Rostagno, M., U. Bindseil, A. Kamps, W. Lemke, T. Sugo, and T.
Vlassopoulos. 2016. Breaking through the zero line: The ECB’s negative
interest rate policy. Presentation at Brookings Institution seminar negative
interest rates: Lessons learned… so far, Washington D.C., June 6, 2016.
Saunders, A., and L. Schumacher. 2000. The determinants of bank interest rate
margins: An international study. Journal of International Money and Finance
19: 813–832.
Wong, K.P. 1997. On the determinants of bank interest margins under credit
and interest rate risk. Journal of Banking & Finance 21: 251–271.
Zarruk, E.R. 1989. Bank margins with uncertain deposit level and risk aversion.
Journal of Banking & Finance 13: 797–810.
3
The Swedish Mortgage Market: Bank
Funding, Margins, and Risk Shifting
Viktor Elliot and Ted Lindblom
3.1 Introduction
The Swedish household indebtedness is among the highest in the world
(BCBS 2016), and the vast majority of this debt is mortgages.1 The
Swedish Bankers’ Association (SBA 2015) reports that almost nine out of
ten homeowner households are indebted, and the debt ratio of these
households, measured as loans in relation to disposable income, has on
average increased steadily since the mid-1990s. According to the Swedish
Financial Supervisory Authority (SFSA 2016), in the first half of this
decade, the share of households with a greater debt ratio than 450% has
almost doubled (from *35% in 2011 to *60% in 2015). This is
explained partly by the conversion of rental apartments, primarily in
urban locations, into condominiums and, partly, as a result of rocketing
property prices. Residential construction in urban areas has not been in
pace with the growth of citizens in these areas, while the interest rate
V. Elliot (&) T. Lindblom

University of Gothenburg, Gothenburg, Sweden
e-mail: viktor.elliot@gu.se
36 V. Elliot and T. Lindblom
environment has been and still is historically low. In the past 2 years, the
repurchasing rate (repo rate) of the Swedish Central bank has been
continuously cut and is, since February 1, 2015, below zero. In addition,
the average monthly interest rates of treasury bills, 2-year government
bonds, and, lately, 5-year government bonds are negative.
The extremely low interest rate environment has put pressure on
mortgage interest rates offered by Swedish banks. At present, both the
short- and the medium-term mortgage rates are down to the crisis and
post-crisis levels in 2009–2010. The long-term rates are at an all-time
low (in modern time) and substantially lower (150–200 basis points)
than they were 6 years ago. Mortgage-lending accounts for approxi-
mately 30% of total lending provided by the major Swedish banks, and
the banks have readily accommodated the growing demand for debt on
the household market. In order to meet the increasing demand for
household debt, the banks have gradually increased their reliance on
market funding, a strategy that ceteris paribus implies higher funding
costs and potentially lower margins. Considering the great importance of
mortgage lending in the balance sheets of the major Swedish banks, it
may appear as a paradox that these banks on average show high prof-
itability in accounting as well as in market value terms. This suggests that
the banks have been able to secure lower funding costs in some other
way. As will be shown in this chapter, the implementation of the covered
bonds legislation [see “the Covered Bond (Issuance) Act (2003:1223”)], on
the Swedish mortgage market in 2004, seems to be one major expla-
nation for their profitability.
In 2006–2008, Swedish banks gradually replaced residential
mortgage-backed bonds (MBS) with covered bonds (CB). The market
for CBs has thereafter grown large in Sweden, making CBs one of the
most important sources of funding for Swedish banks (Sandström et al.
2013).2 Today, about a quarter of the banks’ average total lending is
financed by CBs. Certain properties (these are discussed in greater detail
below) of the CBs make them “often seen as close substitutes for high-
quality government bonds” (Prokopczuk et al. 2012: 1), suggesting lower
risk for investors and lower risk premiums to be paid by the issuing bank.
Hence, the bank can offer homeowners lower interest rates on granted
mortgage loans and still make a “good” profit. This may seem as a
3 The Swedish Mortgage Market … 37
“win-win-win” situation, but it is questionable whether it is sustainable

in the long run. As noted by (Carbó-Valverde et al. 2012: 2) “…banks
might not view MBS and CB as substitutes since there are some real and
regulatory differences between issuing MBS and issuing CB.” We will argue
throughout this chapter that these differences are fundamental to
understand the risk shifting on the Swedish mortgage market and why
Swedish banks have been able to maintain, or even increase, their mar-
gins on mortgages over the past decade.
More specifically, we aim to compare Swedish banks’ mortgage
lending and funding rates over a period of 15 years in order to illustrate
changes in risk and bank mortgage margins stemming from the financial
crisis and the move from the MBS regime to the CB regime. The study
seeks to contribute to the ongoing debate of whether Sweden is heading
for another real estate-related financial crisis. In Sect. 2, we outline the
key characteristics of the MBS and the CB as well as discuss briefly the
Swedish context and the increasing use of CBs on the Swedish market.
Section 3 describes our method, and the results of our analysis are
reported and discussed in Sect. 4. Section 5 concludes the chapter.
3.2 Covered Bonds—Essential Features

To understand the effects on risk shifting and profitability for banks
when moving from the MBS to the CB as one of the key mortgage
funding sources,3 the first part of this section briefly compares the two
securities. In many respects, CBs are similar to MBSs (Carbó-Valverde
et al. 2012). Both have fixed maturities, their principal amount (face
value) is repaid at maturity, and they are collateralized by a pool of
underlying assets primarily in the form of residential mortgages. The
most distinguishing feature differentiating the CB from the MBS is that
the former is held on the balance sheet, whereas the latter is not (Larsson
2013). This means that the holder of a CB retains a dual recourse, i.e., a
high-priority claim on the assets that serves as collateral in the cover pool
and an unsecured claim on the assets of the issuing institution (the
originator) in case of default (Schwarcz 2011, 2013; Martín et al. 2014).
In addition, different from the MBS, the CB cover pool is dynamic
Over-
collateralization House values = 100
25 25 25
75 75 75 Loan Loan Loan

Covered Bonds 75 75 75
(Issued amount)
Cover pool
assets
25 25 25
75 75 75 Loan Loan Loan

75 75 75
Fig. 3.1 Illustration of overcapitalization of CB issues
(Martín et al. 2014), requiring the issuer to continuously replace insuf-

ficient (low-quality) assets in the cover pool by assets of adequate quality
over its full lifetime. The implied “overcollateralization” in terms of “a
surplus of collateral over indebtedness” (Schwarcz 2013: 143) of the CB is
illustrated in Fig. 3.1. As banks are also substituting prepaid and/or
defaulted mortgages with new loans, it “keeps the size of the pool pre-
dictable” (Carbó-Valverde et al. 2012: 7).
Finally, CBs require frequent arm’s-length assessments of the issuing
banks’ management of the cover pool by a third party. This party is
either appointed by the issuer for approval by the regulator (i.e., under a
special law-based framework, which is also referred to as a “legislative”
CB regime) or stipulated by a contractual agreement (i.e., under a general
law-based framework or a “structured” CB regime).
Over the past decade, the CB market has grown tremendously (at least
until 2013 from when it has decreased somewhat). At the end of 2015,
the global CB market included 30 countries with aggregate outstanding
volumes of approximately EUR 2500 bn. The vast majority of the CB
market is located in Europe, but in recent years countries such as
Canada, Australia, New Zealand and Singapore have implemented reg-
ulatory frameworks allowing for CB issuance (ECBC 2016).
3.2.1 The Swedish Market for CBs
The debate about whether to allow CB issuance in Sweden was initiated

in 1993 when the Swedish Bankers Association (SBA) submitted a
white-collar paper to the Treasury Department in Sweden. It was argued
that “Golden Bonds,” as they were commonly called during that time,
could mitigate a shortage of funding vehicles for large investors (primarily
insurance firms) that was expected to rise when new legislation was
introduced in 1994. The SBA suggested that Sweden would follow the
Danish (and German) example and allow single investors an exception to
the EU regulations for large exposures specifically for these golden bonds.
In 1996, a government inquire was issued to shed further light on
whether new legislation related to golden bonds was to be introduced.
The report that followed the inquiry (SOU 1997) concluded that there
were no major reasons to allow golden bonds issuance, and the intro-
duction was put on hold. The SBA continued to argue for CBs, and in
2003, the legislation was changed to allow CB issuance in Sweden.
Like in many other countries in the EU, CBs are issued under a
legislative regime in Sweden. This suggests that there is “a high degree of
certainty regarding the investors’ legal rights and responsibilities in the event
of issuer insolvency, and lower transaction costs in structuring a covered bond
transaction” (Schwarcz 2011: 569). It also means that issues of CBs, as
well as measures that might have to be taken during their lifetime, are
more rigid than under the alternative structured CB regime.4 At the
beginning of 2006, there were three issuing financial institutions on the
Swedish CB market (SBA 2015). The following year three additional
institutions joined, and at present, there are eight issuing financial
institutions on the Swedish CB market. As shown in Fig. 3.2, the out-
standing stock of CBs has more than quadruplet since 2006, making
Sweden the fifth largest CB market in the world in terms of volumes.
In 2014, the four largest banks in Sweden accounted for 83% of the
total stock. Three of these banks issue CBs through their own “building
societies,” which are sanctioned by the SFSA to act as issuers of CBs. As
illustrated in Fig. 3.3, in this model, the funds obtained by the issuing
2,50,000
2,00,000
Denominated in other
1,50,000 currencies
Denominated in
1,00,000 EURO
Denominated in
50,000 domestic currency
0
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Fig. 3.2 Growth (in mEuro) of outstanding CBs on the Swedish market since 2006.
Source Data from the Association of Swedish Covered Bond issuers
Homeowner Parent bank

Asset transfer
Issuing institution
Collateral
Investors Covered bonds
Fig. 3.3 Indirect issuing of CBs through a bank-owned “building society”
institution are transferred to the parent bank for financing its mortgage
lending to primarily homeowners.
The remaining large bank and some of the other institutions are
issuing CBs directly by themselves. Their model is illustrated in Fig. 3.4.
Financial institutions that issue CBs on the Swedish market are legally
required to act as market makers on the secondary market for out-
standing CBs. In order to reduce the risk of not being able to refinancing
maturing CBs, it is common that issuing institutions repurchase a large
share of the CBs about 9–12 months before the maturity date. This is
commonly done by offering investors to exchange maturing bonds for
bonds with a longer maturity and (likely) another interest rate
Bank/Issuer
Homeowner
Collateral
Investor Covered bonds
Fig. 3.4 Direct issuing of CBs through a bank-owned “building society”
Bank A Bank B
Building society:
Market maker issuer of covered
bonds Market maker
Investor 1 Investor 2
Fig. 3.5 Illustration of the roles of actors involved in CB issues. Source Based on
Sandström et al. (2013: 8)
(Sandström et al. 2013). The roles of the different actors on the sec-
ondary market are illustrated in Fig. 3.5.
As shown in Table 3.1, the Swedish CBs are highly rated by inter-
national rating agencies. Due to the high credit ratings, the CBs are
expected to be traded at prices (interest rates) at par with or close to
government bonds (cf. Prokopczuk et al. 2012). The diagram in Fig. 3.9
(see Appendix) reveals that this has been only partly true since these
bonds were introduced on the Swedish market. From 2004 to mid-2007,
the spreads between 2- and 5-year CBs and government bonds were very
small. At some occasions, the risk premium of the 2-year CBs was even
Table 3.1 Credit ratings of CBs issued by Swedish institutions

Moody’s Standard and Poor’s
Swedbank AAA Aaa
Handelsbanken – Aaa
Landshypotek AAA/Stable –
Länsförsäkringar Bank AAA/Stable Aaa/Stable
Nordea AAA/Stable Aaa/Stable
SBAB/SCBC – Aaa
SEB – Aaa
Skandiabanken – Aaa
Source Collected from the webpage of each institution, 2016-09-28
negative. However, after mid-2007, the average risk premiums of both

maturities have, with only a few exceptions, been greater than 50 and
even up to almost 200 basis points. Even during the last 2–3 years, when
Swedish banks have been rather frequently reported as stable, strong, and
solid banks, the risk premiums seem to settle around 50 basis points for
the 2-year CBs and slightly higher for the ones of 5-year. It should be
noted, though, that both are currently traded at extremely low and
occasionally even negative interest rates. It is today more expensive for
banks to finance their mortgage lending through deposits and savings—
even before considering operational transaction costs. How this has
affected the mortgage rates offered by the banks is examined in Sect. 4.
3.3 Sources of Data and Selection of Banks

There is no comprehensive database covering the Swedish mortgage
market, the CB market and the banks mortgage rates together.
Accordingly, we have relied on several different sources to collect the data
for our analysis including the Swedish Bankers’ Association (SBA), the
Swedish Central bank (Riksbanken), the Swedish Financial Supervisory
Authority (SFSA), Statistics Sweden (SCB), and the Swedish banks.
While volumes and bond rates are actual, it should be noted that there
is no publically available data on actual mortgage interest rates.5 To
mitigate this problem, we have relied on the officially offered interest
rates by the banks as a proxy for actual rates paid. However, since
mid-2015, Swedish banks are by law required to reveal their actual

average mortgage rates on a monthly basis providing some indication of
the rates that consumers of mortgages actually pay (this is further dis-
cussed in the next section).
We have selected five banks [Swedbank, Svenska Handelsbanken
(including Stadshypotek), Skandinaviska Enskilda Banken (SEB),
Nordea and SBAB (including Swedish Covered Bond Corporation)]
which together dominate the Swedish markets for residential mortgages
and CBs. The combined market shares of these five banks summed up to
approximately 90% on both markets in 2014, and their market shares
have remained stable over the time period studied, i.e., 2000–2016 (SBA
2015). These five banks also offer publicly available data about their
officially offered mortgage rates over the full period (except Svenska
Handelsbanken from which data are available between 20056 and 2016).
Table 3.2 offers an overview of importance of mortgage lending (be-
tween 30 and 70% of total lending to the public are residential mort-
gages) and CB funding in the selected banks.
In the next section, we analyze the interest-rate environment on the
Swedish mortgage market focusing on bank margins and risk shifting.
Because of the dispersion of data sources that our analysis is based on we
try to, as far as possible, clarify the limitations of the data in connection
to each of the figures presented.
Table 3.2 Key figures (in SEKm) of selected banks as of December 31, 2015
Lending Deposits Outstanding Equity Total
to the from the CBs assets
public public
Nordea 3,132,884 1,776,716 1,024,762 285,169 5,944,393
Handelsbanken 1,866,467 753,855 603,952 128,268 2,522,133
SEB 1,353,386 883,785 310,178 142,798 2,495,964
Swedbank 1,413,955 748,271 531,219 123,342 2,148,855
SBAB 296,981 76,639 187,280 11,848 374,552
Total 8,063,673 4,239,266 2,657,391 691,425 13,485,897
Note that these are group level figures and include the banks’ foreign operations
Source Data from the Swedish Bankers Association and the Annual Reports of
individual banks
3.4 Mortgage Rates of Swedish Banks

In the new Millennium, banks operating on the Swedish market have
increased their lending to households, non-financial business firms and
other organizations in each year but 2009. In total, the banks’ lending
has more than tripled between 2000 and 2015. This corresponds to an
average annual increase by almost 8%. Approximately, half of the lending
is used to finance private homes through mortgages. This lending has in
fact increased annually without any interruption during the specified
period of study. Residential mortgages are offered to variable (3 months)
as well as fixed interest rates over various maturities (1–10 years).7 In the
Appendix, Fig. 3.10 displays how the interest rates, officially offered by
the five major Swedish banks that grant residential mortgages, on average
have developed until today for 3-month, 2, 5, and 10-year fixed mort-
gage rates, respectively.
The diagram in Fig. 3.10 exhibits a clear tendency of decreasing
official interest rates on both short- and long-term residential mortgages.
On average, the offered interest rates by the five banks did increase
between mid-2005 and the financial crisis, but since mid-2011, their
average interest rates have declined steadily. A more detailed analysis,
which is displayed in Fig. 3.11, reveals that the offered interest rates by
the individual banks deviated relatively more at the beginning of the
period. The diagrams in the figure show that this has been the case for
the offered short-term rates, in particular. The officially offered 3-month
rates of the banks differed markedly (up to 85 basis points between the
lowest and highest offered average monthly rate) until October 2007.
Subsequently, these rates have barely deviated at all until April 2015.
Except for a few single months, the standard deviation did on a monthly
basis vary between 1 and 10 basis points only. The fixed 2-year and
5-year interest rates offered by the banks display a similar pattern—albeit
not as pronounced and to the end of 2004 only. After the introduction of
CBs, these rates have converged and with a few exceptions barely differed
until April 2015. The fixed 10-year mortgage rates have been less dif-
ferentiated during the lion part of the period. On average, the 10-year
mortgage rates show the largest deviation in 2000 and in the aftermath of
the financial crisis from 2009 to 2011. Like the other officially offered
mortgage rates, from April 2015 and onward the banks’ highest and
lowest offered 10-year mortgage rates have differed more extensively.
3.4.1 Residential Mortgage Rates and Bank Margins
Figure 3.7 displays the average discount given by the banks from May
2015 to May 2016. We have no information of the average discounts on
the banks’ mortgage rates prior to 2015, let alone to the actual interest
rates paid by their customers. It seems reasonable to assume that dis-
counts have been given in a similar range as displayed in Fig. 3.6, but it
is less likely to assume that the largest discounts are given on shorter-term
lending rates. In a less low rate environment, the opposite may very well
be the case.
Assuming that Figs. 3.10 and 3.11 reflect the interest rates charged by
the banks on their mortgage lending accurately in relative terms; com-
petition in banks’ mortgage lending at the beginning and toward the very
end of the period seems rather week. This observation is in accordance
with the SBA (2011) report in which it is concluded that the banks’
mortgage lending increased markedly from 2000 to 2007. Under this
period, the market shares of the five banks have remained intact.
Average discounts on mortgage rates

0.5
0.45 3 Month 2 Year 5 Year 10 Year
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
Fig. 3.6 Average discounts on the banks’ officially offered mortgage rates the
past year
100%
90%
80% Equity
70%
Other debt
60%
50% Derivatives
40%
Bonds
30%
20% Intermediate
funding
10% Deposits
0%
2010
2013
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2011
2012
2014
2015
Fig. 3.7 Aggregate bank funding 1996–2015. Source Data from the Swedish
Bankers Association
Derivatives/ Deposits Derivatives/ Deposits

other other
Business Business
lending lending
Intermediate Intermediate
funding funding
Derivatives
Other debt
Mortgage Bonds Mortgage
lending lending Equity
Covered
Derivatives bonds
Other debt
Equity
Pre covered bonds Post covered bonds
Fig. 3.8 Pre- and post-CB claims by different bank funds providers on bank assets
in case of bankruptcy
Average market rates

8.00
GB 2-Year CB/MB 2-Year
6.00
4.00
2.00
0.00
8.00
GB 5-Year CB/MB 5-Year
6.00
4.00
2.00
0.00
Average risk premiums on covered/mortgage-based bonds

2 Year 5 Year
2.00
1.50
1.00
0.50
0.00
-0.50
Fig. 3.9 Average market rates and risk premiums on 2- and 5-year covered bonds
(CB/MB) relative government bonds (GB) for 2000–2016
9.00
3 Month 2 Year 5 Year 10 Year
8.00
7.00
6.00
5.00
4.00
3.00
2.00
1.00
0.00
Fig. 3.10 Average official mortgage rates of the five Swedish “mortgage” banks
from 2000 to 2016
However, on average, their lending portfolio changed dramatically. First,

the share of renewed mortgage loans with variable rates declined to
around 50% or less. This share increased again during 2008 and, par-
ticularly 2009, in which year the proportion of new mortgage loans with
a variable (3-month) interest rate peaked at 85%. Second, mortgage loans
with fixed medium-term interest rates over one to 5 years gained pop-
ularity from 2000 to 2007. In contrast to this development, Fig. 3.12
3 Month Mortgage rate Max Min

1.00
0.50
0.00
-0.50
2 Year Mortgage rate
0.50
0.00
-0.50

1.00
0.00
-1.00

1.00
0.00
-1.00
Fig. 3.11 Deviations of the banks’ officially offered mortgage rates from 2000 to
2016
suggests that the average bank earned more on loans with a “variable”
interest rate during this period. From early 2001 to the end of 2007, the
average margin on 3-month mortgage loans was equal to or greater than
the corresponding margins on medium- and long-term fixed rate loans.
Irrespective of the interest rate maturities, the average mortgage margins
have increased significantly from the end of 2011. To what extent this is
the case does of course depends on the actual interest rates charged by the
banks. Considering their average discounts, all mortgage rate margins
become lower. At the end of the period, the average interest rate margins
on the banks’ mortgage loans with long-term fixed rates became more
attractive for the banks in relative terms. On average, their greatest
margins were still on the 3-month and 2-year interest rate maturities.
Our comparative analysis hitherto rests on the assumption that the
average bank continuously matches the interest rate maturities on its
borrowing and lending. However, the financial crisis clearly showed that
Swedish banks were in general borrowing short and lending long (cf.
Lindblom et al. 2011). In Fig. 3.13, the average interest rate margins for
mortgage loans are computed as the difference between the offered
mortgage rates on the loans of the average bank and the 3-month
Average lending rate margins

2.50
3 Month 2 Year 5 Year 10 Year
2.00
1.50
1.00
0.50
0.00
Fig. 3.12 Marginal mortgage lending margins of the average bank from 2000 to
2016
Average interest rate margins if borrowing short and lending long

5.00
3 Month 2-Year 5-Year 10-Year
4.50
4.00
3.50
3.00
2.50
2.00
1.50
1.00
0.50
0.00
Fig. 3.13 Illustration of average interest rate margins if borrowing short and
lending long
funding rate. Figure 3.13 demonstrates what can be interpreted as a

temptation for borrowing short and lending long. Instead of interest rate
margins around 1%, the banks would get considerably greater margins
by financing medium- and long-term fixed rate mortgages with
short-term funds. These funds include ordinary savings and deposits,
which constitute important sources of funds for Swedish banks.
The diagram in Fig. 3.13 suggests that it would be profitable for the
average bank to borrow short and lend long during the first years after the
outbreak of the financial crisis, too. However, the banks almost lent to
the 3-month rate only in 2009, and thereafter, they seem to have adapted to
the gradually implemented new regulatory regime, which is focused on
both sustainable capital adequacy and liquidity reserve requirements for
banks. Figure 3.12 shows that the lending margins of the average bank have
in general been around 1.5–2% from early 2012 and onward. In pace with
the increasingly lower interest rates on government securities and out-
standing CBs, the banks seem to have been able to increase their average
margins steadily on mortgage loans with different interest rate maturities.
3.4.2 Risk Shifting and CB Funding on the Swedish

Mortgage Market
In some countries, like Canada and Australia, the local regulators impose
a cap on CB issuance in order to limit the subordination of depositors to
CB investors. As exemplified by Fig. 3.7, Sweden has followed the
Danish and German tradition without any caps. The figure shows that
starting in 2006, CBs have gradually replaced intermediate funding
among the Swedish banks. The figure clearly shows that this trend started
prior to the financial crisis (2008–2009) and the introduction of new
liquidity regulations (2013) in Sweden.
To illustrate the risk shifting that this change in funding imposes, Fig. 3.8
compares the pre- and post-CB claims by different bank funds providers on
bank assets in case of bankruptcy. As shown in the figure, the non-CB funds
providers are made substantially worse off in a bank insolvency situation and
as will be further discussed in our concluding remarks, and this should be
reflected in the price that banks have to pay for non-CB funds.
3.5 Concluding Remarks

This paper set out compare Swedish banks’ mortgage lending and
funding rates over a period of 15 years. By illustrating changes in risk and
bank mortgage margins stemming from the financial crisis and the move
from the MBS regime to the CB regime, we seek to contribute to the
ongoing debate on whether Sweden is heading for another real
estate-related financial crisis. Our instruments are the lending and

funding rates, and our analysis points to at least three key insights.
Firstly, the move from MBS to CBs means that the funding costs
should, all else equal, go down for the Swedish banks. While there may
exist systemic uncertainties related to the shifting of risk (through
seniority), the banks seem to have been able to translate the reduced
funding costs into higher margins on mortgages. From a crisis perspec-
tive, this may be seen as a good thing as the banks are able to charge
higher premiums, which will work as an extra cushion during a crisis.
Secondly, prior to the crisis banks relied heavily on short-term funding to
finance mortgages. The lessons from the 2007/2009 financial crisis, in
combination with a stricter regulatory framework, have led to a better
matching between the asset and liability side of their balance sheets. This
process should be costly, both in terms of funding and administrative costs.
However, the banks have been able to internalize these costs and still
remain highly profitable, and we show that this seems to work through two
interrelated channels—the move to a CBs regime as discussed above and
through limited competition. The latter is also supported by our previous
work (see Elliot 2015; Elliot and Lindblom 2015, 2016).
Thirdly, some propagators of a soon to strike financial crisis in Sweden
have compared the current situation to the period leading up to the Swedish
financial crisis in the early 1990s. However, several important differences
have been discussed that may have implications for whether this claim is true
or not: (i) the indebtedness growth is much larger this time, (ii) the currency
is no longer fixed, (iii) the inflation is close to zero and interest rates are either
close to or even below zero, and (iv) the financial service sector is currently
moving toward a much stricter regulatory regime, whereas the late 1980s was
marked by deregulation. We add to these differences by providing evidence
that the banks, in contrast to the period prior to crisis in the 1990s, have been
able to maintain high margins while also reducing their risk (through higher
capital ratios, more liquid assets and better matching between assets and
liabilities). Thus, even though we do not wish to make a claim as to whether a
new crisis is on the horizon, this paper indicates that the Swedish banks are at
least better prepared this time.
We strongly suggest that future research extends the analysis related to
the risk shifting outlined in this chapter. It seems particularly relevant to
analyze whether non-CB providers of bank funds charge a higher price

when banks replace MBS funding with CB funding. Such analysis will
need to account for the fact that depositors may be insensitive to bank
credit risk because of deposit insurance schemes. Accordingly, we suggest
the analysis to focus on subordinated bank debt (such as contingent
convertibles and bank equity).
Notes
1. There exists no systematic data on the composition of household debt but
Statistics Sweden estimates that approximately 80% of total household
indebtedness is mortgages.
2. According to ECBC (2014), the Swedish CB market belongs to the five
largest in the world and in 2013; Sweden was the second largest issuer of
new covered bonds.
3. A mortgage loan in Sweden typically consists of four key components:
CBs (both in SEK and Euro), other bonds, deposits and equity, and the
vast majority of bank funding consist of CBs and deposits.
4. As is described and thoroughly explained by, e.g., Schwarcz (2011), both
the legislative and the structured covered bond regimes are each subject to
pros and cons. The structured covered bond regime may be less certain
regarding legal aspects (which are to be contractually specified) and give
rise to higher transaction costs, but this regime tends to offer greater
flexibility for the parties to, for example, adjust to changing market
conditions.
5. Private as well as business customers commonly negotiate discounts in the
range of 10–100 basis points.
6. From 2005, the 3-month rates are available, whereas the other maturities
are made available from 2008.
7. After the financial crisis, banks no longer offer mortgage rates on a daily
basis. In our analysis, the 3-month rate is regarded as a variable rate.
References
BCBS. 2016. BIS Statistical Bulletin, available at https://www.bis.org/statistics/
bulletin1612.htm.
Carbó-Valverde, Santiago, Richard J. Rosen, and Francisco Rodríguez-Fernández.

2012. Are covered bonds a substitute for mortgage-backed securities? Federal
Reserve Bank of Chicago, WP 2011–2014.
ECBC. 2014. ECBC Publication—September—9th edition.
ECBC. 2016. ECBC Publication—August—11th edition.
ECBC. http://ecbc.hypo.org/Content/Default.asp?PageID=503.
Elliot, V. 2015. Performance management systems, regulation and change in
Swedish Banks. Dissertation. BAS Publisher: Gothenburg.
Elliot, V. and Lindblom, T. 2015. Funds transfer pricing in banks: Implications
of Basel III. In Performance Management Systems, Regulation and Change in
Swedish Banks, ed. V. Elliot. Dissertation, BAS Publisher: Gothenburg.
Elliot, V. and Lindblom, T. 2016. Funds transfer pricing in banks under
different market structures. Working paper.
Larsson, C.F. 2013. What did Frederick the great know about financial
engineering? A survey of recent covered bond market developments and
research. The North American Journal of Economics and Finance 25: 22–39.
Lindblom, T., M. Olsson, and M. Willesson. 2011. Financial Crisis and Bank
Profitability. In Bank Performance, ed. P. Molyneux, 83–105. Risk and Firm
Financing, Basingstoke, UK: Palgrave Macmillan.
Martín, Rebeca A., José M. Marqués Sevillano, and Luna R. González. 2014.
Covered bonds: The Renaissance of an old acquaintance, Banks and Bank
Systems 9 (1): 46–60.
Prokopczuk, Marcel, Jan B. Siewert, and Volker Vonhoff. 2012. Credit Risk in
Covered Bonds (working paper, September 20).
Sandström, Maria, David Forsman, Johanna Stenkula von Rosen, and Johanna
Fager Wettergren. 2013. Marknaden för svenska säkerställda obligationer och
kopplingar till den finansiella stabiliteten, Penning- och valutapolitik, vol. 2.
Schwarcz, Steven L. 2011. The Conundrum of Covered Bonds, The Business
Lawyer 66: 56.
Schwarcz, Steven L. 2013. Securitization, Structured Finance, and Covered
Bonds. The Journal of Corporation Law 39 (1): 129–154.
SBA. 2011. Bank- och finansstatistik. Swedish Bankers’ Association: Stockholm.
SBA. 2015. Bank och Finansstatistik. Bankföreningen: Stockholm.
SFSA. 2016. Tillsynen över bankerna och kreditmarknadsföretagen, Dnr
15-15698.
SOU. 1997. Säkrare obligationer? Statens offentliga utredningar 1997:110.
Finansdepartementet: Stockholm.
4
Incapability or Bad Luck? Testing the “Bad
Management” Hypothesis in the Italian
Banking System
Fabrizio Crespi and Mauro Aliano
4.1 Introduction
The pattern of growth of nonperforming loans (NPLs) in the banking
sector of a country has always been considered an important issue in
determining the onset of a banking crisis and the consequent instruments
that should be used by authorities to prevent bank failures.
In this chapter, we add a contribution to the strand of literature
starting with Berger and DeYoung (1997), by testing the “bad man-
agement hypothesis” in the Italian banking sector using a more detailed
dataset about the composition of NPLs: That means that we can dis-
tinguish between substandard/past due loans and restructured exposures,
on the one hand, and bad loans on the other. This possibility allows us to
investigate if and how much the substandard/past due loans and
restructured exposures translate into bad loans over time or, put in a
F. Crespi (&) M. Aliano

University of Cagliari, Cagliari, Italy
e-mail: Fabrizio.crespi@unicatt.it
56 F. Crespi and M. Aliano
different way‚ we can prove if credit managers are able to recover, at least
partially, problematic credits. This part of the analysis is a novel in
literature: Preceding studies have not investigated in this specific way the
worsening in credit quality, which may indeed suggest the management
incapability, ceteris paribus, to manage bad loans. In this regard, starting
from our proprietary database, we insert a measure of bank credit health
among the determinants of NPLs which acts as proxy of a worsening in
the quality of credit.
We created a dataset composed of 48 banks of a single country (Italy)
which cover about 82% of the market (measured as total assets in 2013),
and we investigate a critical period for the Italian banking sector lasting
four years (2010–2013): This stretch of time covers in fact the preceding
period and the aftermath of the government debt crisis in Italy (2011–
2012) and corresponds with a strong increase in NPLs for the entire
banking system.
Considering that all banks in our sample operate mainly in Italy and are
subject to the same economic conditions, we exclude macroeconomic
explanatory variables from our analysis. We also prefer to investigate a
group of banks of one single country, avoiding a comparison with other
banking systems in which different economic conditions and accounting
rules could alter the significance of results (see Barisitz‚ 2011). The choice
of Italy as a laboratory for our analysis is due to the better information
about NPLs that we could get from bank balance sheets and to the critical
importance and enormous volume of NPLs in the Italian banking sector:
just to have an idea, the ratio of NPLs on loans was about 18% at the end
of 2015, compared with a 3% in France and a 2% in the USA.
The results of our study confirm the “bad management hypothesis”
first introduced by Berger and DeYoung (1997). Indeed, we find that
more specialized banks and more efficient banks tend to have a better
quality of credit. More interesting, from our point of view, is the general
relationship between substandard/past due loans, restructured exposures,
and bad loans. Our results indicate that problematic loans tend to
transform into bad loans, demonstrating that credit managers were not
able to implement recovery strategies for these critical positions during
the investigated period. We think our results can be considered an
4 Incapability or Bad Luck? Testing … 57
interesting contribution to the strand of the literature regarding NPLs

and bad management.
The rest of the chapter is organized as follows. Section 4.2 presents an
overview of the main strands of literature about the determinants of
NPLs. Section 4.3 briefly presents the characteristics of our dataset.
Section 4.4 describes the model and the variable used to investigate the
relationships between bank-specific determinants and NPLs. Section 4.5
presents the results of our analysis, and finally, Sect. 4.6 summarizes our
findings.
4.2 Determinants of NPLs: Main Strands

of Literature
The literature on NPLs is extensive and covers different topics regarding
the causes and effects of deterioration of credit quality. However, in the
majority of studies that investigate the determinants of NPLs, three
topics can be considered fundamentals. The first is related to a strand of
literature that investigates macroeconomic explanatory variables.
A second group of studies emphasize the effect of bank-specific charac-
teristics on problem loans; in particular, researchers have long since
investigated the relationship between NPLs and bank efficiency (usually
measured through a cost frontier approach). A third strand of literature
combines macroeconomic (country-specific factors) and microeconomic
variables (bank-specific factors) to explain aggregate NPLs.
4.2.1 Macroeconomic Factors
Various papers about NPLs start from the evidence that behind every
financial crisis there are macroeconomic factors (or systematic factors)
which influence the creditworthiness of borrowers. These studies com-
monly compare data from different countries (see Beck et al. 2013; Klein
2013) over a long period of time with the aim to evaluate the impact of
different phases of the economic cycle on the appearing of NPLs.
The main results of this part of literature can be summarized as fol-

lows. First of all, there is significant empirical evidence regarding the
anti-cyclical behavior of the NPLs, where real GDP growth can be
considered the main driver of NPLs, i.e., higher real GDP growth
translates into more income which improves the debt-servicing capacity
of borrowers. Conversely, when there is a slowdown in the economy, the
level of NPLs is likely to increase as unemployment rises and borrowers
face greater difficulties to repay their debt.
Beck et al. (2013), in a comprehensive study of 75 countries over a
ten-year period, confirm that real GDP growth has a negative impact on
NPLs when considered through a fixed effects model, but, using dynamic
Arellano-Bond estimations, also appears that lagged GDP growth sig-
nificantly affects NPLs with a positive sign; this finding suggests that
bank asset quality deteriorates with a lag in response to positive growth
due to loose credit standards applied during the boom period. In any
case, the authors affirm that economic activity is not able to fully explain
the evolution of nonperforming loans across countries and over time;
additional factors may negatively affect asset quality in countries with
specific vulnerabilities. For example, exchange rate depreciations lead to
an increase of nonperforming loans in countries with a high degree of
lending in foreign currencies to unhedged borrowers; further, an increase
in lending interest rates tend to increase NPLs.
Other macroeconomic variables, which were found to affect bank’s
asset quality, include disposable income (Rinaldi and Sanchis-Arellano
2006), lending interest rate and unemployment (Berge and Boye 2007),
and inflation (Klein 2013).
Rinaldi and Sanchis-Arellano (2006) try to understand what explains
household NPLs in seven euro area countries; their results suggest that,
in the long run, an increase in the ratio of indebtedness to income is
associated with higher levels of arrears. However, if the rise in the debt
ratio is accompanied by a rise in disposable income, the negative effect is
more than offset.
Similarly, Berge and Boye (2007) indicate that households’
debt-servicing capacity generally depends on developments in their
income, debt, borrowing rate, and collateral values. Higher incomes are
expected to contribute to reducing the volume of problem loans.
However, incomes may be unevenly distributed across households.

When unemployment is rising, many households may experience a
substantial reduction in income. Using an equilibrium correction model
of the logarithm of the share of problem loans (to total loans) in the
household sector in Norway, during the period 1993–2005, the authors
find that the share of problem loans could be reduced by 1.2% in the
long run if real disposable income increases by 1%. On the other hand, a
rise in the unemployment rate from 3 to 4% could increase the share of
problem loans by just over 11%.
4.2.2 Bank-Specific Factors
Even if macroeconomic factors are rightly considerable the main cause of

the increase of NPLs during time, they are not able to explain everything.
In the same country, and in the same phase of the economic cycle, banks
do normally register different amounts of NPLs in their balance sheets: It
should then be obvious to think that also bank-specific factors influence
bad loans. And in fact, various researchers have tried to discover signif-
icant relationships between NPLs and endogenous variables. The starting
point of this strand of literature (or at least one of the most important
contributions in this regard) can be considered the paper of Berger and
Deyoung (1997).
The authors draw attention to the links between bank-specific char-
acteristics and NPLs; using a Granger-causality analysis, they test a set of
hypotheses that describe the intertemporal relationship among problem
loans, cost efficiency, and financial capital. Specifically, they indicate four
possible mechanisms, namely “bad luck,” “bad management,” “skimp-
ing,” and “moral hazard,” to formulate predictions of the link between
credit quality and efficiency.
Under the “bad luck” hypothesis, external events precipitate an
increase in problem loans for the bank. The bank is then forced to
increase managerial effort and expenses dealing with the increase in
problem loans.1 Thus, under the bad luck hypothesis, we should expect
increases in NPLs to Granger-cause (i.e., temporally precede) decreases in
measured cost efficiency. Conversely, under the “bad management”
hypothesis, low cost efficiency Granger-causes larger amounts of problem

loans (a deterioration in asset quality) because management’s failure to
control operating costs immediately produces low cost efficiency, sug-
gesting that poor managerial practice causes an increase in problem loans
after a lag. The basic idea is that bad managers do not sufficiently
monitor and control their operating expense, which is reflected in low
measured cost efficiency. Specifically, so-called bad managers exhibit the
following tendencies. They are not adept at credit scoring and select a
relatively high proportion of investments with low or negative net present
values; collateral against loans is improperly valued; and customers are
not sufficiently monitored in order to ensure compliance with the loan
contract. It is important to note that “bad luck” hypothesis and “bad
management” hypothesis have an opposite temporal order, but both
predict that NPLs will be negatively associated with cost efficiency.2
Under the “skimping” hypothesis, it is implied that resources allocated
to underwriting and monitoring of loans affect both loan quality and
measured cost efficiency. Banks face a trade-off between short-term
operating costs and future loan quality. Management may choose to
minimize short-term operating costs by reducing expenditure on moni-
toring borrowers in an attempt to enhance long-term profitability.
Therefore, management delays having to deal with deterioration in asset
quality until an unspecified future date. Thus, under the skimping
hypothesis, we should expect a positive Granger-causation from mea-
sured efficiency to problem loans (i.e., an opposite sign in comparison
with the bad management hypothesis). Finally, the “moral hazard”
hypothesis implies that low financial capital Granger-cause high NPLs;
the idea behind this hypothesis is that banks’ managers have moral
hazard incentives to increase the riskiness of their loan portfolios when
their banks are thinly capitalized. The “moral hazard” hypothesis is then
based on the classical problem of excessive risk-taking when another
party is bearing part of the risk and cannot easily charge for or prevent
that risk-taking.
The results of the study of Berger and DeYoung suggest that the
intertemporal relationships between loan quality and cost efficiency run
in both directions. However, the data favor the bad management
hypothesis over the bad luck hypothesis and the skimping hypothesis.
Finally, decreases in bank capital ratios generally precede increases in
NPLs for banks with low capital ratios (moral hazard hypothesis).
Following the methodology of Berger and DeYoung, Williams (2004)
investigates management behavior in European saving banks from six
European countries (Denmark, France, Germany, Italy, Spain, and the
UK), between 1990 and 1998.3 The results of Williams are mixed:
Managers in German banks exhibit strong statistical evidence of bad
management, while there is weaker statistical evidence of bad manage-
ment in Danish and Italian banks.
Podpiera and Weill (2008) continue along this line of research and
examine the relationship between efficiency and bad loans in the Czech
banking industry from 1994 to 2005. They extend the Granger-causality
model developed by Berger and DeYoung by applying GMM dynamic
panel estimators. Their findings provide empirical evidence in favor of a
negative relationship between decreased cost efficiency and future NPLs
(i.e., the bad management hypothesis). Interestingly, Podpiera and Weil
use two different measures to assess credit quality: the conventional ratio
of NPLs on total loans and a so-called compensated risk taking measure,
which account for the fact that a certain amount of NPLs is normally
expected and accounted for in the interest required on such more risky
loans. Thus, the actual (uncompensated) risk-taking measure of a par-
ticular bank (associated with unexpected events) might be smaller if the
bank gets sufficiently compensated on interest revenues from the entire
loan portfolio. Therefore, a second measure is introduced and formulated
as the share of NPLs in total loans minus the share of interest revenues in
total loans. Indeed, if bank managers choose consciously to increase the
risk of the loans portfolio, we should expect immediately an increase in
interest revenue, and later an increase in NPLs. In this case, the risk of
NPLs could be considered well priced in the conditions of loans, and the
management behavior riskier but justified.
Other studies in the same stream include Karim et al. (2010) and
Louzis et al. (2012); the former investigates bank efficiency and NPLs in
Malaysia and Singapore and reaches conclusions similar to those of
Berger and DeYoung; the latter is a more complex analysis, in which
both macroeconomic and bank-specific determinants are taken into

account (see Sect. 4.2.3).
In most cases, these kinds of studies are run using data from single
countries and not considering macroeconomic factors in connection with
bank-specific determinants.
4.2.3 Micro and Macro Approach
As indicated above, explanations of the size and growth of NPLs can be

traced back to macroeconomic factors or to bank internal characteristics;
ça va sans dire that these two set of causes could be analyzed jointly.
There are indeed few studies which follow this approach; they include,
for example, Salas and Saurina (2002) which compare determinants of
problem loans of Spanish commercial and saving banks using both
macroeconomic and individual bank-level variables; Williams (2004)
which investigates management behavior at European saving banks
located in six different countries; and Klein (2013) which uses four
explanatory bank-level variables, three country (macroeconomic) specific
variables, and two global (macroeconomic) variables. Interestingly, the
results of this last study broadly confirm that both bank-level and
macroeconomic factors play a role in affecting banks’ asset quality,
although the contribution of bank-level factors is relatively small.
Louzis et al. (2012) analyze macroeconomic and bank-specific deter-
minants of NPLs in Greece, in a comparative study of mortgage, busi-
ness, and consumer loans portfolios, using a panel of data spanning from
2003 to 20094. In this study, nine different hypotheses are tested using
dynamic panel estimators, and two of them regard bad management.
Specifically, bad management hypothesis (I) refers to the link between
cost efficiency and future NPLs as in the preceding papers (even if
inefficiency is simply measured using the ratio between operating
expenses and operating income), while bad management hypothesis
(II) investigates the relationship between performance and future NPLs,
following the idea that past performance (ROE) could be interpreted as a
proxy for the quality of management, and should be negatively correlated

with a later deterioration of asset quality.
For all macroeconomic variables, the results of this study are com-
patible with the theoretical arguments, even if their impact is different
depending on the type of loans analyzed.5 Bad management hypothesis
(I) is confirmed by a positive and statistically significant coefficient of the
inefficiency index for all NPLs categories; the ROE indicator is statisti-
cally significant and negatively related to the mortgage and consumer
NPLs, supporting the bad management hypothesis (II) for these types of
loans.6 Moral hazard and diversification hypotheses are rejected.
Finally, Chiorazzo et al. (2016) analyze country-specific determinants
of NPLs jointly to banking-industry-specific determinants for 124 large
European banks located in 21 European countries: Their results high-
light the strong influence of country-specific variables on NPLs, while
the influence of bank-specific variables is rather limited.
The review of the literature reported above indicates, ultimately, that
both macroeconomic and microeconomic factors influence NPLs
increases in the course of time, with the first playing the most important
role. In this chapter, we will focus on bank-specific determinants, fol-
lowing the strand of literature starting with the study of Berger and
DeYoung (1997); the difference of our study relative to preceding
analyses lies in the fact that we can better discern the development of
NPLs during time using a unique dataset; in particular, our aim is to
understand whether bank management is able to recover impaired loans
before they become definitively bad loans or, put in a different way,
whether the bad management hypothesis is demonstrated by the inca-
pability of the bank management in doing that.
4.3 Data Description and Variables

In order to investigate the “bad management hypothesis” in the Italian
banking sector, we created a dataset composed of 48 banks including
data and variables in the period 2010–2013. Data about NPLs were
manually extracted from the unconsolidated balance sheet of single banks
and were then integrated with other data taken from Bankscope (Bureau
Van Dijk) and ABI Banking data (a specific database for Italian banks
created by the category Association). It is important to note that (in the
period investigated) NPLs were accounted in banks’ balance sheet fol-
lowing the accounting rules imposed by Bank of Italy (Rule 272/2008),
which considered four different categories of bad loans/impaired loans,
namely:
(i) substandard loans (loans to customers in temporary difficulties that

can be expected to be cleared up in a reasonable time)
(ii) past due/overdrawn more than 90 days
(iii) bad loans (loans to insolvent customers, even when insolvency is
not ascertained by court)
(iv) restructured exposures (loan for which a bank, upon granting a
moratorium on repayment, renegotiates the loan at lower than
market interest rates).
Information about bad loans and other impaired loans is manually

collected from the notes to the accounts of balance sheets. In particular,
we calculated the ratio between each different form of gross bad/impaired
loans and the total amount of credit to clients.
The classification reported above permits a better understanding of the
credit exposure of banks and give us the possibility to investigate the bad
management hypothesis (also) by looking at if and how much the
substandard/past due loans translate into bad loans over time. It is indeed
reasonable to think that good managers should be able to recover (at least
partially) these kinds of loans before they became definitively bad loans;
on the other hand, if a great amount of substandard/past due loans
became bad loans, we can assert that a poor management is positively
correlated with an increase in NPLs. It is important to stress that the
preceding literature (probably due to lack of data) normally considers the
total amount of NPLs, that is bad loans, substandard or “weak” loans,
and past due loans all together, making it impossible to discern the
internal dynamics and relations among these different categories of
impaired loans.
The selection of banks in our dataset started from the analysis of the
entire banking system that is all the banks surveyed in the ABI Banking
database (558 banks). We then considered independent and holding
banks for which an unconsolidated balance sheet was available, while
banks controlled by foreign companies were discarded. We then selected
the largest 48 banks in terms of total assets (2013) for two reasons:
(i) Our dataset represents numerically about 9% of the banks operating
in 2013, but about 82% of the total assets of the sector, 65% of loans to
domestic clients, and 66% of NPLs of the system; (ii) The detailed data
for the smaller bank unfortunately do not always exist. Table 4.1 reports
a first description of our dataset.
Table 4.1 shows that the majority of banks are located in the north of
Italy (the most industrialized zone of the country) and are commercial
banks operating as limited companies. Only seven banks operate as
cooperative banks (the classification presented in Table 4.1 follows the
classification used in ABI Banking database). To better investigate bad
loans, two other banks were finally excluded because they specifically
engaged in activities (such as private banking and asset management)
which do not produce significant amounts of NPLs.
As indicated above, we choose a specific period of time to investigate
NPLs in the Italian banking sector: Indeed, even if a deterioration of
economic conditions in Italy can be traced back to the outburst of the
financial crisis at the end of 2008, different studies (see Chiorazzo et al.
2016) indicate that a strong increase of NPLs in banks’ balance sheet is
temporarily linked to the government debt crisis (2011–2012). With
respect to the banks in our dataset, Table 4.2 shows how the mean ratio
of impaired loans (i.e., the four categories of bad loans described above)
on credit to clients changed during the investigated period.
The increase of NPLs actually went on also in 2014 and 2015,
reaching an astronomic value of over 300 billion euro for the entire
system. However, we decided to limit our analysis to the period 2010–
2013 because, from 2014, some banks in our dataset started to imple-
ment securitization processes which altered the accounting amounts of
NPLs and, from 2015, the introduction of different accounting rules
makes new data not comparable to the past ones.
66
Table 4.1 Dataset description

Geographical Dimension Commercial banks (ltd.) Cooperative banks Total
area Obs. Total assets 2013 Obs. Total assets 2013 Obs. Total assets 2013
(EUR billion) (EUR billion) (EUR billion)
Center 9 442.10 1 14.99 10 457.09
Large 1 36.34 – 1 36.34
Major 2 275.21 – 2 275.21
Medium-sized 5 87.56 – 5 87.56
Small 1 42.99 1 14.99 2 57.99
F. Crespi and M. Aliano
Northeast 11 558.79 3 90.90 14 649.68

Large – 1 42.68 1 42.68
Major 1 398.31 – 1 398.31
Medium-sized 4 94.52 1 42.11 5 136.64
Small 6 65.95 1 6.10 7 72.06
Northwest 20 895.75 2 75.82 22 971.57
Large 4 297.62 1 45.36 5 342.98
Major 1 393.16 – 1 393.16
Medium-sized 8 130.68 1 30.46 9 161.15
Small 7 74.29 – 7 74.29
South 3 45.87 1 9.34 4 55.20
Medium-sized 3 45.87 – 3 45.87
Small – 1 9.34 1 9.34
Total 43 1942.51 7 191.05 50 2133.55
All commercial, cooperative and 558 2610.71
mutual banks in ABI banking
database
% of our dataset 9% 82%
Source ABI banking Data and BvD data provider
Table 4.2 Mean ratio of impaired loans on credit to clients for the banks in our
dataset
Year Mean 2010 Mean 2011 Mean 2012 Mean 2013
(%) (%) (%) (%)
Commercial banks 8.37 9.86 11.68 14.88
(ltd.)
Center 11.34 13.98 17.62 22.17
Northeast 10.69 12.33 14.50 17.40
Northwest 5.31 6.20 6.88 9.44
South 12.12 13.69 16.47 20.87
Cooperative banks 9.20 10.73 14.67 18.96
Center 14.71 20.29 30.02 37.45
Northeast 8.02 8.82 12.28 15.64
Northwest 6.43 7.34 9.87 15.02
South 12.73 13.72 16.09 18.34
Total 8.49 9.99 12.11 15.46
4.4 Model and Variables

In the first draft of our model, we employ a simple panel regression in
order to examine the bank-specific determinants of the credit quality
(cq). As indicator of credit quality (or ex post risk), we utilize the ratio
between bad loans and accounts receivable (credit to clients). A higher
(lower) value denotes a deteriorating (better) quality of credit quality
ceteris paribus.
Badi;t
cqi;t ¼ ð4:1Þ
ARci;t
where cqi,t is a measure of the credit quality for the i-th bank in the year t,
while Badi,t is the amount of bad loans for the i-th bank in the year t, and
ARci,t is the amount of accounts receivable from clients for the i-th bank
in the year t.
It is important to note that our dependent variable does not consist of
the total amount of NPLs, but only of loans to insolvent customers (even
when insolvency is not ascertained by court) that can be considered no
more restorable. The aim of our analysis is indeed to investigate whether
other forms of impaired loans (still restorable) turn into bad loans after a
lag or, put in a different way, whether credit managers are able to restore
these kinds of impaired loans before they turn into bad loans.
As independent variables, we insert a set of indicators which describe
the economic and financial structure of banks in our sample, the weight
of loans in the balance sheet, the weight of other problematic loans (not
again bad loans), and the operational area. The objective is to catch
bank-specific determinants of bad loans. The specification of our model
is contained in the following formula:
X
n
cqt ¼ a þ bj Xj t1 ð4:2Þ
j
where bj is the coefficient associated with the independent variables j-th

(Xj) at the time t-1. Through a Pool Least Square Method each coeffi-
cient is estimated, without compute in fixed7 and random effects.
In Table 4.3, descriptive statistics of variables used are reported.
Our results show that the mean value of the credit quality (cq) variable
gets worse over the time, increasing from 4.5 to 8.5% in the period. That
is in line with the persistence of negative real economic conditions in
Italy, and this evidence justifies a high standard deviation value. For the
Arc variable, viceversa, a negative trend is recorded, that is there was a
decrease of loans to clients with respect to the total volume of business
(and this aspect may partly capture the so-called credit crunch
phenomenon).
The increasingly positive value for Exp and Ris variables suggests a
possible transformation degree of exposure delayed in bad loans, meaning
that over time critical exposures have become bad loans, both for the
continuation of the economic crisis and for the failure to select (screening
activities) and manage (controlling activities) credit.
The net interest income on total asset (Netinc) presents very low
values that, on average, become negative in 2013 due to losses on credit
and receivables. The ratio between total asset and equity (inversely
captured by the variable Equity) decreases progressively, denoting an
increase of capitalization requirement; however, this measure fails to take
Table 4.3 Descriptive statistics of variables used
cq Arc Exp Ris Netinc Equity Efficiency Struct
Mean 0.0626 0.0006 0.0088 0.0065 0.0009 7.5414 2.0403 0.5392
Median 0.0529 0.0007 0.0056 0.0050 0.0017 7.7205 1.8095 0.6720
4
Std. Dev. 0.0480 0.0003 0.0106 0.0060 0.0093 3.5147 1.7523 1.2925
Skewness 1.2851 −0.7712 2.7909 0.9364 −2.4219 0.5115 −1.3287 -10.7091
Kurtosis 5.4342 2.5418 13.5879 3.3610 17.3394 3.2971 24.5107 130.1650
Jarque-Bera 100.2458 20.7131 1146.0720 29.0992 1832.6470 9.0789 3758.1650 133,037.4000
The table shows the descriptive statistics of the variables used in the Panel Regression (OLS) for the period 2010–2013.
Credit quality (cq) is the ratio between bad loans and accounts receivable to clients, Arc is the ratio between accounts
receivable to client (ARC) and total asset, Exp is the ratio between exposure delayed and ARC, Ris represents the coefficient
associated with restructured exposure to ARC, Netinc is the coefficient of the ratio between net interest income and total
assets, Equity is the coefficient associated with the ratio between equity and total assets, Efficiency is the ratio between
operating income and operating expenses, Struct is given by the ratio total noninterest expenses/ (net interest
income + net fees and commissions)
Incapability or Bad Luck? Testing …
69
account of risk-weighted asset (RWA). For the Efficiency and Struct

variables, an erratic pattern seems to emerge.
As indicated, one of the variables used in our model is given by the
ratio between accounts receivable to clients and total asset. This variable
(named Arc) refers to the percentage of asset invested in loans to clients
and provides a proxy of credit specialization for the banks investigated.
A positive relation between credit quality and Arc (i.e., when the bj
associated is positive) indicates that when the relative (to total assets)
amount of receivables raises the quality of credit falls (or a positive
movement for cq is showed). Vice versa, if bj is negative, the rise in
percentage of credit improves the quality of credit (i.e., cq ratio falls).
Another variable used to define the determinants of bad loans is the
ratio between exposure delayed (Past due) and account receivable to
clients. Following the classification of impaired loans reported above, the
numerator of this ratio refers to a kind of problematic credit which is not
yet a bad loan, but the credit presents a significant delay in payment.
A positive coefficient indicates the attitude of problematic credit to
transform itself into bad loans, or the (in) ability of credit manager to
manage difficulty positions. A negative coefficient, on the other hand,
suggests a positive skill of credit management to recover problematic
receivables. This coefficient indicates the relation between the exposures
delayed at time t-1 and bad debts at time t; if this relation is positive, it
means that during the year the exposures delayed (less severe and
precedent condition) have been transformed in bad debt (most serious
and next condition). Since the macroeconomic and financial scenario is
the same for the banks in the sample, one possible explanation may be
provided by the (in) ability of management to manage exposures delayed.
Another type of impaired loans is restructured exposures, i.e., a
problematic credit for which the bank and client make a deal in order to
define a new payments program. This variable is also measured as a
percentage of accounts receivable to clients, and the interpretation of the
sign of the bj is the same as for exposure delayed.
An indicator of economic structure is specified by the ratio between
net interest income and total assets. A positive value of bj indicates that
the banks with high interest income (scaled on total assets) present a low
quality of credit, and vice versa. This indicator could be then interpreted
in two different and opposite ways. On the one hand, if bj is negative, the
relative rise in net interest income improves the quality of credit, and this
result could be interpreted as a sort of specialization efficiency; on the
other hand, if bj is positive, the net interest income and the quality of
credit are negatively correlated, and this result could suggest that con-
scious managers impose higher interest rate on more risky clients (as in
Podpiera and Weill 2008, that could be interpreted as a right behavior,
because the risk of bad loans could be considered well priced in the
conditions of loans).
The role of capital is studied by the ratio between equity and total
asset; this indicator highlights the effect of the capitalization on bad loans
ratio. If the coefficient is positive, more capital means more troubles in
loans. If the coefficient is negative, better capitalized banks show a low
bad loans to credit ratio.
The bank’s operative framework is captured by a dummy variable that
displayed 1 if the bank has a regional operating zone, 0 elsewhere.
Through this variable, we testify whether, for the Italian banks, the
environment impacts on credit quality.
Finally, we use two variables to investigate operating
efficiency/inefficiency. The operating efficiency is analyzed trough the
ratio between operating income and operating expenses; a negative
coefficient for this variable indicates the management’s attitude to gen-
erate profit from business activity and can be used as a proxy of good/bad
management.
The impact of the costs structure is captured by the following ratio:
total noninterest expenses/ (net interest income + net fees and com-
missions). This ratio quantifies the impact of the costs structure on gross
revenues and may be interpreted as an inefficiency indicator. A positive
value (what we expect) suggests a positive relation between operative
inefficiency and bad loans, an inefficiency that goes behind (deepening)
bad management.
In a further version of the model, we inserted a dummy variable
associated with the use of advanced IRB at the beginning of period
(2010); our aim was to investigate whether the use of more sophisticated
models to value the credit risk of a loan could reduce the following
appearing of bad loans.
4.5 Results
In this section, we show the estimations of the model presented above
using OLS panel regression without fixed and random effect. The results
are presented in Table 4.4.
The negative value of the coefficient Arc means that banks with a high
value of ARC on total assets present a better credit quality than the other
banks (i.e., the ratio of bad loans on credit to clients decrease). This result
suggests that a progressive specialization in credit can improve the quality
of credit.
The positive value of Exp (Exposure delayed) indicates a vicious circle
in which past due loans became nonperforming loans or, from a man-
agement perspective, the incapability of credit managers to recover
problematic credit (a different proof of the bad management hypothesis).
A similar interpretation is provided by the variable Ris; also in this case,
the positive coefficient shows that restructured exposures usually translate
in bad loans, demonstrating a failure in credit management.
The negative value for net interest income (Netinc) suggests that the
banks in our panel which are overspecialized in credit are better posi-
tioned in the management of the credit quality, and this result is in line
with that showed for Arc variable. Summarizing, banks with a greater
Table 4.4 Panel regression. Bank-specific determinants on credit quality

Model A Model B Model C
a 0.003986 0.00672c
Arc −12.18002b −11.67315b −10.74555b
Exp 0.752817a 0.766854a 0.760051a
Ris 0.419537c 0.499378b 0.507247b
Netinc −0.388602b −0.354188c −0.340694c
Bad loans (−1) 1.118938a 1.127141a 1.127228a
Equity 0.0000594
Regional 0.007892b 0.007343b 0.00696b
Efficiency −0.001806c −0.002132c
Size 0.0191c
Struct 0.000872
R-square 0.907171 0.909137 0.909137
a
significative at 99%
b
c
significative at 90%. R-square: 0.91. Total panel (balanced) observations: 138
ratio of credit to total assets and a greater ratio of net interest income to
total assets seem more capable to manage credit risk, imposing higher
interest rates on clients.
The variable equity on total asset is not significant, and similarly the
capitalization level seems not to impact on quality of credit. The dummy
variable, positive and significant, indicates that the regional banks are
more exposed to bad loans. With regard to efficiency, the negative value
indicates a negative relation, as expected, between efficiency and bad
loans; and this indication is partially confirmed by the value showed by
the variable Struct.
Similarly to Chiorazzo et al. (2016), we also find a significant auto-
correlation for the bad loans ratio, suggesting that an increase in bad
loans in one year creates more bad loans in the next year. Finally, the
control variable size (log total assets) has a negative and statistically sig-
nificant impact on credit quality.
To what concern the use of IRB models to asses credit risk, contrary to
other studies, we discovered that the coefficient related to this dummy is
not significant, and also trying with a detailed analysis that considers only
the credit quality and the use of Advanced IRB, give us back a negative
relation, that is Advanced IRB determines falls in credit quality.
Considering that our model could be affected by multicollinearity, we
use variance inflation factor8 (VIF) to check whether there is correlation
between independent variables employed in the models presented.
Logically, we expect a certain degree of correlation, especially among loan
quality variables, but this correlation should result in a lagged relation,
and not in a cross-sectional relation (i.e., intuitively the exposures delayed
at time t are not correlated with exposures restructured at time t, but
eventually with the exposures restructured at time t + 1), and then could
be considered a conversion factor of the progressive credit worsening.
As can be seen in Table 4.5, all the variables employed in the model
are less than 49 showing no problematic with VIF values.
Moreover, to better explain the relations between dependent variable
and independent variables, we run a simple redundant period fixed effects
tests. The F-Statistics, and the relative Prob. values, contained in
Table 4.6, lead us to reject the hypothesis that period fixed effects are
significant (at least at the 95% level).
Table 4.5 VIF test results

Arc 1.065 1.235 1.316
Exp 1.370 1.261 1.266
Ris 1.282 1.199 1.199
Netinc 1.114 1.131 1.136
Bad loans (−1) 1.565 1.435 1.449
Equity 1.282
Regional 0.132 1.136 1.149
Efficency 1.111 1.366
Size 1.064
Struct 1.266
The table shows the VIF’s values for the models displayed in the Table 4.4
Table 4.6 Redundant Fixed Effects Tests

Period F 2.35 1.38 2.42
Period Chi-square 5.07* 2.91 5.18*
The table shows the F-Statistics (F) and Chi-square values for the redundant fixed
effects tests. *** significative at 99%, ** significative at 95%, * significative at 90%
Table 4.7 Panel regression. Bank-specific determinants on credit quality. GMM

Arc −9.617761c −5.63208c −5.558272c
Exp 0.770208a 0.782203a 0.784275a
Ris 0.405222c 0.501241c 0.492964c
Netinc −0.336999b −0.263731 −0.259638
Bad loans (−1) 1.128687a 1.137686a 1.139673a
Equity 0.000238
Regional 0.007417a 0.007325a 0.007224b
Efficency −0.001189 −0.001423
Size 0.010226
Struct 0.000526
The table shows the estimations of the coefficients for the models given by a panel
generalized method of moments (GMM); we added constant to instrument list.
a
b
significative at 95%, c significative at 90%. R-square: 0.91
Further analyses are developed to check the persistence of the data; in

order to do that, we used dynamic panel data (system GMM) technique
proposed by Arellano and Bond (1991). Results displayed in Table 4.7
confirm the intensity and direction of the relationships identified in

Table 4.4, for the variables Exp and Ris.
4.6 Conclusion
The results of our analysis, as reported in Sect. 4.5, confirm the “bad
management hypothesis” first introduced by Berger and DeYoung
(1997). Indeed, we discovered that more specialized banks (higher ratio
of loans to clients on total assets and higher ratio of net interest income
on total assets) and more efficient banks (higher ratio of operating
income on operating expenses and lower ratio of total noninterest
expenses on net interest income + net fees and commissions) tend to
have a better quality of credit. Nevertheless, is rather surprising that the
use of IRB models is not significant in reducing NPLs during time.
More interesting, from our point of view, is the general relationship
between past due loans, restructured exposures, and bad loans. As our
results indicate, problematic loans (past due and restructured) tend to
transform into bad loans, demonstrating that credit managers were not
able to implement recovery strategies for these critical positions during
the investigated period. And even if part of these results could be
attributed to the stressed macroeconomic conditions of the time, it is licit
to affirm that bad management plays a part.
These last results are useful to better understand the present situation
of the Italian banking system: Indeed, the total amount of impaired loans
in the balance sheets of Italian banks is often much greater than bad loans
alone, also nowadays. Just to have an idea, gross bad loans at the end of
2015 for the whole banking system amounted to about 200 billions (see
Bank of Italy, Statistic Bulletin, I 2016), but the total amount of gross
NPLs (bad loans and other impaired loans) was 338 billions.
If the same process of transformation of impaired loans into bad loans
registered in the period investigated in our analysis should persist also in
the future, it is then easy to forecast that the level of bad loans in the
system will remain very high, preventing banks to increase significantly
the amount of credit notwithstanding the nonconventional impulses of
monetary policy that we have seen in the last years. Moreover, the
solution to the problem of NPLs in Italy, through securitization and/or
government guarantees, could take much more time than that requested
by ECB to banks in problematic situations (such as Monte Paschi di
Siena).
Finally, it is not inappropriate to suggest that better models to
investigate the dynamics of credit quality during time (probably, the
current IRB models used by banks are not able to forecast the deterio-
ration of credit quality in a period of stress economic condition), and
more efficient procedures to manage problematic credits (from the first
moments of their appearance in the balance sheet) should be
implemented.
Notes
1. Extra operating costs include, for example, additional monitoring of
borrowers, the expense of analyzing and negotiating possible workouts
arrangements, and the cost of disposing of collateral if default later occurs.
2. It must be stressed that these two assumptions are not mutually exclusive,
as the relationship may be bidirectional.
3. Actually, Williams (2004) estimates two different measures of bank effi-
ciency to test the hypotheses of Berger and Deyoung, namely operating
cost efficiency and profit efficiency. Moreover, problem loans are mea-
sured using the ratio of loan loss provision-to-loans instead of other
typical balance sheet measure such as the ratio NPLs/total loans or
NPLs/total assets.
4. The dataset is created using supervisory data for only nine largest Greek
banks, even if they accounted for 87.68% of the Greek banking system.
5. For example, the quantitative impact of GDP growth on mortgage NPLs
is attenuated compared to the NPLs of other loan types. Moreover, for all
NPLs categories, the estimation results indicate that the coefficients of the
macroeconomic variables are fairly stable across different models with
different bank-specific variables.
6. On the other hand, the ROE indicator for the business NPLs is
insignificant. The authors suggest that this may signify that the effect of
management quality is mainly reflected on the efficiency of households’
credit granting procedures, which are primarily based on the development

of quantitative modeling techniques, while the quality of case-by-case
assignment procedures, which characterize business loans granting, does
not differ substantially among banks.
7. The redundant fixed effects tests present a Prob. a value higher than 0.10
for the three following models.
8. VIF is expressed as followed:
1
VIFi;j ¼
1 R2i;j
where, R2i;j represents the R squared when the i-th explanatory

variables is regressed to the j-th explanatory variable.
9. For the VIF’s measure thresholds, see also O’Brien (2007): “We
demonstrate that the rules of thumb associated with VIF (and tolerance)
need to be interpreted in the context of other factors that influence the
stability of the estimates of the i-th regression.”
References
Arellano, M., and S.R. Bond. 1991. Some tests of specification for panel data:
Monte Carlo evidence and an application to employment equations. Review
of Economic Studies 58 (2): 277–297. doi:10.2307/2297968.
Bank of Italy (2016), Statistic Bulletin I (2016).
Barisitz S. 2011. Nonperforming loans in western Europe—A selective
comparison of countries and national definitions, Oesterreichische
Nationalbank. Focus on European Economic Integration Q1/13.
Beck R., P. Jakubic, and A. Piloiu. 2013. Non performing loans. What matters
in addition to the economic cycle, European Central Bank, Working paper
series, No 1515.
Berger, A., and R. Deyoung. 1997. Problem loans and cost efficiency in
commercial banks. Journal of Banking & Finance 21: 849–870.
Chiorazzo, V., V. D’Apice, F. Masala, and P. Morelli. 2016. Non-performing
loans in the wake of crises: What matters for large European banks?. ABI, Temi
di economia e finanza: Country vs. Bank Determinants.
Karim, M.Z.A., S. Chan, and S. Hassan. 2010. Bank efficiency and

non-performing loans: Evidence from Malaysia and Singapore. Prague
Economic Papers 2: 2010.
Klein, N. 2013. Non performing loans in CESEE: Determinants and impact on
macroeconomic performance, IMF Working Paper, WP/13/72.
Louzis, D.P., A.T. Vouldis, and V.L., Metaxas. 2012. Macroeconomic and
bank-specific determinants of non-performing loans in Greece: A compar-
ative study of mortgage, business and consumer loan portfolios. Journal of
Banking & Finance 36 (4) 1012–1027, ISSN 0378-4266, http://dx.doi.org/
10.1016/j.jbankfin.2011.10.012.
O’Brien R. 2007. A caution regarding rules of thumb for variance inflation
factors. Quality & Quantity: International Journal of Methodology, 41 (5):
673–690.
Podpiera, J., and L. Weill. 2008. Bad luck or bad management? Emerging
banking market experience. Journal of Financial Stability 4: 135–148.
Rinaldi, L., and A. Sanchis-Arellano. 2006. Household debt sustainability:
What explains household non-performing loans? An empirical analysis, ECB
Working Paper.
Salas, V., and J. Saurina. 2002. Credit risk in two institutional regimes: Spanish
commercial and savings banks. Journal of Financial Services Research 22: 203–224.
Williams, J. 2004. Determining management behavior in European banking.
Journal of Banking & Finance. 28 (10): 2427–2460.
5
Why Do US Banks React Differently
to Short Selling Bans?
Daniele Angelo Previati, Giuseppe Galloppo, Mauro Aliano
and Viktoriia Paimanova
5.1 Introduction and Motivations

of the Study
Financial crisis brought significant decreases in market indices, led to
active selling of stocks, and raised the possibility of a total collapse. Short
selling ban was expected to bring lower stock price volatility and raise
investor’s confidence. In this context, a policy intervention can change
the net expected present value of an individual bank, basically because
such kind of interventions aims to reduce the speculative selling pressure
on a single title stock, according to policy regulators. Consequently, it
should calm down the price reduction and net expected present value of
D.A. Previati (&)

Bocconi University, Milan, Italy
e-mail: daniele.previati@uniroma3.it
G. Galloppo V. Paimanova
University of Tuscia, Viterbo, Italy
M. Aliano
University of Cagliari, Cagliari, Italy
80 D.A. Previati et al.
every single stock. Second, the intervention may reduce both volatility
and probability of default of financial companies.
Previous studies showed a rather controversial role of selling bans on
stock price reaction, because it could even result in crashing of stock
market liquidity and increase of bid-ask spreads.
The main research hypothesis, investigated in our paper, relates to the
effect on stock prices in terms of financial returns and risks, caused by
short selling restrictions during the financial crisis in the US banking
sector. To be specific, we deal with two research questions. The first
research hypothesis, which corresponds to the title of the paper, aims to
verify, whether there are different reactions of banks in accordance with
their structural characteristics. In this sense, we operate in two ways: On
the one hand (Sect. 5.4.1), we perform an analysis of the temporal
dynamics of firms’ fundamentals; on the other, we check the action of the
short selling ban on bank value, so as proxied by its stock price (we
discuss it in Sect. 5.4.2). The second research hypothesis is to verify the
effect of short selling ban on the systematic risk component of the
intervention, and whether the structural characteristics of the bank cause
reduction of this component (please, see Sect. 5.4.3).
There are three main findings in this research. The first one shows
how short ban restrictions affected firm fundamentals in the USA. In
reality, it did not cause any big improvements in profitability of firms,
leverage, and liquidity. So, short selling ban did not bring the results that
might have been expected by official bodies, when implementing it in the
short-time period and did not make lots of changes in the long-time
period. Second finding relates to our observation about the stock price.
One year after the ban period for US banks, it happened to be higher,
then when we coupled with size and earnings per share control variable
measures. In terms of price change, it is also sensitive to asset turnover
and liquidity measures. Finally, when we consider results connected to
bank risk profiles, we conclude that there are changes in volatility in
terms of overall and systematic risk. We observe that banks are partic-
ularly effective, in terms of overall volatility reduction, earrings, and short
leverage metrics.
The main contribution of our paper to previous studies is twofold. For
our best knowledge, we are the first to discriminate the reaction by a large
5 Why Do US Banks React Differently to Short Selling Bans? 81
set of control variables describing the economic status of financial

companies, so that there may be differences in the response across banks.
The title of this paper goes in this direction. Second, we argue that this
kind of monetary policy interventions influences not only stock prices of
financial companies, but also their risk profiles, as well as systematic risk,
that is the most interesting component in periods of financial turbulence
(Bali et al. 2012; Campbell et al. 2005; Savor et al. 2016).
This paper consists of following parts. A literature survey is included in
the next Sect. 5.2. We present our data and empirical design in Sect. 5.3
and our main results in Sect. 5.4. In Sect. 5.5, we summarize our
findings and discuss their policy implications.
5.2 Literature Review

Short selling is a common tool to be used when periods of crises are
coming. The practice of borrowing shares and selling them at a lower
price should be studied in its influence on different economic and
financial processes, such as market liquidity, volatility and market price.
The consequences of short selling, especially under stressful market
conditions, require some tests on market downturns due to the ban. That
is why our research is related to a number of existing studies, which
confirm our statements and provide ideas on stabilizing the market.
Beber and Pagano (2013) found that bans were detrimental for liq-
uidity (even for financial stocks, where bans are associated with large
bid-ask spreads), in the most degree for stocks with small capitalization
and no listed options, slowed down price discovery mostly on bear
markets and failed to support prices. Moreover, they found that the
adverse liquidity effect of bans was stronger for stocks without listed
options. The effect of short selling bans on liquidity is measured by the
quoted percentage bid-ask spread and the Amihud illiquidity ratio; the
effect on the speed of price discovery is done by capturing the extent to
which individual stock returns correlate with past market returns; and the
effect on the overpricing stocks is measured by the excess returns on
stocks subject to bans. They use regression analyses to measure short sale
restrictions and base the bid-ask spread models on adverse selection and
inventory holding risk. Furthermore, the bans failed to support prices

with the exception that may be explained by the confounding effect of
the concomitant TARP announcement of the US financial stocks.
In contrary, Appel and Fohlin (2010) show that bans can improve
market liquidity and lower volatility. They found that bans did not harm
but even improved market quality, based on effective spreads and
volatility. They build a model based on idea, that the bid-ask spread is
lower, when regulators impose a ban on short selling and when only
those who already own a stock can sell it. Their results, received by the
use of panel regressions, suggested that short sellers were more likely to
trade on information, the limits on short selling reduced information
transmission and price discovery and ban reduced liquidity. Their
observation of the impact of the short selling ban on stock volatility
demonstrated that imposing of short selling bans reduced volatility in
equity markets. Moreover, they found that bans improved liquidity and
reduced volatility. The lifting of the ban led to increase the volatility of
financial stocks in relation to non-financial stocks.
The effect of short selling on volatility requires deep attention, espe-
cially in times of financial crises. Mattarocci and Sampagnaro (2010)
examine whether the short selling bans from financial companies of
Italian Stock Exchange affected the daily volatility and returns of the
shares. They suggest a different impact of the short selling prohibition,
which depends on focus on risk or performance of bank shares to the
ban. They use a dual methodological approach, which is the analysis of
standard deviations pre- and post-restriction, and analysis of asymmetric
conditional volatility by ANOVA and GARCH. They proved that
volatility increases in most of the stocks, covered by the ban, and there is
a direct relationship between the increase in volatility and prohibitions,
imposed by the authority despite the original market function of the ban.
Moreover, they suggest that elimination of short selling can actually
trigger market instability. Short selling can be used for correction of
market trends and prevention of market bubbles.
Charoenrook and Daouk (2005) checked if short selling restrictions
have no effect on skewness and if it declines with introduction of options
trading. They found that the absence of short selling reduces liquidity
and found no evidence that short selling disrupts orderly markets by
causing panic selling, high volatility or market crashes. They proved that
short sales increased market quality. By the use of panel regression and
event study, authors conclude that short sales enhance market quality
and affect market returns.
Initially, short selling was implemented in order to reduce the negative
pressure on financial market stocks and supposed to end in a lower
volatility. In fact, according to another research paper (Schwartz and
Norris 2009), the market behaved in an opposite way and brought a
short-term increase in volatility for all observed firms. Moreover, as it
turned out, small firms were affected in the days just after the ban, while
large ones felt its influence before and during the ban implementation.
It is worth of saying that short selling has a direct impact on risk,
which requires further careful studies. Felix et al. (2014) examined how
the European short sale ban affected jump and contagion risk of banned
and non-banned stocks. It resulted in the evidence that short positions in
banned stocks decreased while they increased for non-banned stocks. On
using extreme value theory and implied volatility skews, they found that
jump risk abruptly rose for all stocks and contagion risk decreased for
banned stocks and increased for non-banned stocks after imposition of
the ban. They showed that the increase in jump risk was caused by the
imposition of the ban itself, rather than by information flow, options
trading volumes and stock-specific factors. They show that the increased
risk levels are especially evident for the banned financial stocks. An
increased jump risk may provoke financial contagion and increase sys-
tematic risk. They found that the contagion risk dropped for banned
stocks after the ban. So, their results proved that short sale bans reduced
contagion risk in the financial sector.
Some following studies suggest that short sale bans do not impact
stock prices and can even contribute to their decline and negatively
impact on market quality. For example, Helmes et al. (2010) proved that
imposing constraints on short selling reduce trading activity, increase
bid-ask spreads and price volatility, which was done with the use of
univariate and multivariate fixed effects panel regressions.
Boehmer et al. (2013) studied emergency order (SEC decision in
2008) that temporarily banned about 1000 financial stocks. They found
that the ban affected larger stocks in a greater degree than the small
quartile stocks of firms. It ended in a low market quality, as it was

measured by quoted spreads, effective spreads, and volatility. Authors
checked the hypotheses by the following fact. If shorting is banned,
bid-ask spreads would narrow, because liquidity providers face less
adverse selection. On the other side, they questioned how a shorting ban
can influence market quality, if short sellers are important liquidity
providers. They describe the effects of shorting ban graphically and in
firm-pair fixed effect panel regressions. They prove that shorting ban
eliminates a substantial subset of trading activity and it might worsen
market liquidity.
Battalio et al. (2012) focus on investors, who short stocks for longer
periods, because they are overpriced according to them. They are
studying the effect of short selling bans of 2008. They examine the link
between market downturns and short selling; evaluate evidence on the
bans’ effectiveness in limiting share price declines of 2008 and the costs
imposed by these bans. To measure the impact of short selling on stock
prices, regression models are used. Authors conclude by saying that bans
seem to have unwanted effects of raising trading costs, lowering market
liquidity and preventing short sellers from cases of fraud and earnings
manipulation, however, they do not drive price declines on markets.
If the ratio of informed to uninformed short sellers rises, prices react
more negatively to unexpected short interest news (Kolasinski et al. 2013).
They examine the market response (around 1000 US financial stocks) to
short interest announcements and find that rule changes increase the ratio
of informed to uninformed short sellers. They find that the negative rela-
tion between returns and changes in short volume becomes stronger during
the ban. The price reaction to one standard deviation increases in the ratio
of short to total volume and becomes more negative by economically and
statistically significant points. They use the Diamond and Verrecchia
(1987) empirical model, which predicts an increase of information content
of short sales and causes prices to better reflect private information, which
they call the “short restriction” effect.
Short selling ban can lead to inflation in banned stocks when buyers
have to pay more during the period of the ban. The previous analyses of
effects of the ban on short selling of US financial stocks imposed by the
SEC in September 2008 (Harris et al. 2013) showed that price inflation
was found to be lower for stocks with greater short interest before the
ban. Moreover, they found that price inflation is the strongest for stocks,
where no listed options trade. The estimation is done with the use of the
factor-analytic model, where they compare the actual and predicted
banned stock index returns in two separate timeframes outside of the ban
period.
The effect of short selling ban can be different for financial and
non-financial firms (Hasan et al. 2010). They find that during the crises
period, the short selling of financial firms stock was not much greater
than of non-financial firms when short sellers rationally short sold the
financial company stocks with the greatest subprime and insolvency risk
exposures. Moreover, they investigated whether short selling activity
rationally reflected financial companies’ insolvency risk exposure with the
use of a direct measure of the exposure to the subprime market.
Grullon et al. (2015) found that the increase of short selling brings
price fall, affects asset prices and impacts on financing and investment
decisions. He proved that even “the uptick rule can have a significant
effect on the equity prices of financially constrained firms and appear to
be a binding constraint on the equilibrium level of short selling.”
Moreover, he discovered that when removed, short selling influences
short selling activity and stock prices “even before the actual suspension
of the constraints.”
5.3 Data and Methodology

In this section, we present the research design to measure the impact of
short selling bans on financial companies. We have implemented several
methods investigating the reaction of banks to short selling bans in terms
of stock price reaction and checking what has happened from the risk
side. In Table 5.2, we show the relevant event date of short selling bans.
First, we conducted a descriptive analysis by observing what was going
on at different firm fundamentals for certain periods back and forth with
respect to the dates of the interventions of short selling bans. We then
wondered what effect can be observed on the stock price, as a proxy for
the entire banking firm value, when the action of the short selling ban is
implemented. We perform pooled regressions to understand what banks

reacted better and worse with respect to their fundamental firms. To this
end, we run two models (Mod1 and Mod2) as follows:
X X X
Mod1 : Pi;t ¼ a þ bj Xj;i;t þ kj Wj;i;t þ cExpk;i;t
j j k
X X ð5:1Þ
þ dLevk;i;t þ nSizek;i;t þ ei;t
k k
where the dependent variable in the model is the stock price considered
at the end of selling ban period, for the i-th stocks. We considered also a
dynamic version of the Eq. 5.1, as follows:
X X X
Mod2 : DPi;t ¼ a þ bj DXj;i;t þ kj DWj;i;t þ cDExpk;i;t
j j k
X X
þ dDLevk;i;t þ nDSizek;i;t þ ei;t
k k
ð5:2Þ
where DPi;t represents the change of stock price between the start of short
selling restrictions and the same observations one year ahead for the i-th
banks. Same interpretation counts for all the other independent variables.
Xj;i;t is a vector of control variables accounting for the level of Profitability
(Asset_Turnover, EPS_Annualized, Ebit, EPS_Growth, Grossmargin,
Return_On_Asset, Return_On_Cap); Wj;i;t is a vector of variables repre-
senting the level of firm’s Liquidity (CF_CASH_FROM_OPER,
CF_NET_INC); dLevk;i;t includes control variables accounting for
Leverage (CURRATIO, LTBTBT); Expk;i;t is a vector of control variables,
stands for Expectations (BEST_ANALYST, BEST_TARGET_PRICE);
and lastly nSizek;i;t is a vector accounting for Size control variables
(CURR_MAR_CAP, TOT_COM_EQY) of each i-th stock taking into
account (see Table 5.1 for the description of variables).
Table 5.1 Fundamental variables description

Variable Description
ASSET_TURNOVER Asset turnover ratio is the ratio of the value of a
company’s sales or revenues generated relative to
the value of its assets
CUR MARKET CAP Current market capitalization and means a total value
of all outstanding shares of a company
CUR_RATIO Current ratio and indicates whether a company is able
to pay its obligations
EBIT Earnings before interest expenses and income taxes
EPS_ANNUALIZED Earnings per share in last fiscal year
EPS_GROWTH The raise or down of earnings before extraordinary
items, when making a comparison between the
observed current period and the same time last year
LT_DEBT_TO_TOT_ASSET All interest-bearing financial obligations that are not
due within a year
RETURN_ON_ASSET Return_on_asset stands for ROA and shows how
profitable a company is in relation to its total asset
TOT_COMMON_EQY Represents an equity measure, which counts only the
common stockholders without the preferred ones
CF_NET_INC Stands for cash flow on net income and explains the
ratio between a stock’s price and its cash flow per
share
RETURN_ON_CAP Return_On_Cap is a metric that measures the return
that an investment generates for capital
contributors. It is a proxy of ROE, a measure of a
corporation’s profitability by revealing how much
profit a company generates with the money
shareholders have invested
GROSS_MARGIN It represents the percent of total sales revenue that
the company retains after incurring the direct costs
associated with producing the goods and services
sold by a company
CF_CASH_FROM_OPER Cf_Cash_From_Oper represents Cash Flow from
Operating activities
BEST_ANALYZED Best_Analyst represents the consensus rating based on
analyst recommendations
BEST_TARGET_PRICE Best_Target_Price consists in analyst price short-term
forecast
This table reports fundamental variables description. Source Authors’ elaboration
When it comes to control variables, we mention some details about

the use of different analysis levels and their role in explaining effects on
stock price changes.
Profitability: These variables are used to estimate the impact of
profitability measure on stock price changes. We expect a positive rela-
tionship between profitability and stock price changes. Asset turnover
ratio is the ratio of the value of a company’s sales or revenues generated
relative to the value of its assets. Ferrer and Tang (2016) find a positive
relation between changes in stock prices and asset turnover ratio; their
results show that asset turnover ratio, coupled with other financial ratios,
would have a positive impact on the year-on-year change in stock price.
Also, Earnings before interest expenses and income taxes, Earning per
Share, Gross Margin, Return on Asset (ratio of net income to total assets)
and Return on Capital (ratio of net income to capital) might be included
in profitability indicators. Under this perspective, many studies have
analyzed the effect of banking business diversification on profitability and
the relation on stock prices, but their evidences are often contradictory
(Maudos 2017).
Leverage: The capital structure is arguably one of the most important
decisions a company face. On using these variables, we estimate the effect
of leverage changes (debt/capital structures) on stock prices. Current
Ratio is given by the ratio of current assets and current liabilities and
indicates whether a company is able to pay its obligations. This indicator
represents both Leverage and Liquidity indicators, following Ferrer and
Tang (2016). The impact on stock price (Shieh et al. 2012), regarding
the US market, depends on market characteristics and momentum effect.
LT_DEBT_TO_TOT_ASSET considers all interest-bearing financial
obligations, that are not due within a year on total asset, and may rep-
resent another leverage measure. Cai and Zhang (2011) find a negative
effect of the change in a firm’s leverage ratio on its stock prices for
non-financial firms. However, the evidences provided (on a dataset that
consists of 20 US banks) by Papanikolaou and Wolff (2014) indicate that
leverage contributes to both total bank risk and systemic risk and
increases risks and volatility.
Size: Through these variables, we analyze the impact of banks’ size on
stock price changes. Current market capitalization, current enterprises
value, and common equity represent the size of companies. Using the
vector autoregressive (VAR) process on US monthly stock market data
(from 1926 to 2006), Ho et al. (2011) find a highly and negative rela-
tionship between small and large cap stock prices with a cyclical pattern.
Liquidity: Cash flow from operating activities and cash flow from net
income represent a liquidity measure in stricto sensu. Khan et al. (2016)
investigate the relation between funding liquidity and banks’ risk for US
bank holding from 1986 to 2014 and find that banks facing lower
funding liquidity risk take more risk or more instability of stock price
changes.
Expectation: The nature of expectation variables is seen through the
following. Based on a sample of 23,632 analyst recommendations of
1106 banks, Premti et al. (2016) find that analyst recommendations are
more informative for riskier banks and subject to a higher degree of
information asymmetry. The impact on stock prices is unpredictable and
depends on analysis period.
To analyze the impact of short selling restrictions from risk side, we
investigated, via pooled regressions, what was going on with overall and
systematic risks in the presence of short selling restrictions. Focus on such
fundamental firms is more effective in limiting the risk for banks.
To be specific, we select all stocks belonging to local market banking
index and we run two models as follows:
X X X
Mod3 : DOverallRIski;t ¼ a þ bj DXj;i;t þ kj DWj;i;t þ cDExpk;i;t
j j k
X X
k k
ð5:3Þ
X X X
Mod4 : DSysti;t ¼ a þ bj DXj;i;t þ kj DWj;i;t þ cDExpk;i;t
j j k
X X
k k
ð5:4Þ
where OverallRIski;t represents standard deviations computed in an

estimation window of 252 days, Systi;t stands for systematic risk, also
known as “undiversifiable risk,” affecting the overall market, not just a
particular stock or industry.
Following Damodaran’s approach (1999), we use the following for-
mula to determine the systematic risk for i-th securities:
systi ¼ bi rmkt ð5:5Þ
where systi is the systematic risk for i-th security, rmkt the standard
deviation of market return, bi is the Beta regression (OLS method)
calculated as:
ri ¼ ai þ bi rmkt ð5:6Þ
where ri is the return of i-th security and rmkt the return of market. The
unsystematic risk is obtained as:
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
unsysti ¼ r2i systi ð5:7Þ
where r2i represented the variance of the return of i-th security.

In order to avoid outliers, namely extremely deviant cases with Cook’s
D greater than 1, we perform robust regression using iteratively
reweighted least squares.
Data are collected from various sources. For the period from January
2009 to December 2013, we draw information from specific national
FSA agency regarding the period of short selling restrictions (see
Table 5.2 for viewing specific event dates). Single stock and general
index prices are downloaded by Datastream. Fundamental firms are
extracted by Bloomberg.
The sample from our analysis is composed of 82 banking companies of
USA—Standard & Poor’s 500 Index. It covers the period from January
2005 to December 2014. Our dataset consists of daily observations, total
returns, and fundamental balance sheet data. All-time series are obtained
Table 5.2 Short selling ban interventions in US financial sector
Country Authority From Till Type of restriction Exceptions Type of targeted
firms
(FSC)
USA Securities 07/21/2008 08/20/2008 Naked short selling Market Financial institutions
and ban makers
Exchange (mandatory pre-
Commission borrowing)
(SEC)
This table reports Short Selling Ban Interventions in US Financial Sector. Source Authors’ elaboration starting from single
country FSA web site
5 Why Do US Banks React Differently to Short Selling Bans?
91
92
Table 5.3 Descriptive statistics

D.A. Previati et al.
n° Best Best cash Cash Curr Current Current Debt Common EPS EPS ROA ROC Daily
stocks analyst target flow flow enterprise market ratio % to equity annual growth % % return
rating net value total %
income asset
Mean 82 3.68 60.36 768.06 381.66 57,366.10 24,711.63 1.60 16.46 19,374.12 1.37 48.93 2.99 6.84 0.06%
Standard 82 0.43 90.10 6615.95 1811.55 161,409.85 39,128.80 1.14 15.52 36,271.12 40.48 473.98 5.69 13.60 0.07%
deviation
Skewness 82 −0.15 10.12 11 4 39 38 3.54 33.33 38.18 3.99 2.19 7.54 16.20 0.10
Kurtosis 82 2.63 131.75 82.43 568.13 27.40 15.88 20.65 4.89 15.26 1605.35 143.75 29.34 123.64 3.02
This table provides the panel data descriptive statistics for the sample used in our elaboration for the period of January 2006–December 2013. The analyses and the values
contained in the table are developed starting from daily observations. The variable Best Analyst Rating is calculated as a weighted average of opinions of various analysts
(5 = buy; 4 = high performance; 3 = hold; 2 = low performance; 1 = sell) and indicates the analyst recommendation and their consensus on single stock. The variable Best
Target indicates the projected analyst price level. The variables Cash Flow, Cash Flow Net Income, Current Enterprise Value, Current Market and Common Equity represent the
fundamental balance sheet information. The variable Debt to Total Asset indicates the ratio between Total Asset and Debt. EPS Annual % and EPS Growth represent the
earning per share and the rate of growth of earning per share, respectively. ROA % and ROC% indicate, respectively, the Return On Asset and the Return On Capital. Daily
Return is the daily stock return. Zeta Altman represents the Altman Z-score for predicting bankruptcy
from Bloomberg. Table 5.3 provides an overview of the descriptive

statistics associated with the banned financial stocks.
Table 5.3 contains the mean, standard deviation, skewness and kur-
tosis values of the fundamental variables used in the analysis organized
by: expectation (Best Analyst Rating and Best Target Price), corporate
liquidity (Cash Flow and Cash Flow Net Income), enterprises values
(Current Enterprise Value and Current Market Capitalisation), leverage
(Current Ratio and Debt to Total Asset) and performance (Common
Equity, EPS Annual %, EPS Growth, ROA %, ROC % and Daily
Return) and risk (Zeta Altman). Daily closing stock prices for individual
financial institutions and for market indices are also retrieved from
Bloomberg.
5.4 Empirical Results

We present empirical results according to the logic showed above. In
Sect. 5.4.1, we present banks’ fundamental analysis, then in Sect. 5.4.2
we report empirical results in terms of stock price reaction. Then we
conclude in Sect. 5.4.3, considering risk analysis in terms of both overall
and systematic measures.
5.4.1 Banks’ Fundamental Analysis
We performed our test by considering fundamental variables related to each

specific bank, listed in the US Sp500 Financial Index, and in terms of 1 year
before the event and 1 and 2 years after the event. Specifically, we aimed to
reach our research objective, that is, to explore the stock price performance in
these countries in their response to stock short selling ban announcements, as
it is a remedy to be used by central banks in order to provide a stable situation
on the market and control over the price decline. The analysis contains
14 variables, where every variable corresponds to a certain year of analysis
and is grouped according to its characteristics, e.g., Profitability
(ASSET_TURNOVER, EBIT, EPS_ANNUALIZED, EPS_GROWTH,
RETURN_ON_ASSET, RETURN_ON_CAP, GROSS_MARGIN),
Leverage (CURR_RATIO, LT_DEBT_TO_TOT_ASSET), Liquidity (CF_

CASH_FROM_OPER, CF_NET_INC), and Size (CURR_MAR_CAP,
CURR_ENT_VAL, TOT_COM_EQY). We were measuring stock price
reaction to two dates of announcements of short selling ban (initial and final
date).
On following the presentation of fundamental variables in Sect. 5.2,
we provide our main results and policy implications from descriptive
statistics analysis contained in Table 5.4.
Profitability: We started our observation from how did short selling
restrictions affect profitability indices. ASSET_TURNOVER demon-
strated decline in the US profitability indices in their reaction to short
selling ban announcements for both observed event dates, which was
followed by their reduction in the first year after the ban in the short
term and got growing meaning only for EPS_ANNUALIZED, ROA and
ROC 2 years after the ban, that means that the US market needed some
time in order to respond to short selling ban, however, in the year of
announcement in the short term, it did not bring any improvement in
the US profitability fundamentals.
Leverage: Furthermore, we decided to explore the reaction of leverage
fundamentals to short selling ban. The US leverage fundamentals showed
that CURR_RATIO demonstrates price decline to short selling in the
short term, in contrast, there are very little changes in the long term. As
for LT_DEBT_TO_TOT_ASSET, it was raised in the short period,
however, without any big changes in the long term.
Size: Size fundamentals were influenced by short selling restrictions in
the following way. We should notice reduction of both
CURR_MAR_CAP and CURR_ENT_VAL size variables for the US
banks in the short term. At the same time, it is worth of saying that there
are no improvements of size fundamentals for the US banks in the long
term. We can conclude that short selling did not cause many improve-
ments in firm’s size.
Liquidity: Then, we observed liquidity fundamentals of our sample
countries, in their reaction to short selling ban. We found that short
selling caused a rise of the US firms’ liquidity by its announcement in a
short-term perspective for CF_CASH_FROM_OPER. However, it
ended in a general reduction of indices in the long time. CF_NET_INC
Table 5.4 Fundamental firms’ analysis

Years t = −1 t=0 t = +1 t = +2
Profitability
ASSET_TURNOVER 0.23 0.21 0.19 0.19
0.23 0.21 0.19 0.19
EBIT 938.25 556.77 386.83 372.86
919.59 515.64 386.83 372.86
EPS_ANNUALIZED 5.74 0.19 1.55 2.02
5.76 0.07 1.55 2.02
EPS_GROWTH 50.12 −3.16 −23.46 60.81
50.71 −3.29 −23.46 60.81
RETURN_ON_ASSET 4.44 3.57 0.27 2.18
4.44 3.57 0.27 2.18
RETURN_ON_CAP 10.80 8.10 0.61 5.59
10.80 8.09 0.61 5.59
GROSS_MARGIN 3.20 3.04 2.97 3.41
3.20 3.04 2.97 3.41
Leverage
CURR_RATIO 0.22 0.16 0.18 0.18
0.22 0.16 0.18 0.18
LT_DEBT_TO_TOT_ASSET 14.53 16.01 17.30 16.97
14.54 16.01 17.30 16.97
Liquidity
CF_CASH_FROM_OPER 49.92 1971.30 1027.80 1496.51
127.58 1979.57 1027.80 1496.51
CF_NET_INC 651.18 299.36 313.96 379.25
644.86 289.08 313.96 379.25
Size
CURR_MAR_CAP 28980.29 22269.71 17460.62 22158.81
29440.91 22250.52 19768.79 20582.19
CURR_ENT_VAL 54697.59 44384.54 34065.30 39443.27
54712.16 43427.62 35349.04 38765.32
TOT_COM_EQY 15474.20 15629.03 15695.61 19413.22
15430.28 15654.07 15695.61 19413.22
This table provides the fundamental firms’ descriptive statistics for all the banks
listed in US general index according to time (years), ahead and back,
corresponding to short selling restrictions period. The first line of each variable
corresponding to the observation detected at the day (t = 0) of short selling ban
start, or 1 year before (t = −1) or 1 and 2 years (t = +1, t = +2) ahead, respectively.
The second line of each variable reports variable values corresponding to the
observation detected at the day (t = 0) of short selling ban end, or 1 year before
(t = −1) or 1 and 2 years (t = +1, t = +2). The variables are grouped according to
firm fundamental characteristics and legend of each variable is provided by
Table 5.1
reduced indices of US banks (first announcement), but caused raises of

US firm fundamentals in the long time.
Our analyses of how short ban restrictions affected firm fundamentals
in US banking system showed that it did not cause any big improve-
ments in profitability of firms, leverage and liquidity, as well as no much
positive effect on firm size. So, short selling ban did not bring the results,
that might have been expected by official bodies, when implementing it
in the short-time period and did not cause lots of changes in the
long-time period. For sure, the effectiveness of short selling is a key point
for policy makers on their decisions to see if it can bring serious
improvements in firm’s characteristics. However, it is very important for
policy makers to pay attention to a local level, as our results are not the
same for the observed countries. Moreover, we would like to underline
that some variables show a greater sensitivity to firm fundamentals than
the others. To tell more, it turned out that short selling brought almost
no improvements on firm’s fundamentals at the time of its closing date
and only slight changes can be seen in 2-year time.
5.4.2 Stock Price Reaction
Next, we decided to study the reaction of bank stocks, therefore, their

price response to final date of short selling and the change of the price
(between the final and the initial date).
The results about the price changes (see Table 5.5) at the end of the
ban period (we refer in the text as Mod1) and 1 year after the end of the
ban interventions (we refer in the text as Mod2) are more interesting.
This part of the analysis is a novel in the literature.
The situation of price change 1 year later the short selling ban closing
(Mod2) involves the following considerations. The characteristics of our
country samples, by a number of banks, are as follows: We consider 82
banks listed in US Sp500 Financial Index, while the average value of the
stock price change a year after the short selling ban closure is −42%. The
first reading exhibits that American banks have shown a price fall.
Profitability: When considering ROA as a profitability indicator, its
contribution (which is normally expected to be positive) seems not to be
Table 5.5 Stock price change
USA Size Profitability Leverage Liquidity Expectations Full model
Coeff. St. Coeff. St. Dev Coeff. St. Dev Coeff. St. Coeff. St. Coeff. St. Dev
Dev Dev Dev
Price final date
f_cur_mkcapt −0.0004 0.0004 0.0004*** 0.0002
f_tot_com −0.0005** 0.0002 0.0015*** 0.0004
f_assetturnover −38.8896 31.9368 −38.8896 31.9368
f_ebit −0.0006 0.0012 −0.0012 0.0018
f_EPS_an 0.5083*** 0.1823 2.4566*** 0.1233
f_EPS_growth 0.0130 0.0148 0.0094*** 0.0084
f_grossmargfn 0.1996 0.2230 −1.8838 2.6300
f_roa 0.4266 0.7047 −0.4468 0.6564
f_roc −0.0458 0.2060 0.2180 0.3410
f_curratio -3.1233 5.3238 −7.1038 12.8120
f_lt_db_t 0.0231 0.1099 0.0737 0.3591
f_cf_net 0.0004 0.0005 0.0018*** 0.0007
f_cf_in 0.0044*** 0.0030 −0.0205*** 0.0043
f_best_an 0.7887*** 0.0451 −5.8691*** 1.6420
f_best_targ_price −7.5403*** 1.6376 0.7887*** 0.0451
Delta price
d_cur_mkcapt 0.0028*** 0.0004 0.0004*** 0.0002
d_tot_com 0.0001 0.0001 0.0001 0.0001
d_assetturnover −27.6598 44.3376 −27.6598 44.3376
d_ebit −0.0013** 0.0006 −0.0013** 0.0006
d_EPS_an 0.5968*** 0.1695 0.5968*** 0.1695
d_EPS_growth 0.0043*** 0.0030 0.0043*** 0.0030
d_grossmargin −0.8629 0.8499 −0.8629 0.8499
d_roa 0.2803*** 0.2281 0.2803*** 0.2281
d_roc −0.0734 0.1142 −0.0734 0.1142
d_curratio 17.5205 30.3783 −2.2000 4.0666
(continued)
97
98
USA Size Profitability Leverage Liquidity Expectations Full model

Coeff. St. Coeff. St. Dev Coeff. St. Dev Coeff. St. Coeff. St. Coeff. St. Dev
Dev Dev Dev
d_lt_db_t −1.7512 1.3282 −0.3937* 0.1977
d_cf_net_in 0.0001 0.0006 0.0001 0.0002
d_cf_cash_in −0.0260*** 0.0033 −0.0038*** 0.0010
d_best_an −2.3662 1.8207 −3.9638*** 1.5300
d_best_targ_price 0.6246919*** 0.0137 0.6719*** 0.0397
In this table, we report the effect it can be observed on the stock price, as a proxy for the entire banking firm value, when the action of the short selling ban is implemented. To
this end, we perform pooled regressions to understand what banks reacted better and worse with respect to their fundamental firms
P P P P P
Mod1 : Pi;t ¼ a þ bj Xj;i;t þ kj Wj;i;t þ cExpk;i;t þ dLevk;i;t þ nSizek;i;t þ ei;t where the dependent variable in the model is the stock price considered at the end of selling
j j k k k
ban period, for the i-th stocks, Xj;i;t is a vector of control variables accounting for the level of Profitability (ASSET_TURNOVER, EPS_ANNUALIZED, EBIT, EPS_GROWTH,
GROSSMARGIN, RETURN_ON_ASSET, RETURN_ON_CAP), Wj;i;t is a vector of variables representing the level of firm Liquidity (CF_CASH_FROM_OPER, CF_NET_INC), dLevk;i;t
includes control variables accounting for Leverage (CURRATIO, LTBTBT), Expk;i;t is a vector of control variables, stands for Expectations (BEST_ANALYST, BEST_TARGET_PRICE)
and lastly nSizek;i;t is a vector accounting for Size control variables (CURR_MAR_CAP, TOT_COM_EQY) of each i-th stock taking into account. We considered also a dynamic
version of the Eq. 5.1, as follows
P P P P P
Mod2 : DPi;t ¼ a þ bj DXj;i;t þ kj DWj;i;t þ cDExpk;i;t þ dDLevk;i;t þ nDSizek;i;t þ ei;t where DPi;t represents the change of stock price between the start of short selling
j j k k k
restrictions and the same observations 1 year ahead for the i-th banks. Same interpretation counts for all the other independent variables. ***, **, * denote that estimates are
statistically significant at the 1, 5 and 10% levels
estimated by market participants, especially for Mod1. While a positive

and significant contribution is provided by EPS indicators for two
models, classical profitability variables and those related to the levels of
earnings are not effective in contrasting the price fall.
Leverage: Results highlight a positive contribution to the containment
of the price fall for the degree of leverage in Mod2 through long term and
for total asset ratio.
Size: Evidence shows the importance of current market capitalization
as a size variable and its positive contribution to stock price in both
Mod1 and Mod2, while common equity does not provide a significant
contribution to the explanation of stock price changes in two models.
Liquidity: The contributions of liquidity variables to stock (Mod1)
and stock price changes (Mod2) are ambiguous and depend on the type
of indicator used. Using cash flow from net income, the impact is pos-
itive on both models.
Expectations: The contribution of analysts’ expectations is positive, in
the sense of contribution to the price reduction due to the variation of
recommendation, i.e., increasing the level of recommendation on indi-
vidual securities is an increase of the reduction of the price fall. Even a
long-term debt improved results in a better price change.
In the model used to measure the sensitivity of the price change 1 year
from the closing of the ban, some information useful for policy makers
comes out. In particular, there are considerations about the results
coupled with the size variable and mostly to those related to cash levels
and those related to the expectations of the analysts in the sense of
recommendation issued. It seems that policy makers should pay more
attention to the size and to the banks’ liquidity levels, rather than to
profitability variables.
5.4.3 Risk Analysis: Total and Systematic Risk
When making our analyses about the effect of short selling on stock price
risk attitude, we divided our results into overall and systematic risk
measure (see Table 5.6).
100
Table 5.6 Total and systematic risk analysis

Panel B USA Size Profitability Leverage Liquidity Expectations Full model
Coeff. St. Dev Coeff. St. Dev Coeff. St. Dev Coeff. St. Dev Coeff. St. Dev Coeff. St. Dev
Overall risk
d_cur_mkcapt 0** 0 0.0000 0.0000
d_tot_com 0 0 0.0000 0.0000
d_assetturnover 0.1173*** 0.0662 0.1173*** 0.0662

d_ebit 0 0 0.0000 0.0000
d_EPS_an −0.0004** 0.0002 −0.0014* 0.0008
d_EPS_growth 0*** 0 0.0000*** 0.0000
d_grossmargin −0.0001 0.0039 0.0008 0.0038
d_roa 0.0026*** 0.0010 0.0020*** 0.0010
d_roc −0.0020*** 0.0005 −0.0019*** 0.0005
d_curratio −0.0207 0.0211 −0.0034 0.0184
d_lt_db_t 0.001691*** 0.0009 0.0008 0.0009
d_cf_net 0 0.0000 0.0000 0.0000
d_cf_in 0.0000 0.0000 0.0000*** 0.0000
d_best_an −0.0203*** 0.0062 −0.0110 0.0069
d_best_targ_price −0.0001** 0.0000 −0.0003 0.0002
Systematic risk
d_cur_mkcapt 0.0000 0.0000 0 0
d_tot_com 0.0000 0.0000 0*** 0
d_assetturnover 0.0585*** 0.0314 0.0616*** 0.0305
d_ebit 0.0000 0.0000 0 0
d_EPS_an −0.0001 0.0001 −0.0009*** 0.0004
d_EPS_growth 0** 0.0000 0 0
d_grossmargin −0.0006 0.0019 −0.0003 0.0018
d_roa 0.0009*** 0.0005 0.0006*** 0.0005
d_roc −0.0007*** 0.0002 −0.0006*** 0.0002
(continued)
Panel B USA Size Profitability Leverage Liquidity Expectations Full model
Coeff. St. Dev Coeff. St. Dev Coeff. St. Dev Coeff. St. Dev Coeff. St. Dev Coeff. St. Dev
d_curratio −0.0109 0.0090 −0.0050 0.0087
d_lt_db_t 0.0007*** 0.0004 0.0004 0.0004
d_cf_net_in 0 0 0 0
d_cf_cash_in 0 0 0*** 0
d_best_an −0.0052* 0.0029 −0.0005 0.0033
d_best_targ_price 0 0 −0.0002*** 0.0001
In this table, we report the effect of short selling restrictions on total and systematic risk. To this end, we perform pooled regressions to understand what banks reacted better
and worse with respect to fundamental firms
P P P P P X X X
Mod 3 : DOverallRIski;t ¼ a þ bj DXj;i;t þ kj DWj;i;t þ cDExpk;i;t þ dDLevk;i;t þ nDSizek;i;t þ ei;t and Mod 4 : DSysti;t ¼ a þ bj DXj;i;t þ kj DWj;i;t þ cDExpk;i;t
j j k k k
j j k
P P
þ dDLevk;i;t þ nDSizek;i;t þ ei;t where OverallRIski;t represents standard deviations computed in an estimation window of 252 days; Systi;t stands for systematic risk, also
k k
known as “undiversifiable risk,” affecting the overall market, not just a particular stock or industry; Xj;i;t is a vector of control variables accounting for the level of Profitability
(ASSET_TURNOVER, EPS_ANNUALIZED, EBIT, EPS_GROWTH, GROSSMARGIN, RETURN_ON_ASSET, RETURN_ON_CAP); Wj;i;t is a vector of variables representing the level of firm
Liquidity (CF_CASH_FROM_OPER, CF_NET_INC); dLevk;i;t includes control variables accounting for Leverage (CURRATIO, LTBTBT); Expk;i;t is a vector of control variables, stands for
Expectations (BEST_ANALYST, BEST_TARGET_PRICE); and lastly nSizek;i;t is a vector accounting for Size control variables (CURR_MAR_CAP, TOT_COM_EQY) of each i-th stock
taking into account (d_ represents the difference of the variable values at the beginning and 1 year after the short selling ban). ***, **, * denote that estimates are statistically
significant at the 1, 5 and 10% levels
101
With regard to overall risk and systematic risk dynamic models,

hereafter, we expose some main findings. We focus in particular on the
category of variables that from present analysis, they result the most
effective in the containment of overall and systemic risk component.
Profitability: US banks appear to be effective in reducing the overall
risk by increasing the level of earnings per share annualized. The growth
of earnings per share is also statistically significant in the reduction of
overall risk. Therefore, the policy makers must pay deep attention to
dynamics of earnings of banks, in order to better understand their impact
on the local banking system and decisions of short selling ban. Also, the
increase of profitability in terms of ROC is effective in the overall risk
containment. The increase in ROA is not effective in overall risk con-
tainment. So, the American banking system seems to be more sensitive in
reducing the risk on profitability return, calculated on bank capital. Also,
there is an evidence of the effectiveness of ROC ratio, as well as of the
overall risk. The increase of asset turnover results in increase of both
overall and systematic risk components. As for systematic risk, it contains
a specific risk component in relation to increase in earnings per share
annualized level, so it is related to overall risk too.
Liquidity: From our results, we evidence the increase of liquidity in
terms of cash flow. It appears to have a non-restraining effect of the
overall risk.
Expectations: Even the increase in target price, expressed by analysts,
is useful in the containment of the systematic risk.
Overall, we can conclude that policy makers of the US banks must
keep an eye on earnings and profitability variables in terms of ROC, as
fundamental firm, which is more effective in containing the overall and
systematic risk. These findings appear to be worthy of being considered
by local policy makers when they want to analyze the potential effects of
their decisions about short selling ban on the local banking system.
5.5 Discussion and Limitations of the Study

The use of short selling ban in different countries caught high attention
of policy modeling. The attitude to it was split into two points of view,
whether such actions could stabilize the market or not, and raised some
doubts concerning a speculative character of policy implementations and
rationality. It resulted in a certain influence on stock demand and
reduced stock prices in a different degree in every country. Our paper is one of
the first ones trying to explain the evidence of different bank price reactions
in terms of stock market conditions. Therefore, the degree of the
appropriate level of policy actions aimed to smoother the selling ban
pressure could be discussed if stock market and bank characteristics are
well-known by policy institutions. Recent events about Brexit gave room,
worldwide, for debating about the effectiveness and convenience of policy
measures such short selling bans. From the evidence of occurred trading
days, it is clear to see a noticeable increase in volatility on global markets.
Therefore, banks got record trading volumes when immediate aftermath
of Brexit and when usual trading levels were increased up to 10 times and
brought commissions 10 times more. Investors are seen very concerned
about banking which is expressed in large selling of bank stocks world-
wide. There was a risk-off trading on financial markets when investors
returned back to the less volatile asset class. At the same time, banks
experienced extremely negative volatility events, because their value
shrinks and their capability to continue their traditional banking activi-
ties was concerned. Indeed, the price fall corresponds to a destruction of
the value of banks, and it also influences the future lending policy.
For these reasons, it is very important to investigate what can happen to
banks under the short selling ban regime, in terms of both potential
destruction of value and increase in volatility. To this end, in this section,
we try to light policy implications related to our results. To understand
better how these situations can be read in light of short selling ban mea-
sures, it is useful to start by summarizing our results and then identify the
policy ideas related to various evidence we found in the same order we
present in empirical result section. With respect to fundamental firm
analysis, and focusing just on ban periods, for the US banks, we do not
observe any improvement in the profitability indices. It might also be due

to the short period taken into account. Also, 1 year later the start of the ban
period, we have no evidence of improvement. Likewise, there is slight or no
improvement with regard to the level of leverage and liquidity. On ana-
lyzing these results from a policy maker’s point of view, authorities should
think of implementing this intervention in market conditions not much
compromised. In fact, we observe improvements neither in profitability
nor in leverage and liquidity bank conditions. It makes us realize that short
selling ban actions cannot be implemented to achieve improvements in
economic characteristics of the banks. They can only become buffer actions
temporary stressful conditions of market volatility.
When it comes to analysis of price changes in the end of the ban
period and 1 year after the end of ban interventions, we conclude that it
is very important for policy makers to be aware of, which bank charac-
teristics are particularly sensitive to short selling ban, before it actually
starts.
When we analyze the specific characteristics of banks to decide which
fundamental firm should be considered, we turn to overall response in
terms of stock price change, return results and unpredicted evidences.
To be specific, we found that the classical profitability variables (ROA,
ROC and EBIT) and those related to the levels of earnings are not
effective in contrasting the price fall. In addition, evidences suggest that
policy makers consider positive contribution of recommendation (as
market expectation) on stock price. Therefore, the policy maker must not
pay too much attention to the current banks’ profitability, before taking
actions on ban restriction.
In fact, he should be focused on checking the banks’ health, by
considering the most long-term debt (as leverage indicator) and levels of
liquidity (measured by cash flow from net income). These two funda-
mental firms in fact, in the present context, have shown their effective-
ness in supporting the stock price. Policy makers should also pay
attention to the size of the bank, measured by market capitalization, due
to the fact that the bigger the market size, the higher the stock price.
Under this perspective, it might be useful to hypothesize differentiated
short selling interventions according to the size of the bank.
For volatility containment that is connected with stock price fall, it is

usually the second milestone to the base of the interventions of short
selling ban. Specifically, policy makers must pay deep attention to the
dynamics of an indicator of profitability, the earnings (EPS annualized
and growth) of banks, in order to reduce the impact of the overall risk on
the local banking system. Under this perspective, policy makers should
also consider the increase of profitability to be effective in terms of ROC,
when in the overall risk containment. Our evidences could be of some
interest also in the perspective of recent stress test exercises. Since stress
test is conducted on the assumption of a static balance sheet that is made
by banks with respect to the P&L, revenue, and costs, it all should be in
line with the constraints of zero growth and a stable business mix. Under
this perspective, it might be interesting to national policy makers to
understand how banks react to short selling bans, and mostly, what kind
of fundamentals are there for different market reactions under a short
selling ban regime. Since short selling restrictions are implemented in
high volatility and price fall scenario, its analysis could enrich the
knowledge of how banking systems could react in adverse conditions that
represent the same logic on which stress test exercises are based on.
We conclude this study by presenting some limitations of present
analysis. We think that we mostly deal with two kinds of limitations.
From one side, the ban period occurs in a limited time period, so we face
the limited data variability that means, changes in firm fundamental
occur in months and so, if we deal with a very short period of time, this
means that we have limited availability in changing of fundamental firm
data. This could, in some way, affect estimations we obtain by running
statistical models. In this context, we could also have some difficulties
(the most serious of the previous case to tell the truth) in using macro
variables that in some ways could be useful in describing local banking
systems. From the other side, we have a well-known problem of an
analyst interested in the estimation of a treatment effect of a given policy
program. Here we try to perform ex-post evaluation of policy actions
(namely short selling bans) via evidence-based statistical analysis. From a
statistical point of view, some recent studies (see, e.g., Cerulli 2012)
claim that it needs to perform counterfactual causal analysis in order to
better discriminate evidences coming out by performing econometric
models. In our case, we do not think it is useful to perform counter-

factual analysis, because we focus only on one sector (banking indeed),
and ban measures we examined were focused on a single sector. We
prefer to think of our evidence as descriptive statistics results and not as
casual effects of a certain policy action, because in this case, we would
have to approach, even with limitations, what we have already discussed
but with different statistical tools.
References
Appel I., and C. Fohlin. 2010. Shooting the messenger? The Impact of Short Sale
Bans in Times of Crisis. Pennsylvania: Department of Finance, the Wharton
School, University of Pennsylvania, 1–35.
Bali, Turan G., Stephen J. Brown, and Mustafa Onur Caglayan. 2012.
Systematic risk and the cross section of hedge fund returns. Journal of
Financial Economics 106 (1): 114–131.
Battalio, Robert H., Hamid Mehran, and Paul H. Schultz. 2012. “Market
declines: What is accomplished by banning short-selling?” Permanent DOI at
https://www.newyorkfed.org/medialibrary/media/research/current_issues/ci18-5.
pdf.
Beber, A., and Macro Pagano. 2013. Short-selling bans around the world:
Evidence from the 2007–2009 crisis. Journal of Finance 68: 343–381.
Boehmer, E., C.M. Jones, and X. Zhang. 2013. Shackling short sellers: The
2008 shorting ban. The Review of Financial Studies 26 (6): 1363–1400.
Cai J., and Z. Zhang. 2011. Leverage change, debt overhang, and stock prices.
Journal of Corporate Finance 17 (3): 391–402, ISSN 0929-1199, http://
dx.doi.org/10.1016/j.jcorpfin.2010.12.003.
Campbell, John Y., Christophe Polk, and Tuomo, Vuolteenaho. 2005. Growth
or glamour? Fundamentals and systematic risk in stock returns. National
Bureau of Economic Research No. w11389.
Cerulli, Giovanni. 2012. “ivtreatreg: A new STATA routine for estimating
binary treatment models with heterogeneous response to treatment under
observable and unobservable selection.” CNR-Ceris Working Papers 03/12.
Charoenrook A., and H. Daouk. 2005. A study of market-wide short-selling
restrictions. New York: Department of Applied Economics and Management,
Cornell University 1–43.
Diamond, Douglas W., and Robert E. Verrecchia. 1987. Constraints on

short-selling and asset price adjustment to private information. Journal of
Financial Economics 18 (2), 277–311.
Felix L., Kräussl R., and P. Stork. 2014. The 2011 European short sale ban: An
option market perspective. LSF Research Working Paper Series, 14 (2), 1–45.
Ferrer, R.C., and A. Tang. 2016. An empirical investigation of the impact of
financial ratios and business combination on stock price among the service
firms in the Philippines. Academy of Accounting and Financial Studies Journal
20 (2): 104–115.
Grullon G., S. Michenaud, and J. Weston. 2015. The real effects of short-selling
constraints. The Review of Financial Studies 28 (6): 1737–1767.
Harris L.E., Namvar E., Phillips B. 2013. Price inflation and wealth transfer
during the 2008 SEC short-sale ban. Journal of Investment Management 2:
1–43.
Hasan I., N. Massoud, A. Saunders, and K. Song. 2010. Should short-selling be
restricted during a financial crisis? 1–48. Stern School of Business, New York
University: New York University.
Helmes, Uwe, Julia Henker, and Thomas Henker. 2010. “The effect of the ban
on short selling on market efficiency and volatility.” Permanent DOI at
http://epublications.bond.edu.au/cgi/viewcontent.cgi?article=1532&context=
business_pubs.
Ho, K.Y., B.D. Ernst, and Z.Y. Zhang. 2011. Assessing the dynamic
relationship between small and large cap stock prices. MODSIM 2011—
19th International Congress on Modelling and Simulation—Sustaining Our
Future: Understanding and Living with Uncertainty, p. 1554.
Khan M.S., Harald Scheule, and E. Wu. 2016. Funding liquidity and bank risk
taking. Journal of Banking & Finance, Available online 21 Sept 2016, ISSN
0378-4266, http://dx.doi.org/10.1016/j.jbankfin.2016.09.005.
Kolasinski, Adam C., Adam Reed, and Jacob R. Thornock. 2013. Can short
restrictions actually increase informed short selling? Financial Management
42 (1), 155–181.
Mattarocci G., and G. Sampagnaro. 2010. Financial crisis and short selling: Do
regulatory bans really work? Evidence from the italian market. Academy of
Accounting and Financial Studies Journal 14 (4): 1–23.
Maudos, J. 2017. Income structure, profitability and risk in the European
banking sector: The impact of the crisis. Research in International Business and
Finance 39: 85–101, Part A, ISSN 0275-5319, http://dx.doi.org/10.1016/j.
ribaf.2016.07.034.
Papanikolaou N.I., and C.C.P. Wolff. 2014, Oct. The role of on- and off-balance-
sheet leverage of banks in the late 2000s crisis. Journal of Financial Stability 14:
3–22, ISSN 1572-3089, http://dx.doi.org/10.1016/j.jfs.2013.12.003.
Premti A., L. Garcia-Feijoo, and J. Madura. 2016. Information content of
analyst recommendations in the banking industry. International Review of
Financial Analysis, Available online 21 Nov 2016, ISSN 1057–5219, http://
dx.doi.org/10.1016/j.irfa.2016.11.005.
Savor, Pavel, and Mungo Wilson. 2016. Earnings announcements and
systematic risk. The Journal of Finance 71 (1): 83–138.
Schwartz L., and K. Norris. 2009. The impact of temporary short selling
restrictions on the volatility of financial stock prices: does firm size matter? In
Academy of Business Education, Proceedings 10 (19).
Shieh, S.-J., C.-Y. Lin, and P.-H. Ho. 2012. Large changes in stock prices:
Market, liquidity, and momentum effect. Quarterly Review of Economics and
Finance 52 (2): 183–197.
6
Reputational Risk in Banking: Important
to Whom?
Ewa Miklaszewska and Krzysztof Kil
6.1 Introduction
Protecting a financial institution’s reputation is among the most signif-
icant challenges facing financial firms, and trust in the integrity of the
financial sector is the cornerstone of its stability and growth. The
financial crisis of 2007–2009 and the post-crisis restructuring period
have brought an increased interest in the reputational risk, particularly in
the banking and financial sector. Crisis and post-crisis restructuring
always results in an increased interest in the issues of trust and corporate
culture, as scandals and excesses of the pre-crisis period come to light,
and the amounts spent to rescue banks raise public opposition (Walter
2013). Moreover, as the empirical research has indicated, the reputa-
tional risk increases with the scale and profitability of banks, making the
subject even more relevant in a global system characterized by a highly
E. Miklaszewska (&) K. Kil

Cracow University of Economics, Kraków, Poland
e-mail: uumiklas@cyf-kr.edu.pl
110 E. Miklaszewska and K. Kil
concentrated banking markets (Fiordelisi et al. 2011). The crisis caused

multibillion losses and revealed problems with strategic priorities and the
failure of risk management systems in large global banks. Consequently,
there has been a renewed interest in the creation of stable and functional
risk culture in global banks. Thus, the aim of this chapter is to analyze
why reputational risk is important for banks, and what are the incentives
to manage it.
Reputational risk is often analyzed within an operational risk frame-
work. The Basel Committee (BCBS 2001) described the latter as one of
the three main categories of banking risks and defined it as a possibility of
direct or indirect loss resulting from inadequate or failed internal pro-
cesses, actions of people or systems, or losses related to the impact of
external events. Although the definition was quite broad, the reputational
risk, as well as the strategic one, has not been included. The methodology
to manage and measure operational risk has been advancing rapidly in
recent years, fueled by a number of well-publicized scandals (the bank-
ruptcy of Barings, problems of Société Générale due to rogue traders and
the Allied Irish Bank, and UBS due to unauthorized trading), and also,
huge sums paid by banks and insurance companies after 2008 crisis to
settle allegations of sales abuses illustrate the point. However, as it took
over a decade to develop an acceptable infrastructure for operational risk
management, the reputational risk is only at the beginning of a similar
process.
Reputational risk is not a new concept, but the efforts to manage it as a
self-standing type of risk and not within an operational risk framework
are quite recent. However, it is more difficult to manage reputational risk
than other risk categories, as it is difficult to define and quantify, or
separate it from the impact of other events (ACE 2013). Consequently,
in the empirical part, this chapter proposes a methodology to measure
reputational risk, based on the bank stakeholders’ perspective. The rep-
utational risk is approximated by an integrated indicator: Stakeholder
Reputation Score (SRS). Then, panel regression models are used to
examine its impact on bank performance, for listed banks in Central and
Eastern Europe (CEE-11). The aim of the empirical part is to analyze
whether there is a reputational premium, i.e., what are the incentives to
manage the reputational risk in banks.
The chapter is organized as follows: Sects. 2–3 review the approaches

to define the reputational risk, Sect. 4 analyzes the literature on factors
causing reputational risk, Sect. 5 reviews the approaches to measure
reputational risk, Sect. 6 describes the empirical methodology and
summarizes the results of the panel data models aiming at measuring the
reputational performance premium for CEE banks, approximating rep-
utational risk by Shareholder Reputational Score, and the last section
concludes the chapter.
6.2 Reputational Risk from a Regulatory

Perspective
Risk appears with every banking product and operation, and managing
risk constitutes an everyday bank activity. Risk can be defined as
uncertainty concerning the return or outcome of an investment or an
action. Risk management is a process by which managers identify, assess,
monitor, and control risks associated with financial institutions’ activities
(Koch and MacDonald 2015). Its objective is to minimize negative
effects on the financial result and capital of a bank. However, in financial
institutions, risk can be treated both as a threat and also as an oppor-
tunity (Marcinkowska 2014). Banks manage risk at many levels, taking
account of both macro- and micro-factors, in many cases external to the
decisions taken by bank. In many cases, risk is interconnected, both
within a bank and in the whole system. Risk management encompasses
the process of identifying risks to the bank, measuring exposures,
ensuring that an effective capital monitoring program is in place, mon-
itoring risk exposures and corresponding capital needs on an ongoing
basis, taking steps to control or mitigate risk exposures, and reporting to
senior management and the board on the bank’s risk exposures and
capital positions (BCBS 2011). In the future, the new challenges will be
coming from expanding regulations, raising customers’ expectations due
to technological progress and the emergence of new types of risks
(McKinsey 2015).
Historically, the efforts in managing risk by banks tend to focus on

credit and market risk. However, risk management in banking has been
transformed over the past decade, largely in response to regulations that
emerged from the global financial crisis. Reputation risk was not inclu-
ded in the recommendations of the Basel Committee on the modeling of
risk in the banking sector. Basel II (2004) and Basel III (2010) kept
reputational risk out of pillar one capital requirement, and reputational
risk is currently not subject to any specific capital requirements in the
EU. Capital Requirements Directives applicable to EU countries require
only that the competent authorities evaluate reputational risks arising
from securitization transactions and that financial institutions develop
methodologies to assess the possible impact of reputational risk on
funding positions (Dey 2016). In the USA, reputational risk is one of the
Federal Reserve System’s categories of safety and soundness and fiduciary
risk (credit, market, liquidity, operational, legal, and reputational) and
one of the three categories of compliance risk (Business Insurance 2016).
Reputational risk—damage to an organization through loss of its
reputation—can arise as a consequence of operational failures, as well as
from other events. Both operational and reputational risks belong to a
similar area, as operational problems can have negative consequences for
bank reputation, affecting client satisfaction and shareholder value.
However, those risks can also include a broader set of incidents, such as
fraud, privacy protection, legal risks, and physical (e.g., infrastructure
shutdown) or environmental risks. In light of the significant number of
recent operational risk-related losses incurred by banks, in June 2011, the
Basel Committee published the “Principles for the Sound Management
of Operational Risk,” which incorporated the lessons from the financial
crisis. The eleven principles cover governance, risk management envi-
ronment and the role of disclosure, and address the three lines of defense:
business line management, an independent operational risk management
function, and an independent review. In 2014, the Committee con-
ducted the review in the form of a questionnaire, involving 60 system-
ically important banks in 20 countries, in which the banks self-assessed
their implementation of the principles. A key finding of the review was
that banks have made insufficient progress in implementing the princi-
ples (BCBS 2014). Hence, in 2014, the Basel Committee proposed a
revision to its operational risk framework that set out a new approach for
calculating operational risk capital. In addition, the Financial Stability
Board stressed the importance of operational risk in the post-crisis
environment, defining it as a synthetic one, including people risk, out-
sourcing risk, internal and external fraud, money laundering, and tech-
nology risk (FSB 2012).
In 2009, the Basel Committee passed the document addressing the
need to strengthen risk management by banks, in which the reputational
risk was defined as a multidimensional process, based on the perception
of other market participants (BCBS 2009). Reputational risk was
explained as the actual or potential risk related to earnings or capital,
arising from negative perception of financial institutions by the current
and potential stakeholders (customers, counterparties, shareholders,
employees, investors, debt-holders, market analysts, other relevant par-
ties, or regulators) that can adversely affect a bank’s ability to maintain
existing, or establish new, business relationships and its continued access
to sources of funding, including the interbank market or the securitiza-
tion processes. In this document, the Basel Committee stressed the need
to manage reputation risk, identifying its sources and taking it into
account when testing the resilience of a bank business model to external
shocks (BCBS 2009). The Fed’s Commercial Bank Examination Manual
defines reputational risk as “the potential that negative publicity
regarding an institution’s business practices, whether true or not, will
cause a decline in the customer base, costly litigation or revenue
reductions” (Business Insurance 2016).
6.3 Reputational Risk as Internal

and External Factor
Risk management is result oriented, with different priorities given to
avoidance of operational and reputational problems and a different time
horizon for maximizing the value of the company. The reputational risk
is associated with faulty strategy, poor management and leadership, or a
wrong system of incentives, inadequate supervision, and problematic
corporate culture. Reputational risk can be defined as the risk of eco-

nomic losses associated with a negative image of the bank by the clients,
supervisors, regulators, and the public. This and similar definitions
stressed that reputational risk is multidimensional and reflects the per-
ception of other market participants.
It can also be defined as the risk to bank goodwill, which is not
associated with deterioration of book value and is typically reflected in a
falling stock price (Walter 2013). There is also a problem of time frame.
In most cases, the effects of a scandal or unexpected loss are immediate.
The loss is seen as a signal that the company has a weak control envi-
ronment. Shareholders may also sell shares if they believe that future
losses are inevitable. However, there are also cases of more prolonged
problems with corporate culture, which gradually erode customers’ and
business partners’ trust. In some cases, reputational problems have a
negative impact on the financial results, but there are also the opposite
cases (Marcinkowska 2013).
Reputational risk is not regulation or compliance driven, but deter-
mined by stakeholder expectations. Steinhoff and Sprengel (2014)
observed that risk awareness is probably the most important factor for
risk reduction, so it should be placed inside the corporate governance
framework, particularly in “who is responsible for what” approach.
However, corporate culture is also a very broad concept and can be
defined in many ways (Guiso et al. 2006). The development of corporate
culture is a continuous process, where the results are visible in the long
term. Its definitions emphasize that it rests on a set of values shared by a
community, which affects its organization and motivates behavior within
the organization (Carretta and Sargiacomo 2016). The period of crisis
often results in an increased interest in corporate governance; however,
changes in prudential regulations correcting errors in risk management
are usually easier than the long-term changes in the corporate culture of
market participants (Walter 2013). However, there are some mechanisms
which can be used in enhancing trust, such as codes of ethics, internal
anti-fraud systems, independent ethics audits, and reputational indices.
Indirect measures involve membership of professional associations or
self-regulatory organizations, which protect the reputation and discipline
among its members, setting standards in codes of conduct and
developing mechanisms of better risk assessment processes (Morris and

Vines 2014; Marcinkowska 2013).
Reputational risk is usually due not to incidental events, but is the
result of long-term poor decision-making processes. The causes are often
linked to the pressures on results, the asymmetry of the profit-to-risk
ratio, conflict of interest related to the complexity of bank business
models, and compensations based on bonuses (Walter 2013). Financial
services differ significantly from the industrial sector. Key stakeholders of
banks are depositors, creditors, and the government (insurance). As
banks are financed largely through debt, shareholders have a lesser
importance than in corporations. However, bank governance prioritizes
shareholder interests, particularly when ownership is concentrated in
institutional investors with a large risk tolerance. Consequently, gover-
nance of financial institutions may accept excessive operational risk,
which may erode shareholder wealth and may fail to meet the expectation
of other stakeholders (Dow 2014).
Inside the banking sector, reputation is often treated in the same way
as a “brand,” i.e., an intangible asset that can be impaired by operational
mistakes or inappropriate behavior. In this approach, reputational risk is
a derivative risk, arising as a result of damaging action (Steinhoff and
Sprengel 2014). Reputation may also serve as a cushion against losses,
i.e., companies with a better reputation suffered less severe declines in
market value during the crisis periods although the empirical evidence
varies in this respect—in some cases good reputation softens the impact
of failures; in others, it may be dangerous, as other objective indicators of
strength, such as capital or liquidity, may seem irrelevant. The third way
is not to treat it as an asset, nor as a kind of equity capital, but as a set of
obligations toward stakeholders, which have to be fulfilled (Steinhoff and
Sprengel 2014). Thus, reputation can be summed up as having three
main manifestations:
• reputation as asset (stakeholders’ goodwill),

• reputation as liability (stakeholders’ expectations), and
• reputation as capital (buffer against failure, helping to maintain
goodwill when failing to meet expectations).
The impact of reputation on performance is a direct consequence of

the interaction of those domains (Steinhoff and Sprengel 2014).
6.4 Reputational Risk in Global Surveys

The strategy of the largest global banks has evolved from simple, com-
mercial institutions, providing selected services for a specific customer
segment, to complex conglomerates, serving millions of customers in
many countries. Traditionally, the financial services industry worked
according to easily understandable principles, with clearly defined risk
profiles, but in the last 20 years those divisions were blurred, and new
players, such as hedge and equity funds, were offering para-banking
services (Rajan 2005). However, the strategy of a “financial supermarket”
and a “too big to fail” scale turned out to be very risky. Although among
the top causes of the global financial crisis was a systemic risk associated
with the activities of large global banks, after the crisis, their role has been
further strengthened. In many countries, post-crisis restructuring took a
form of mergers and acquisitions, particularly of investment banks by the
universal ones in the USA or merging the nationalized banks to control
losses (the Netherlands and the UK). So the question of managing the
reputation risk in the process of acquisition is another important chal-
lenge (Schoenmaker 2011; Dermine 2006).
The 2008 financial crisis had a significant effect on bank reputation
and trust, and only recently can we observe a gradual rebound of trust:
Financial services have recorded an 8-point increase from 43% in 2012
to 51% in 2016 on a global basis. Financial services, however, are still the
least trusted industry among those surveyed by the Edelman Trust
Barometer (2016). Inside the industry, employees are more trusted than
senior executives and CEOs to communicate about topics like financial
earnings, crises, and the treatment of customers. In the USA, the
Reputation Institute compared the financial industry problems with past
reputation of tobacco firms. In the post-crisis period, the financial sector
has been obliged to pay an incredible amount of litigation expenses, with
the most notable being JP Morgan paying a 13 billion dollar settlement
to the US government over behavior leading to the crisis in 2014,

Deutsche Bank investigated for tax evasion and money laundering, in
addition to Libor fixing in 2012, or large banks fined for the Libor
scandal in 2015. However, in 2016 for the first time, the large banks
have risen in the US ranking—of the 33 banks evaluated, ten banks had
an “excellent” reputation among their customers, compared to eight in
2015 (American Banker 2016). Other surveys have also shown that
inside the banking industry, the best reputation has divisions related to
new technologies, e.g., Internet banking and ATM, though not tele-
phone banking (Ernst and Young 2014).
As early as in 2005, the Economist Intelligence Unit Report observed
that protecting a firm’s reputation is the most important and difficult task
facing a firm’s managers and reported that in a survey of 269 senior
executives, responsible for managing risk, reputational risk emerged as
the most significant threat to business out of a choice of 13 categories of
risk. Reputational risk was defined as an event that undermined public
trust in bank products or brand (The Economist 2005). Reputation is
based on aggregate past experience; however, it is directed toward the
future and reflects the expectations concerning the firm (Edelman Trust
Barometer 2014). Customers satisfied with the services of the bank have
a greater loyalty which helps to improve the bank image and its com-
petitive position (Fiordelisi and Molyneux 2009). In contrast, problems
with bank reputation can lead to (Eccles et al. 2007):
• loss of current or prospective customers,

• loss of employees or managers in the organization,
• departure of current or future business partners, and
• an increase in the cost of financing through a loan or capital markets.
The growing awareness of reputational risk is reflected in an annual

survey conducted by the European Banking Authority and reported in
“Risk Assessment of the European Banks.” This document includes a sec-
tion on reputational risk, particularly assessing its impact on consumer
confidence (EBA 2014, 2015, 2016). The reports showed a growing
awareness of the reputational risk in the European banking sector, as
indicated by 33% of responding banks in 2013, 44% in 2014, and 68%
in 2015. Numerous case studies and empirical studies showed that

reputational risk is particularly important for large global banks and those
with relatively low capitalization, so it should be an important subject of
supervisory concern. According to EBA reports, particularly a detri-
mental impact on consumers had failures with regard to rate
benchmark-setting processes, the misselling of banking products, and
more recently misconduct related to foreign exchange rates, violations of
trade sanctions and redress for payment protection insurance, and floors
for mortgage loans at variable interest rates. The scope of identified
detrimental business practices remains wide and misconduct costs remain
high. The share of banks indicating that they have paid out more than
one billion euros in compensation, litigation, and similar payments
increased in 2015 to 32% of participating banks (16% in 2014 and only
8% in 2013) (EBA 2014, 2015, 2016). Efforts to adjust culture and risk
governance are the most widely considered approach to address reputa-
tional and legal risks (85% in 2016), an increase from less than 50% of
respondents in previous surveys. However, in the 2016 Report, only
about 10% of surveyed banks indicated their intention to adjust products
and business models in an effort to address reputational and legal risks.
Kaiser (2014) analyzes two surveys conducted by KPMG among the
G-SIBs (the Global Systemically Important Banks) in 2013 and 2014
and responded to by ten banks and a survey of the German banks,
responded to by 18 institutions, 13 of which belong to the 20 biggest
German banks in 2012. In the surveys, 60% of both global and German
banks asserted that reputational risk stands on its own, rather than being
a consequential risk, or triggers to other risks; however, most banks did
not include it in their risk inventory and admitted that it is not explicitly
addressed in their risk strategy. Another question showed that only 55%
of the G-SIBs and 60% of the German banks prioritized their stake-
holders, in order to manage reputational risk more efficiently. German
banks gave the highest priority to customers, while global banks gave top
priorities to customers, employees, and regulators. The surveys demon-
strated that banks put the main emphasis on the self-assessment of
reputational risk, only supplementary emphasis on expert opinions,
interviews with senior management, and analysis of press and social
media, and that they register and report losses due to reputational risk
mainly as a part of an operational risk database, so although banks were
aware of the need to include reputational risk in their overall risk
mapping, in everyday life, they dealt with it in an operational risk
management framework.
6.5 Problems with Measurement

of the Reputational Risk
Efforts to manage operational risk have been successfully quantified in
the last decade, but for reputational risk, the typical approach is still to
monitor it inside the broadly defined “risk culture.” What gets measured
gets managed (Diermeier 2008), but quantification of reputation risk is
extremely difficult as there is no universally accepted methodology and
the concept is broad. If we define reputational risk as unexpected losses
due to the reaction of stakeholders to an altered perception of an insti-
tution (Kaiser 2014), there are many possible ways of approximating this
risk. Moreover, reputational risk does not act in isolation and, on the
contrary, is interrelated to many other types of risks. Some sources of
gain/loss in the reputational capital include economic performance,
stakeholder interface, and legal interface, which can be reflected in client
flight, loss of market share, investor flight and increase of cost of capital,
and talent flight and increase of contracting costs (Walter 2016).
Assuming that reputational risk is managed through strong corporate
governance, another approach is to create indexes which measure the
quality of firms’ corporate governance structure and link it to stock
price-based performance of the company, assuming that the change in
corporate governance index is a signal of quality of firm management
(Fox et al. 2016).
Empirical studies typically focus on various surveys, case studies, or
media coverage of detrimental events. There is also a lack of tools to link
reputational risk with financial performance, and it is unclear how rep-
utation risk can impact capital (Diermeier 2008). In many companies,
reputational problems are still considered rather as a problem of public
relations than a strategic one, and the response is frequently inadequate

to the scale of the damage. The problem of reputational risk measure-
ment is still aggravated for CEE banks, as the stock markets are not
efficient in discounting information, so the panel data models using stock
market information may be misleading.
Assessing reputational risk is most often not an objective process, but
rather it is a subjective assessment that could reflect a number of different
factors. Reputation could be perceived as an intangible asset, synony-
mous with goodwill, which is difficult to measure and quantify.
Consistently strong earnings, a trustworthy board of directors and senior
management, loyal and content branch employees, and a strong cus-
tomer base are just a few examples of positive factors that contribute to a
bank’s good reputation (Business Insurance 2016).
Establishing a strong reputation provides a competitive advantage.
A good reputation strengthens a company’s market position and increases
shareholder value. It can even help attract top talent. Communication
between a bank and its stakeholders can be the foundation for a strong
reputation. Bank examiners may consider whether an institution
responds to customer concerns; whether the stock analyst recommends
buying or selling and why; and what the shareholders, employees, or
general public are saying about the institution. They also consider
whether the institution is expanding outside its normal geographical area
and is supportive of the community. On-site, examiners will talk to both
bank employees and management to get a sense of corporate ethics.
Examiners will assess whether an institution’s expertise is adequate and
controls are in place to oversee growth if the institution should engage in
riskier products or enter into new business lines (Brown 2016).
Also, the agencies, such as Standard & Poor’s, Moody’s, and Fitch,
have significantly increased their emphasis on reputational risks related to
corporate governance. The rating agency’s primary focus is the ability
and willingness of an entity to make full and timely payment of debt
service on its financial obligations. However, a damaged reputation can
significantly affect the performance and, ultimately, the ability to borrow
capital. For example, S&P issued a statement saying that costs associated
with the Costa Concordia disaster had negatively affected the firm’s
operating performance in 2012. Another example of the importance of
reputation in obtaining the rating score is public universities in the USA,

which rely heavily on their reputation and brand as a strategic asset
(Business Insurance 2016).
A measure that is sometimes used is the difference between the
immediate costs of a crisis and damage to a firm’s market capitalization in
the period following a crisis event (ACE 2013). Another frequent
approach in modeling reputational risk is to analyze it within an opera-
tional risk framework, assuming that operational loss events can lead to
significant reputational losses, and to check the impact of bank reputa-
tional problems on bank market capitalization (Perry and De Fontnouvelle
2005). The reputational loss is defined as market value loss that exceeds
announced operational loss (Eckert and Gatzer 2015). Another frequent
approach is to conduct an event study analysis of the impact of operational
loss events on the market values of financial institutions by examining a
firm’s stock price reaction to the announcement of particular operational
loss events such as internal frauds, estimating the Reputational Value at
Risk at a given confidence level, which represents the economic capital
needed to cover reputational losses over a specified period (Micocci et al.
2009).
6.6 Empirical Analysis of the Reputational

Risk in the CEE Banking
Reputation can be perceived not only as a problem, but as an asset,
contributing to a performance premium. The empirical part adopts this
approach, examining the relationship between an indicator of the rep-
utational risk (Shareholder Reputational Score) and bank performance.
To test the role of reputational risk for bank performance in CEE-11
countries, the panel data model with fixed effects was used (with
Hausman and Breusch-Pagana tests), based on individual bank data from
Bankscope. In the sample, 42 banks listed at CEE stock exchanges were
analyzed (15 from Poland, 12 from Croatia, 4 from Bulgaria and
Slovakia, 3 from Romania, and 1 from the Czech Rep., Hungary,
Lithuania, and Slovenia), for which the rating information from at least
one of the three major agencies: Standard & Poor’s Rating Services,
Moody’s Investors Service Inc., or Fitch Ratings Ltd., was available.
The first step was to construct an index of reputational risk; the
following one was to test its impact on bank performance. In the model,
reputation risk was represented by a three-dimensional, synthetic index:
Stakeholder Reputation Score (SRS). The index is based on the per-
spectives of three major bank stakeholders, according to the following
formula:
SRS: (a) market participant perspective + (b) client perspective + (c)
investor perspective.
Those three perspectives were approximated by:
SRS: (a) credit agencies’ bank ratings + (b) deposit growth + (c) bank
stock returns.
There is a long debate on the relevance of the rating information and
rating agencies’ credibility, particularly after the global crisis (Grothe
2013; Eckert and Gatzer 2015), but nevertheless credit rating encom-
passes a broach range of information. Credit ratings express credit rating
agencies’ forward-looking opinion about the creditworthiness of an
obligor—the capacity and willingness to meet its financial obligations in
full and on time (S&P 2016) and represent an evaluation of the quali-
tative and quantitative information on the prospective debtor. In the
model, the ratings were employed both at a country level (CR) and at the
bank level, included in the SRS index.
The three dimensions in SRS (a, b, and c) were calculated as follows:
a. ratings: scores from major credit agencies were used and the average
score (arithmetic mean, in points) was established as in Table 6.2, on
a scale of 1–16, adjusted by rating perspective of ±0.5% points; a
stable outlook did not cause adjustments in the assessment;
b. deposits: the annual growth rate of current deposits from the
non-financial sector was used (converted to points); and
c. stock return: the annual rate of return from bank stock was used,
adjust by splits and dividends paid (in points) (Table 6.1).
Table 6.1 Scoring scale used in the model

Rating agency assessment Model score
S&P Fitch Moody’s
AAA AAA Aaa 16
AA+ AA+ Aa1 15
AA AA Aa2 14
AA− AA− Aa3 13
A+ A+ A1 12
A A A2 11
A− A− A3 10
BBB+ BBB+ Baa1 9
BBB BBB Baa2 8
BBB− BBB− Baa3 7
BB+ BB+ Ba1 6
BB BB Ba2 5
BB− BB− Ba3 4
B+ B+ B1 3
B B B2 2
B− B− B3 1
Point values of the three dimensions (a, b, and c) of the SRS were
calculated by assigning each year a numerical value to each decile for each
indicator and for the whole group, in the following way:
• 0 points for the median for the entire group in a given year;
• from −5 to −1 respectively for deciles from 1 to 5; and
• from 1 to 5 respectively for deciles from 6 to 10.
Consequently, the SRS index ranges from −15 to +15 points for the
three indicators and represents an approximation of the bank’s reputa-
tional risk.
The next step was to run a panel data model, for the period 2009–2014.
The dependent variables were the long-term, comprehensive indicator:
Multi-Level Performance Score (MLPS) and the short-term, simple indi-
cator: Return on Equity (ROE). MLPS was defined as the sum of points
awarded in five key areas for long-term evaluation of bank performance:
three performance indicators (ROE, cost-to-income ratio and loans-to-asset
ratio) and two sustainability indicators (Z-score and NPL) (Miklaszewska
and Kil 2015). Thus, MLPS = ROE + C/I + L/A + Z-score + NPL.
Table 6.2 Description of explanatory variables

Symbol Description Rationale/Data source
a. Macroeconomic variables
D GDP Real GDP growth rate Macroeconomic business cycle (World
(%) Bank: World Development Indicators)
HHI Herfindahl-Hirschman Banking market concentration
index for credit (BSCEE Review and ECB Database)
institutions
SB Total bank assets (% of Size of the banking sector
GDP) (Raiffeisen Research 2015)
CR Country LT credit rating Country credit standing (Bankscope,
rating agencies’ internet sites)
b. Bank-level variables (data source: Bankscope)
ln_TA Logarithm of total assets Bank size
(in USD)
SRS Reputational risk index Approximation of reputational risk
L_D Loans-to-Deposits ratio Bank funding risk
NeII_NoIOI Net interest income/ Income diversification (bank business
Total non-interest model)
operating income
S_TA Securities/Total Assets Market risk
LA_DSTF Liquid assets/Deposits Liquidity risk
and short-term funding
The score was calculated as follows: For each indicator, the whole group was
divided into ten deciles, and the median value is 0 (neutral); each subsequent
deciles above the median for the ROE, L/A, and Z-score ranged from 1 to 5,
and each successive deciles below the median had negative value and ranged
from −1 to −5. For C/I and NPLs, the signs were the opposite. This indi-
cator has a simple interpretation: The higher the value of the MLP score, the
better the assessment of the bank’s results. The panel data model with fixed
effects was used, which measured the impact of reputation risk (approxi-
mated by the SRS score) on bank performance, measured by the compre-
hensive index Multi-Level Performance Score (MLPS) and profitability
indicator (ROE). For robustness, bank stock rate of return (RR) as depen-
dent variable was also tested, but the SRS was insignificant for that model.
The explanatory variables are defined in Table 6.2.
The results of estimations for the reputational effects on bank per-
formance are summarized in Tables 6.3 (for the comprehensive MLPS)
and 6.4 (for the ROE).
Table 6.3 Panel data estimations for MLPS, CEE-11, 2009–2014

Control variables
const −79,050
0.121
D GDP 0.369 *
0.068
HHI −249,297 *
0.078
SB 2351
0.827
CR −3789 ***
0.008
ln_TA 7173 **
0.030
SRS −0.265 **
0.011
L_D 0.218 ***
0.000
NeII_NoIOI −0.012 **
0.017
S_TA −0.039
0.688
LA_DSTF 0.178 **
0.026
R2 0.856
R2 corrected 0.837
Note ***, **, * correspond to 1%, 5%, and 10% significance level
Source Own calculation
The estimation results presented in Tables 6.3 and 6.4 indicate that
analyzing bank performance, both approximated by short-term ROE and
by a comprehensive MLP score, the index of bank reputation SCR
(similarly like the country’s rating CR on a macroeconomic level) not
only has a positive impact, but on the contrary affects bank performance
strongly negatively, similarly as the HHI concentration index. Factors
with a positive impact on bank performance were the size of the bank, its
financing strategy, the asset risks, and the high level of GDP growth.
Thus, the empirical results are contrary to the expectations: For CEE-11
stock-listed banks, large risky banks with low reputational score were best
placed for best results, both in a short-term (ROE) and in a long-term
(MLPS) perspective.
Table 6.4 Panel data estimations for ROE, CEE 2009–2014

Control variables 2009–2014
const −187,278 *
0.082
D GDP 0.121
0.747
HHI −504,163 *
0.076
SB 21,042
0.288
CR −2037
0.424
ln_TA 12,325 *
0.072
SRS −0.357 *
0.081
L_D 0.168 **
0.048
NeII_NoIOI −0.003
0.672
S_TA 0.488 **
0.012
LA_DSTF 0.292 *
0.067
R2 0.639
R2 corrected 0.489
Note ***, **, * correspond to 1%, 5%, and 10% significance level
Source Own calculation
6.7 Conclusion
The reputational risk literature and surveys, analyzed in this chapter,
suggested that banks should treat reputational risk as a separate class of
risk and analyze it beyond the framework of operational risk and cor-
porate governance. It should not be narrowed down to “public relation”
response to crisis events, but treated as a strategic type of risk, with a
strong potential to harm the value of the company.
However, as the reputational literature and many case studies indicate,
it is very difficult to categorize and quantify reputational risk, as it can
arise as a consequence of other risks and many events. The panel data
models for listed banks in CEE-11 countries, analyzed in this chapter,
have also indicated that proper management of reputational risk may not
be important (and even harmful) for an assessment of bank performance,
which may explain why many banks dealt with reputational risk mainly
in the context of minimizing loss after a scandal. Consequently, there
seem to be incentives to disregard reputational risk in an operational and
strategic bank management and deal with it only with crisis events.
References
ACE. 2013. Reputation at Risk. London. www.acegroup.com/global-assets/
documents/EuropeCorporate/ThoughtLeadership/ace_reputation_at_risk_
july_2013.pdf.
American Banker. 2016. Survey of Bank Reputation. www.americanbanker.com.
BCBS—Basel Committee on Banking Supervision. 2001, January. Consultative
Document: Operational Risk, BIS.
BCBS—Basel Committee on Banking Supervision. 2009, July. Proposed
Enhancements to the Basel II Framework. Consultative Document Bank for
International Settlement.
BCBS—Basel Committee on Banking Supervision. 2011, June. Principles for the
Sound Management of Operational Risk, BIS.
BCBS—Basel Committee on Banking Supervision. 2014, October. Review of
the Principles for the Sound Management of Operational Risk.
Brown, W.J. 2016. Understanding Reputational Risk: Identify, Measure, and
Mitigate the Risk, www.philadelphiafed.org/bank-resources/publications/src-
insights/2007/fourth-quarter/q4si1_07.
Business Insurance. 2016, August 1st. Reputation Risk in Focus at Credit Rating
Agencies.
Carretta, A., and M. Sargiacomo, (eds.). 2016. Doing Banking in Italy:
Governance, Risk, Accounting and Auditing Issues. Milano: McGraw Hill.
Dermine, J. 2006. European Banking Integration: Don’t Put the Cart Before
the Horse. Financial Markets, Institutions & Instruments 15 (2): 57–106.
Dey, S.K. (2016). Reputational Risk in Banking—The Current Approach and a
Way Ahead, TCS Financial Solutions, www.tcs.com.
Diermeier‚ D. 2008‚ March. Measuring and Managing Reputational Risk. Risk
Management, 12 (3).
Dow, S. 2014. Managing Stakeholder Expectations. Reputational Risk Management

in Financial Institutions: Risk Books Incisive Media Investments Ltd.
Eccles, R., S. Newquist‚ and R. Schatz. 2007, February. Reputation and its Risk.
Harvard Business Review.
Eckert, Ch. and N. Gatzer. 2015. Modeling Operational Risk Incorporating
Reputational Risk: An Integrated Analysis for Financial Firms. FAU Working
Paper, www.vwrm.rw.fau.de/files/2016/05/Reputation_Risk_2015-08-14_WP.
pdf.
Edelman Trust Barometer. 2014, 2016. Trust in Financial Services. www.
edelman.com/insights/intellectual-property/2016-edelman-trust-barometer/
state-of-trust/trust-in-financial-services-trust-rebound.
European Banking Authority. 2014, 2015, 2016. Risk Assessment of the European
Banking System. https://www.eba.europa.eu/documents.
Ernst and Young. 2014. Building a Better Working World. Global Consumer
Banking Survey.
Fiordelisi, F., and P. Molyneux. 2009. Shareholder Value in Banking.
Basingstoke: Palgrave Macmillan.
Fiordelisi, F., S. Soana, and P. Schwizer. 2013, May. The Determinants of
Reputational Risk in the Banking Sector. Journal of Banking & Finance
37 (5): 1359–1371.
Fox, M.B., J.R. Gilson, and D. Palia. 2016, August. Corporate Governance
Changes as a Signal: Contextualizing the Performance Link. Working Paper
323, European Governance Institute.
FSB. 2012. Increasing the Intensity and Effectiveness of SIFI Supervision Progress
Report to the G20 Ministers and Governors. www.fsb.org/wp-content/uploads/
r_121031ab.pdf.
Grothe, M. 2013, December. Market Pricing of Credit Rating Signals. ECB:
Working Paper no. 1623.
Guiso, L., P. Sapienza, and L. Zingales. 2006. Does Culture Affect Economic
Outcomes? Journal of Economic Perspectives 20 (2): 22–48.
Kaiser, T. 2014. Reputational Risk Management across the World: a Survey of
Current Practices. In: Reputational Risk Management in Financial Institutions,
London: Risk Books Incisive Media Investments Ltd.
Koch, T.W., and S. MacDonald. 2015. Bank Management. Boston: Cengage
Learning.
Marcinkowska, M. 2013. Kapitał relacyjny banku: Kształtowanie relacji banku z
otoczeniem. t. I, Łódź: Wydawnictwo Uniwersytetu Łódzkiego.
Marcinkowska, M. 2014. Corporate governance w bankach, teoria i praktyka.

Łódź: Wydawnictwo Uniwersytetu Łódzkiego.
McKinsey. 2015, December. The Future of Bank Risk Management, McKinsey
Working Papers on Risk, McKinsey & Company.
Micocci, M., G. Masala, G. Cannas and G Flore. 2009. Reputational Effects of
Operational Risk Events for Financial Institutions, University of Cagliari
Working Paper, http://www.actuaries.org/AFIR/Colloquia/Rome2/Micocci_
Masala_Cannas_Flore.pdf.
Miklaszewska, E. and K. Kil. 2015. The Impact of the 2008 Crisis on the
Banking Sectors of the CEE-11 Countries: Multi Level Performance Index as a
Synthetic Measure of Bank Risk Adjusted Performance. Econometrics, 4 (50).
Morris, N., and D. Vines. 2014. Capital Failure: Rebuilding Trust in Financial
Services. Oxford: Oxford University Press.
Perry, J., and P. De Fontnouvelle. 2005, October. Measuring Reputational
Risk: The Market Reaction to Operational Loss Announcements, Federal
Reserve Bank of Boston.
Raiffeisen Research. 2015, June. CEE Banking Sector Report.
Rajan, R. 2005, November. Has Financial Development Made the World Riskier?
Working Paper Series 11728, NBER. http://www.nber.org/papers/w11728.
Schoenmaker, D. 2011, April. The European Banking Landscape after the Crisis.
Policy Paper, 12, DSF.
S&P: Standard & Poor’s Ratings Definitions 2016, February 1. www.
standardandpoors.com/ratingsdirect.
Steinhoff, C., and R. Sprengel. 2014. Governance as the Starting Point for a
Reputational Risk-Management Process. Reputational Risk Management in
Financial Institutions, London: Risk Books Incisive Media Investments Ltd.
The Economist. 2005, December. Reputation: Risk of Risks. The Economist
Intelligence Unit.
Walter, I. 2013. Reputational Risk and Financial Crisis. In: Global Asset
Management, eds. M. Pinedo and I. Walter. New York: Palgrave Macmillan.
Walter, I. 2016, February. Reputational Risk and Large International Banks.
Financial Market Portfolio Management, 30 (1): 1–17.
7
The Business Model of Banks: A Review
of the Theoretical and Empirical Literature
Stefano Cosma, Riccardo Ferretti, Elisabetta Gualandri,
Andrea Landi and Valeria Venturelli
7.1 Introduction
The business model (BM) has become a key concept in banking litera-
ture. The topic’s relevance is due to the impact of the crisis on bank
profitability and risk levels, leading to new challenges for bank managers,
analysts and regulators.
S. Cosma R. Ferretti E. Gualandri (&) A. Landi V. Venturelli

University of Modena and Reggio Emilia, Modena, Italy
e-mail: elisabetta.gualandri@unimore.it
S. Cosma
e-mail: stefano.cosma@unimore.it
R. Ferretti
e-mail: riccardo.ferretti@unimore.it
A. Landi
e-mail: andrea.landi@unimore.it
V. Venturelli
e-mail: valeria.venturelli@unimore.it
132 S. Cosma et al.
From the managerial point of view, the crisis has caused an in-depth
review of banks’ strategies and enhanced their ability to change/adapt
both their business mix and their market positioning in the different
strategic areas where they compete.
In the years since the outbreak of the crisis, three main drivers have
spurred a significant change in banks’ strategic plans.
Firstly, a new adverse economic context, with a combination of slow
economic growth and historically low levels of interest rates. Both phe-
nomena depress the prospects for traditional bank intermediation, since
they lead to less, higher-risk lending while simultaneously squeezing
profit margins on the loans–deposits circuit.
Secondly, the re-regulation introduced in the wake of the crisis is
triggering strategic changes in BMs to adapt balance sheet structures to
new regulatory requirements: liquidity, high-quality capital, more stable
funding resources and bail-inable debt.
A third driver concerns the structural configuration of the main banking
systems, which affects banks’ ability to handle the fast pace of technological
innovation and its impact on products and distribution channels.
This Darwinian economic context opens the question of which banks
are in the best position to succeed and, at the same time, which banks are
going to become the victims of this much more competitive arena.
Business model analysis (BMA) has become the conceptual framework
used by analysts and regulators in the attempt to identify banks’ main
strategic behaviours and their implications in terms of competitiveness
and future performance and stability.
As far as banking regulation and supervision are concerned, BMA is
the basis for a proactive response and aims to reveal the key vulnera-
bilities of the different banking business models (BBM). This conceptual
framework, embedded in the Supervisory Review and Evaluation Process
(SREP), has a central role in the 2015 and 2016 Thematic Review by the
Single Supervisory Mechanism (SSM). In banking supervision, BMA is
an important tool for revealing a bank’s main vulnerabilities in the short
run and the viability and sustainability of its strategic plans in the short
and medium terms. The supervisory assessment not only regards the risks
each bank has undertaken and therefore its vulnerability (idiosyncratic
7 The Business Model of Banks: A Review of the Theoretical … 133
risk in a microprudential perspective), but also its contribution to sys-

temic risk, in a macroprudential perspective.
For financial analysts and investors, the BM is an important factor in
the evaluation of banks’ ability to create value. In a phase of high
financial market volatility and high equity capital needs, it is crucial to
understand how the market assesses and evaluates a bank’s restructuring
process and its changes in strategies and business mix.
A key issue in BMA is the identification of banks’ BM types: which
variables and typical characteristics should be considered? Can these variables
be clustered into relatively homogeneous groups to develop a peer analysis
aimed at identifying the relative strengths and weaknesses of different BMs?
These questions are at the centre of recent banking literature, which
has revealed wide differences in the approaches used to classify the
strategic variables on which BM definition is based. This diversity
originates from conceptual schemes that are not always made explicit in
the literature and this means that results are not always easily comparable.
Our review of banking business model (BBM) literature aims to
deepen and specify the definition of BMs by drawing on the main
concepts adopted in strategic management literature.
Relying on a definition of BM that takes into account what banks do
and how they do it, we classify the plurality of contributions that address
the subject of banking BM themes and summarise our main findings. In
doing this, we specify the nexus linking BM literature with other
approaches to bank strategic choices, mainly in diversification studies.
Lastly, we relate the main pointers obtained from the literature reviewed
to the assessment scheme adopted by supervisory authorities to evaluate the
viability, sustainability and key vulnerabilities of banks’ current BMs.
7.2 The Definition of Business Model

In management literature, the use of BM analysis has increased significantly
since the mid-nineties under the pressure of technological innovation and
the expansion of traditional competitive areas through virtual networks.
The BM concept has been examined in depth in different fields of
economic literature, such as Information and Communication
134 S. Cosma et al.
Technologies (ICTs) and e-business management, organisation theory

and strategy studies. This variety of approaches implies that, even now,
there is not a widely accepted definition of BM and there are many
meanings of the keywords “business model” (Klang et al. 2014).
7.2.1 ICT and e-Business Stream
In the last two decades, the BM has become a focal concept for
researchers in the e-business stream; they consider it as a way of analysing
competitive behaviour and explaining firms’ performance in competitive
environments characterised both by intensive use of ICTs in production
and distribution processes and by a rise in the importance of stakeholder
networks (suppliers, partners, customers, etc.) in value creation.
This literature has developed a BM concept that aims to embrace all the
elements and relationships that enable IT-based or Internet-based firms to
generate value. It follows the idea that the system (a sort of gestalt) creates
more value than the sum of its individual parts and the BM is essential to
enhance it (Amit and Zott 2001; Zott et al. 2011). Therefore, according
to these analyses, a BM is interpreted as a representation of the set of
decisions, activities and relationships between them that explain how an
organisation creates, delivers (to its stakeholders, including customers)
and captures value (Osterwalder and Pigneur 2012), building a sustain-
able competitive advantage in specific markets (Morris et al. 2005).
Often these contributions tend to give greater emphasis to specific
components of this systemic representation. Some stress how firms generate
value, i.e. the value proposition (Baden-Fuller and Haefliger 2013), or how
they optimise the cost/revenue structure (i.e. value capture); others focus on
the way in which relationships with the enterprise’s network (suppliers,
customers, delivery channels, partners, competitors) increase value.
Notwithstanding the different focus on BM components, there is
consensus on the idea that BM has value at the corporate level and offers
a useful holistic perspective for understanding not only what businesses
firms do (e.g. what products and services they produce to serve the needs
of customers in addressable market spaces) but also how they do it (e.g.
how they bridge resources and product markets in serving the needs of
customers). So that BM becomes a new unit of analysis which puts

emphasis on a firm’s activity system to create value as well as to capture it
(Zott et al. 2011).
7.2.2 Organisational Theories and Resource-Based

View
In organisation theory, the BM concept has been developed as an inte-

grated presentation of the company, in order to contribute to process
management and process-based organisation design. Not much has been
published from the organisational point of view, and the role of this
approach is marginal in comparison with the management or strategy
streams. In these works, the BM concept focuses to a greater extent on
the business’s operational aspects and value creation within the firm. In
this stream of the literature, the BM generally refers to the firm overall
and its structural components (Wirtz 2011). It has been seen as a tool for
the abstraction of an entire company and its architecture.
Various works within the e-business stream concentrate on the com-
ponents of a systemic representation of the firm, with a resource-based
view (RBV) approach, considering the firm as a combination of resources
and competences. A RBV approach is also used by Osterwalder (2004) in
his BM ontology, in order to represent the knowledge and exchanges of
both tangible and intangible resources (competences, services, etc.)
between organisational players to identify the ways in which the firm
generates value (Chen et al. 2014). However, resources only represent a
firm’s potential. They are necessary but not sufficient for success.
Competitive advantage derives from the ability of the company in its
entirety to activate, coordinate and integrate resources to achieve per-
formances better than those of its competitors (Penrose 1959).
Although useful for representing a firm’s organisation, the RBV is
insufficient to describe the ability to generate value without the aid of a
BM. Resources in themselves do not generate value for the customer;
value is generated by relationships and transactions with the customer
based on them. DaSilva and Trkman (2014) define the BM as a repre-
sentation of “a specific combination of resources” which through
136 S. Cosma et al.
transactions generate value for both customers and the organisation,

combining the RBV with transaction cost economics. The
resource-based approach runs through many works, but its theoretical
rationale has difficulty in understanding how a BM is able to create value
in a way that differs from other management literature concepts (DaSilva
and Trkman 2014). One BM theory, in which a firm’s organisational
outcomes are affected by managerial competences, expertise, ability to
learn and execution, is very appealing. Nevertheless, in the literature,
there is no agreement on the nature of the individual components and
this approach does not clearly explain the contribution of internal
resources and cannot be precisely distinguished from managerial litera-
ture (George and Bock 2011).
7.2.3 Strategy Stream
Traditionally, the strategy stream approach relates a company’s creation

of value to its ability to identify a strategy (or combination of strategies)
able to provide a competitive advantage.
Taking the competitive environment as a reference, strategy is the
creation of a unique, valuable position for the firm through the devel-
opment of a competitive advantage. Strategy involves a different set of
activities and the choice of the specific way in which the firm competes
(Porter 1996; Teece 2010). It is useful to analyse the concept of strategy
on two different levels.
At the first level, there is a corporate strategy (Ansoff 1965), which
concerns the choice of the size and diversification of the company’s
business portfolio. This level involves the definition of the various
strategic business areas (SBAs) which reflect the product–market–tech-
nology combination chosen.
Corporate strategy refers to the firm’s high-order (first level) long-term
choices in terms of diversification, vertical integration, internationalisa-
tion, growth (acquisitions and new ventures), size, governance structures,
capital allocation to the different SBAs and disinvestment. It answers the
question “Where are we going to compete?”
At the second level, there is the business strategy, which identifies how
to compete and how to achieve competitive advantage in each SBA. The
business strategy’s specific objective is the sustainable competitive
advantage a company can achieve and the relationship between that
advantage and the industry, by which it is defended. It answers the
question “How are we going to compete?”
The SBA identifies the actual (physical or virtual) situation where the
business strategy is applied. It is a combination of customers, products
and internal resources, where the specific internal relations need a par-
ticular entrepreneurial and strategic approach.
Within the “strategy stream”, two different meanings of BM can be
identified.
On the one hand, in a more operational perspective, the BM is a way
of representing and analysing (and, not least, validating, through analysis
of internal consistency) the value generated by the strategies, and on the
other hand, it is a way of maximising value through the best operational
structure and articulation (Shafer et al. 2005). Business strategies focus
on the market and external competition, while the BM concentrates on
the ability to optimise the business internally and/or within the network.
It defines the activities through which the objectives defined by the
business strategies can be pursued to improve or optimise competitive
advantage (Mottura 2011).
An alternative systemic perspective positions the BM closer to strategy.
It can be defined as the concrete choices that derive from the actual
combination of corporate/business strategies and the various activities
and economic levers involved (price variables, control of costs, customer
segments, quality, distribution channels, degree of relationship, tech-
nology, productive processes, etc.).
This second approach provides the rationale for connecting the key
company strategic choices to their main consequences. BM is the rep-
resentation of a subset of key choices implemented and their main
consequences. Choosing a specific BM (policies, assets and governance)
means choosing a specific way of operating and creating and capturing
value for the firm’s stakeholders. Strategic choices establish the BM.
Therefore, the BM is not the strategy, but is the direct result of the
strategy a firm implements (Casadesus-Masanell and Ricart 2010).
138 S. Cosma et al.
7.2.4 Strategic Groups
According to the strategy stream literature, one way to analyse the firm
within a specific industry is to compare its performance with that of its
main competitors, in order to identify any strategic groups. The literature
tells us that a strategic group is made up of firms which follow the same
or similar behaviour along key strategic dimensions, with regard to
specific criteria such as product range, size, internationalisation, tech-
nology and vertical integration (Porter 1996).
The group of firms builds mobility barriers on these key strategic
dimensions and the investments they require to separate the individual
group from the external competition. These barriers help to explain
firms’ competitive advantage, their performance and the effects of the
external context on their profitability.
Recent empirical literature has led to the identification of the strategic
group with the BM, as in the large number of studies which employ
cluster analysis to group similar strategic behaviours and performance.
The empirical results of these analyses are controversial, and their
methodology is not always considered effective for analysing differences
in firms’ performance within the strategic group (Short et al. 2007; Leask
2004).
The employment of cluster analysis has usually focused on economic
variables which only provide a very indirect, sketchy picture of actual
strategies, and therefore, the attempt to classify firms’ strategies on the
basis of very broad differences between firms is able to identify general
but not specific strategies.
At the same time, the finer the classification criteria used, the more
homogeneous but also the narrower the groups become in terms of
profitability and strategic behaviour. Moreover, when the analysis deals
with strategies as context specific, as specific actual choices bounded by
available resources, competitive position, management attitude to risk
and industry structure, the strategic group BM is replaced by the indi-
vidual firm.
In Fig. 7.1, we classify the different approaches in management lit-
erature with respect to BM definitions, in order to assess their usefulness
Fig. 7.1 BM definition: the different approaches in the management literature
in relation to the firm’s operations. They are arranged on two axes: we

place the operational/strategic level on the horizontal axis and the firm’s
level of organisation, from industrial sector through to individual lines of
business, on the vertical axis.
7.3 The BBM Literature

Following the broad conceptual perspective of strategic management
studies and its adaptation to bank strategic studies, we review and classify
the BBM literature, taking as reference a definition which considers the
set of both bank portfolio choices and management abilities intended to
exploit market positioning in the different business areas.
Portfolio choices are indicated by the different mix of SBAs, which
reflects how “first level long-term strategies” (like growth, diversification,
internationalisation and attitude to risk) translate into organisational
features of different combinations of products/customers/resources.
These strategies, also referred to as corporate strategies, reflect the firm’s
history, its experience and capabilities and the competitive context.
140 S. Cosma et al.
Long-term strategies—stable in a stable economic context—were

stressed in the pre-crisis era when banks set their strategic plans against
the prospect of continuous growth in the demand for financial services.
Then, after the crisis burst, retrenchment and a consequent deep revision
of growth, diversification and internationalisation strategies prevailed,
along with a strengthening of capital positions. This was followed by
significant changes in the business mix, which affected the composition
of banks’ assets and liabilities as well as their earnings.
At a second strategic level, we set business strategies, the managerial
choices that pursue revenue enhancement through customer segmenta-
tion and product differentiation, cost efficiency and risk management in
the different business areas, capturing value from the business mix
adopted. How these strategies are successfully implemented will depend
on the specific managerial abilities of the bank’s organisation.
Both long-term and business strategies are affected by macroeconomic,
competitive and regulatory variables. These context variables have ex ante
effects on banks’ strategies and their BMs in so far as they influence the
hypothesis the strategic plans are based on. Ex post they directly affect the
way in which business strategies achieve the targeted results.
In Fig. 7.2, we show the different strategic levels through which we
represent the concept of the BM. This schema enables us to classify the
main contributions to BBM literature and underline how the different
approaches focus on a different identification of strategic variables.
An initial classification of the literature is based on the different
emphasis placed on the identification of bank peer groups sharing similar
BMs.
Several studies follow the literature on strategic groups and aim to find
evidence of how the banking industry can be classified using just a few
different bank BMs with different performance with respect to economic
and financial context.
The main characteristic of these studies is the distinction between
what the bank is doing and how the bank is doing it, so that business
composition is identified with the BM concept, whereas other strategic
variables are considered as the outcome of portfolio choices. Under our
Fig. 7.2 Strategic components of banking business model (BBM)
schema, this approach implies a representation of the BM in terms of

SBAs, which are approximated by asset/liabilities and/or income com-
position, whereas business strategies are implicitly evaluated in the out-
come analysis. It is important to underline that this definition of BM can
lead to the attribution of performance results to the business mix even
when the former are due to the bank’s skill in managing the single
business areas.
This approach can be traced in the work of Ayadi and Groen (2014)
and Ayadi et al. (2016a). Following the pioneering work of Passmore
(1985) and Amel and Rhoades (1988), these authors employ a two-stage
procedure for the BM analysis of European banks. The first step adopts
cluster analysis to group banks on the basis of asset and liability com-
position. Then, they evaluate how the bank clusters perform with respect
to a very broad set of indicators concerning both performance results and
strategic behaviours such as risk exposure, loan growth and interna-
tionalisation. Six indicators of asset and funding composition used in
cluster analysis identify four large, distinct groups that differ from each
other in their retail and financial market orientations. A comparison
between bank clusters on the basis of risk-return frontier confirms that
investment banks have higher risk and volatility, while diversified retail
banks seemed to perform better during the financial crisis thanks to their
higher revenue stability.
142 S. Cosma et al.
Roengpitya et al. (2014) also use the cluster algorithm, along with the
adoption of some selection criteria and balance sheet ratios, to classify the
BMs of a large sample of listed and non-listed banks from 34 countries,
during 2005–2013. They compare three distinct bank clusters with
respect to outcome variables. When valued in terms of performance, the
retail-funded group (high share of loans on total assets and high reliance
on deposits) displays the highest average level and the lowest variability of
profitability over time. The trading banks (half of the assets in tradable
securities and predominantly funded in the wholesale market) are the
group with the highest volatility of return on equity and cost base. The
wholesale-funded group stands between the other two groups in terms of
return levels and volatility. The study finds significant shifts across dif-
ferent BMs before and after the crisis: two-fifths of the banks classified as
wholesale funded or trading in 2007 ended up with a retail-funded BM
in 2013. The performance statistics show that the change in banks’ BM
induced a prevailing worsening in profitability.
A different result with respect to migration between bank clusters over
time is shown by a study presented in the ECB Financial Stability Review
(2016) in the classification of European banks during 2007–2014.
Comparing the bank clusters based on size, asset/liability and income
composition indicators, they find that most banks remained in the same
group, revealing “sticky” BMs which have difficulty in adapting to a
changing environment or the anticipation of stress.
De Meo et al. (2016) adopt an original fuzzy clustering technique
based on a broad set of asset/liability mix indicators of listed and
non-listed European banks (77 for 15 countries) for 2006–2014. They
identify three main clusters of banks: retail, diversified and investment
banks. Each group was then subdivided on the basis of four EBA clas-
sification criteria (systemic relevance, dimension, organisational com-
plexity and cross-border activity) considered by the authors to be
attributes of strategic choices. Among the eight resultant peer groups,
retail banks show the highest return on assets in the years preceding the
financial crisis but the worst performances at the peak of the sovereign
debt crisis, due to the deterioration in credit quality. Among them, small
banks with limited cross-border exposure and a low degree of income
diversification (non-complex retail banks) were hardest hit by the
increasing credit risk. The study analyses the effects of macroeconomic
variables on the performance of the different peer groups: as expected,
economic growth, yield curve and sovereign risk are the most significant
variables affecting retail banks, whereas due to their dependence on
non-interest income, investment and diversified banks displayed a sig-
nificant exposure to financial markets. One methodological aspect of the
analysis must be underlined: probabilistic clustering tends to make the
performances of different BBM more similar, a sign that bank-specific
strategies may be more important than membership of a strategic group
in explaining bank performance.
The specificity of bank strategies is the focus of Mergaert and Vennet’s
(2016) analysis. They define BM in terms of the strategic variables that
reflect the management’s long-term choices (latent strategies) with regard
to asset and liability composition, capitalisation, income structure and
the bank’s risk profile. The common variances of these variables define
two broad BBM: retail and diversified. The authors underline the fact
that these models are graduated and use common factor analysis to
evaluate both how these long-term strategies are implemented and their
impact on performance. The authors conclude that there is a substantial
variation in the effects of the BM between different bank types and show
that retail-oriented banks perform better in terms of both profitability
and stability and that diversification improves profitability, but also
increases the likelihood of distress.
A different approach to the grouping of European banks is employed
by Bonaccorsi et al. (2016). They classify 112 significant European banks
following a step procedure based on threshold values of balance sheet
parameters including size, lending propensity and international credit
exposure. The authors use data published by the European Banking
Authority (EBA) and the European Central Bank (ECB) further to the
comprehensive assessment, which allows them to define portfolio com-
position by counterpart type, showing that large domestic and other
lending banks are more exposed towards SMEs and retail real estate
144 S. Cosma et al.
secured loans than large international or diversified banks. This com-

position explains the higher level of profitability of lending banks but also
their greater cyclical sensitivity. The study points to macroeconomic
conditions as the main driver of current differences in profitability across
bank types, whereas riskiness seems to reflect both differences in bor-
rowers’ risk profiles and the extent to which banks use IRB models. In
particular, the ratio between risk-weighted and unweighted exposure
(risk density) is lower for large banks able to both tailor riskiness to each
individual position more effectively and, in some cases, manipulate risk
weights, thus creating a bias towards lower risk density.
How the management of risk weights is linked to banks’ chosen BM is
the theme of the study by Ayadi et al. (2016b). Applying the Ayadi et al.
(2016a) cluster approach and using the same group classification, the
authors provide evidence of the different degree of regulatory arbitrage
across bank BMs. Notably, IRB adoption seems to have a positive effect
on the riskiness of retail diversified banks, signalling that regulatory
arbitrage is occurring within this banking BM.
An alternative strand of BBM literature adopts a wide definition of
BM that combines portfolio choices with many other business and
context variables. According to our schema, this approach has the merit
of considering many strategic aspects of a BM, although these studies
often fail to make a clear distinction between long-term strategies,
business mix and business strategies.
A second feature of these studies is their emphasis on banks’ different
strategic behaviours rather than the identification of strategic groups. In
some cases, BM variables are compared across the main institutional
bank groups or considering different bank sizes.
A further characteristic of this approach is the focus on bank riskiness
and the identification of which BM variables most affect bank vulnera-
bility. The focus on risk reflects the perspective of bank supervisors and
their concern for the consequences of bank strategies on default events.
This approach is central to the work of Altunbas et al. (2011), who use
a broad set of bank characteristics to identify BMs. Three risk measures
of a large sample of European and US banks are regressed on groups of
indicators collected in the pre-crisis period. These should denote different
banks’ BMs: asset, funding and income composition variables along with
indicators like loan growth, capital ratio, total assets and a number of
variables that account for major macroeconomic and institutional factors.
Therefore, their definition of BM includes business mix variables along
with some other strategic variables concerning growth, capitalisation and
size. These strategic variables, along with the reliance on short-term
market funding, are statistically significant in explaining bank distress.
The main indication concerns the significant, high impact on banks’ risk
of the aggressive expansion in loan growth in the pre-crisis years, as
evidence of the relaxation of credit standards and a deterioration in asset
quality. In addition, the ratio of loans to total assets is positively related
to bank risk as well as bank size. With regard to funding and income
composition, the study finds evidence that relying on deposit funding
reduces the probability of a bank rescue, whereas non-interest income
reduces the likelihood of distress during the crisis. Conversely, the use of
wholesale funding increases the bank’s risk.
Köhler (2014) follows a similar approach, relating Z-logscore to
business mix and loan growth variables for a large sample of European
banks. The analysis evaluates the relationship to the main institutional
bank categories: commercial, saving and cooperative, and investment
banks, with a focus on listed banks. In Köhler (2015), the same risk
indicator is regressed on two main business mix variables (non-interest
income share and non-deposit funding as a fraction of total assets) and
then integrated with many other control variables. The approach is
similar to that adopted by Demirguc-Kunt and Huizinga (2010) for an
international sample of 1334 banks in 101 countries leading up to the
2008 financial crisis. The econometric study by Köhler confirms some
results found in the bank diversification literature, which point to the risk
of shifting a bank’s operations onto the financial markets (securities and
wholesale fund market). For savings and cooperative banks, a larger share
of their income from non-traditional activities generates more return
stability, but the banks themselves become less stable due to the increase
in their share of non-deposit funding. This contrasts with investment
banks, which become riskier when they increase their non-interest
income and will be significantly more stable if their share of non-deposit
funding rises. This may be because retail and investment banks diversify
146 S. Cosma et al.
in different ways. The latter derive most of their non-interest income

from securities-related activities that incorporate a market risk, whereas
the former earn their diversification revenue mainly from banking-related
services. This signals the importance of keeping these two fundamentally
different types of activities separate when studying the relationship
between bank risk and diversification (DeYoung and Torna 2013; Brighi
and Venturelli 2014). Along with diversification, lending growth is also
an important determinant of bank risk that significantly differs across
countries, due to both the different aggregate credit growth and the
reduction in bank lending standards and collateral requirements during
booms.
A ECB study (2016) of a sample of 143 Euro area banking groups
during 1995–2014 also regresses the z-score variable on several
bank-specific BM characteristics, including some business mix measures
(including retail ratio, income diversification and short-term borrowing),
cost-to-income ratio, a leverage ratio and size. Other explanatory vari-
ables regard macroeconomic conditions and structural market features.
Pre-crisis, income diversification is associated with higher default risk,
whereas during and after the onset of the financial crisis more diversified
banks displayed lower default risk levels. During the whole period, a
higher default risk for global systemically important financial institutions
(G-SIBs) contrasted with an overall reduction in riskiness for smaller, less
complex banks. This result is in line with Köhler’s findings that diver-
sification is beneficial up to a point, beyond which banking group
complexity is prejudicial to bank stability. Increasing recourse to
short-term borrowing also has significant riskiness implications for
G-SIBs, whereas if they shift their funding mix towards deposits, they are
able to reduce their risk exposure.
The bulk of these studies focuses on the nexus between BM variables
and individual bank risk, and only a few of them deal with the effects of
strategic choices on systemic risk.
An analysis of the nexus between BM variables and measures of
individual and systemic bank risk based on market values is proposed by
Van Oordt and Zhou (2014), who rely on stock market data from CRSP
of US Bank Holding Companies from 1991 to 2011. Drawing on the
literature on market risk, the authors identify two aspects of banks’
systemic risk: bank tail risk and the linkage between a bank’s tail risk and
severe shocks in the financial system. As expected, they find a stronger
dependence between large banks and systemic risk, with a positive
association between size and sensitivity to severe shocks in the financial
system (approximated by severe changes in the financial sector index).
The same positive relationship with severe financial shocks is found for
non-interest income share, confirming that banks’ involvement in these
activities is relevant not only for microprudential but also for macro-
prudential regulation. With regard to asset/liability composition, the
study points out that lending-focused BMs are significantly associated
with higher levels of tail risk, but with lower systemic linkage. For the
deposits-to-assets ratio, they find similar results on the relationship to
financial shocks. Growth strategies are associated with an increase in
sensitivity to large shocks in the financial system, whereas banks with
higher capital ratios show a significantly lower exposure to systemic risk.
As already underlined, the analysis of the interrelations between sys-
temic risk and some main bank characteristics has considerable impli-
cations for regulation: the breakdown of systemic risk clearly indicates
that regulators must choose the right balance between micro- and
macroprudential objectives.
7.4 The Literature on Bank Diversification:

The Nexus with BBM Analysis
The review provided above demonstrates that the literature on bank BMs
is closely linked to that on diversification strategies. The link clearly
emerges from the empirical analysis centred on the nexus between
diversification activities and measures of banks’ performance, with the
former usually approximated by asset/liability and income composition
indicators and in particular by the distinction between net interest
income and non-interest components. These measures thus highlight the
scope of corporate strategy, or in other terms, the results of the strategic
portfolio banks decide to develop. At the same time, the so-called control
variables (i.e. size, economic efficiency and risk profile) used in
148 S. Cosma et al.
diversification studies can be considered as a proxy for bank choices at

the business strategy level.
If the focus is on the main commercial banks, it is reasonable to
assume that the differences in their BMs concern mainly the different
intensity with which the functional diversification1 process, in terms of
an array of products and services and customer segments, has been car-
ried out. In recent decades, the development of financial markets and the
increasing complementarity between the banking and securities segments
of financial intermediation have contributed to the characterisation of
banks’ BMs: the securitisation process is emblematic of this change. In
many countries, the development of the asset management business has
been favoured by banks’ diversification strategies. Banks are also the main
investors in bonds and in particular have continued to play an important
role in the coverage of sovereign debt.
For these banks, decisions about their BM and the competitive
advantages that may result are interwoven with key strategic decisions
concerning size/growth and diversification.
The goal of achieving optimum size and exploiting economies of scope
in the offering of a wider range of products and services was central to the
strategies of many banks, at least until the outbreak of the financial crisis.
In retrospect, it is easy to see that this approach was based on an over-
estimation of the prospects for growth in the demand for banking
products and services, and a clear underestimation of the operational
complexity and risk profile related to larger size and wider SBAs.
From a theoretical point of view, the existing banking literature
focuses on the question “should banks diversify their portfolios or should
they specialise?” since both pros and cons can be identified. Among the
recognised benefits, the possibility of exploiting economies of scope may
lead to an increase in performance through cost savings or revenue
improvements (Teece 1982; Herring and Santomero 1990; Llewellyn
1996; Klein and Saidenberg 1997; Campa and Kedia 2002; Elsas et al.
2010), along with a reduction in the degree of information asymmetry
(Diamond 1984, 1991; Rajan 1992; Stein 2002) and the agency costs of
managerial discretion (Stulz 1990; Stein 1997; Gertner et al. 1994).
These benefits have to be traded off against the costs associated with
diversification. In particular, increasing the size and scope of a bank’s

activities introduces the “cost of complexity”, which at some point may
outweigh the benefits that can be achieved (Rajan et al. 2000; Graham
et al. 2002). Moreover, diversified institutions can suffer (DeYoung and
Roland 2001) from earnings volatility, lower switching costs for clients
and higher operational and financial leverage (Demsetz and Strahan
1997; DeYoung and Roland 2001), increasing the volatility of earnings
and hampering risk-adjusted performance measures.
While the theoretical literature has effectively addressed the reasons for
and economic effects of greater diversification of business, empirical
studies only estimate the implications of functional diversification at a
general level, by testing the nexus between some aggregated indicators of
business mix and measures of banks performance.
Most studies are centred on the US banking industry, following the
implementation of Gramm Leach Bliley in 1999. With few exceptions,2
these contributions find that a shift towards non-interest activities
worsens the risk-return trade-off because the costs of diversification
outweigh the benefits, mainly due to the increased volatility of these
activities (DeYoung and Roland 2001; Stiroh 2004; Stiroh and Rumble
2006; Laeven and Levine 2007; Goddard et al. 2008); moreover, this
finding is valid for both financial holding companies and smaller insti-
tutions such as credit unions.
Fewer studies deal with European banks and those which are available
provide similar results regarding the effect of diversification on bank
performance. Among them, Mercieca et al. (2007), examining a sample
of 755 small European banks for the period 1997–2003, find that small
European banks do not gain from their diversification strategy because
the higher volatility of net interest income outweighs the benefits of
diversification, implying lower risk-adjusted returns, and this is linked to
small banks’ lack of expertise in managing new lines of business. Lepetit
et al. (2008) find that for a set of European banks from 14 countries
during 1996–2012, expansion into non-interest income-generating
activities displays higher risk and higher insolvency, and this is particu-
larly true for smaller banks and those driven by commission and fee
activities. Baele et al. (2007), using a sample of listed banks from 17
European countries during 1989–2004, confirm Stiroh’s finding (2006)
150 S. Cosma et al.
that banks that rely more on non-interest sources of income have sys-
tematically higher market betas and hence higher systematic risk.
These findings may be affected on the one hand by measurement
problems linked to the definition of diversification used and on the other
hand by the lack of consideration of the possible interaction between
diversification and banks’ other characteristics.
In this sense, the degree of information granularity disclosed by banks
in relation to the nature of fee-based revenues allows a more precise
evaluation of the nexus between diversification and performance and can
affect the final results.
Gallo et al. (1996) showed the importance of distinguishing between
the different components of non-interest income. In particular, com-
bining bank and mutual fund activities improved the profitability and
reduced the risk of US bank holding companies during 1987–1994.
More recently, DeYoung and Rice (2004), DeYoung and Torna
(2013) recognise that different fee-generating activities show different
production and risk-return characteristics and hence are likely to have
different impacts on the probability of financial distress and insolvency.
The authors identify three categories of non-interest income, and the
results point out that higher involvement in asset-based non-traditional
activities such as venture capital, investment banking and asset securiti-
sation is associated with higher probability of failure for financially dis-
tressed US banks and that an increase in pure fee-based non-traditional
activities such as securities brokerage and insurance sales reduced the
probability that banks would fail during the crisis.
The recent studies on the diversification of Italian banks benefit from
detailed, public data on bank income composition. Cotugno and
Stefanelli (2012) use a panel data set comprising 4038 observations
relative to Italian banks for 2005–2010 and find a positive relationship
between product diversification and bank performance, also in terms of
risk-adjusted measures. On a sample of 145 Italian banks during 2006–
2008, Vallascas et al. (2012) reveals that institutions that were diversified
within narrow activity classes before the financial crisis experienced large
declines in performance during it. By contrast, diversification across
broad activity classes, such as lending and capital market activities, did
not cause performance losses during the crisis. Brighi and Venturelli
(2016) use bank-level data on 491 Italian banks during 2006–2012 to

investigate the impact of functional and geographical diversification on
bank performance during the 2008 financial and the 2010 sovereign debt
crises. Both crises negatively affected bank profitability, but banks that
were more diversified, in terms of both revenue and geographical
diversification, were less penalised in terms of risk-adjusted profitability.
Results differ for the sample of mutual and non-mutual banks, with the
former benefiting more from geographical and the latter from functional
diversification.
The importance of the degree of information disclosure is captured
well in a recent study by Williams (2016), which models the relationship
between bank revenue composition and bank risk using data drawn from
the confidential regulatory returns of Australian banks. At first glance,
consistently with previous international evidence, it is seen that banks
with lower levels of non-interest income as a proportion of total bank
revenue and higher revenue concentration are less risky, but at the same
time, some types of non-interest income are risk reducing when the
effects of bank specialisation are considered. To study this in greater
detail, bank revenue is broken into six categories, and the results
underline the existence of some portfolio diversification benefits from
trading and investment income.
Turning to the theory that findings relating to diversification may be
influenced by a failure to consider interactions between diversification
and banks’ other long-term choices, the possible effect of interaction
between size and diversification is accurately described in De Jonghe
et al. (2015). Examining a panel of 16,507 bank-year observations,
distributed over 15 years and 76 countries, the authors identify a neg-
ative interaction between size and non-interest income in their rela-
tionship with systemic risk. In other terms, non-interest income reduces
large banks’ systemic risk exposures, whereas it increases those of small
banks. In particular, small banks are more likely to lack the expertise
needed to handle a wide array of products and services or manage
complex financial products. Moreover, they are not subject to in-depth
external scrutiny, so they may be more inclined to engage in riskier
activities; on the other hand, larger banks are typically subject to more
external scrutiny, which may discourage excessive risk taking, and they
152 S. Cosma et al.
can count on more sophisticated risk management techniques and a

more experienced management team. So the concepts of size and scope
should not be analysed in isolation, since they are strictly interrelated.
Summing up, the results of diversification studies are strongly influ-
enced by the consideration of the nature of non-interest income and the
simultaneous interaction with banks’ long-term choices. As a conse-
quence, BMA cannot limit its scope to the same metrics used in the
diversification literature, since from the heterogeneity of the results it is
clear that there is an optimal mix between size, risk and revenue diver-
sification that calls for an integrated approach extended to the analysis of
diversification, which is just one component of BMA. These findings also
influence the measurement aspect of BMA; from a methodological point
of view, BMA requires the implementation of techniques that enable the
simultaneous consideration of the different dimensions involved:
long-term strategies, business mix and business strategies.
7.5 Banks’ Key Vulnerabilities

in the Supervisory Assessment Scheme
The analysis of the recent developments in BBM literature suggests
interesting key points on the conceptual framework adopted by super-
visory authorities: how they are approaching this theme and to what
extent they share the perspectives emerging from the studies discussed
above.
To this end, it is first of all interesting to trace the birth and evolution
of BMA as a proactive supervisory instrument.
The increasing interest of supervisory authorities in BMA stems from
the crisis (2007–2008). BMA was pioneered in the UK after the
Northern Rock crisis and the failure of the “light supervision” applied by
the Financial Services Authorities (FSA). The Turner Report (FSA 2009)
named the FSA’s supervisory approach, based on the idea that BM risks
were better assessed and balanced with returns by top management and
boards of directors (BoD) than by bank regulators and supervisors, as one
of the causes of the crisis. “Light supervision” was focused mainly on the
operation of appropriate systems and controls within the supervised

institutions. Changes in the supervisory philosophy were put in place
from 2009 onward, with the introduction of a more intrusive, systemic
revised approach to be implemented by a new authority: the Prudential
Regulation Authority (PRA), operative since 2013. The revised super-
visory approach is based on an “Intensive Supervision” model with the
pendulum shifting from trust in market discipline, with supervisory
intervention mainly after something had gone wrong, to a proactive
regulatory and supervisory action, with a forward-looking perspective
(Moloney 2012).
Within the new approach, PRA (Bank of England FSA 2012) gives an
important role to BMA as a proactive, forward-looking instrument with
two main aims: from an idiosyncratic point of view, to examine the
threats to the viability of a bank’s BM and its key vulnerabilities; from a
systemic point of view, to identify possible adverse effects on other
participants in the system from the way in which the institution conducts
its business. The key aspects for identification of a bank’s vulnerabilities
are an assessment of its sources of revenues, the related risks and funding,
and the analysis of its strategy and the business plan. The second step is
peer analysis to identify each bank’s position within its strategic group
and evaluate any outlier BMs and management practices, and their
contribution to systemic risk.
BMA is now embedded in the SREP, Pillar II of the Basel Capital
Accord, and is intended to reveal a bank’s key vulnerabilities in the short
run and the viability and sustainability of its strategic plans in the short
and medium term. The aim of BMA is to assess not only each bank’s
risks and therefore its vulnerability, meaning its idiosyncratic risk in a
microprudential perspective, but also its contribution to systemic risk, in
a macroprudential perspective. Within this framework, BMA was
introduced by the EBA (2014), as the first of four key elements, followed
by the assessment of internal governance and institution-wide control
arrangements, risks to capital and adequacy of capital to cover these risks,
and risks to liquidity and adequacy of liquidity resources to cover these
risks.
The ECB also identified BMA as a key area of the supervisory activity
of the SSM in its Thematic Review in 2015 (ECB Banking Supervision
154 S. Cosma et al.
2015) and in greater detail in 2016. Under EBA Guidelines (EBA 2014;
ECB Banking Supervision 2016), the elements of BM analysis are:
identification of banks’ main activities; assessment of the business envi-
ronment; analysis of the forward-looking strategy and financial plans;
assessment of the BM’s viability (within one year), sustainability (within
three years) and sustainability over the cycle (more than three years); and
assessment of key vulnerabilities (Lautenschlager 2016; ECB Banking
Supervision 2016). Through this analysis, the supervisors aim to
understand the implications of BM characteristics for banks’ overall
riskiness. The peer analysis follows.
The SSM approach is based on both quantitative and qualitative
analysis and should incorporate a forward-looking perspective, linked to
financial planning, business plan analysis and macroeconomic and
market trends. The scheme of analysis identified by the EBA is quite
exhaustive, and different aspects are considered when focusing on the
BM adopted by each bank, with the aim of revealing its viability, sus-
tainability and key vulnerabilities due to risk assumptions. The specific
levels of granularity of information on different aspects, product/business
lines, breakdown of income and cost streams, impairment provisions and
key ratios required by the SSM are not disclosed; the criteria for the
definition of peer groups, and the banks included, are also not officially
disclosed. The BMA of supervisory authorities is currently being devel-
oped from basic to more sophisticated analysis.
Interesting points for consideration emerge from the scheme of anal-
ysis provided by the EBA guidelines (2014), concerning the theoretical
and methodological framework underlying the work of the supervisory
authorities, also in relation to the main findings of the theoretical and
empirical literature on BMA, set out above.
One initial comment relates to the methodological approach of the
SREP, where BMA precedes and supports the three subsequent analysis
stages, assessment of governance, ICAAP and ILLAP. BMA (EBA 2014)
is intended to pinpoint the determinants of BMs and the adequacy of
their returns over time, while the other three areas analyse risks, risk
management models and risk governance. In fact, the assessment of these
areas should aid in the overall assessment of the viability of the current
BM and the sustainability of the strategies, the main objective of BMA,
which, therefore, should combine with rather than complementing an

evaluation that investigates the strong and weak points of the BM
adopted at the various strategic levels. Thus, a holistic approach is needed
in BM evaluation, linking risk analysis to the main strategic areas, both
corporate, where the guidelines for the types and amounts of risk to be
taken are decided, and at the business strategy level, with regard to risk
management in the various business areas. Above all, it is essential to
maintain a close connection between business mix decisions, the allo-
cation of resources and risk management. A multidimensional, transverse
approach is the way to strengthen the final synthetic evaluation, with
banks classified into four groups and an additional class for “failing or
likely to fail” institutions, for which specific supervisory measures must be
defined for each bank.
Another point of reflection stems from an assessment of the BM’s
sustainability and viability through two levels of analysis: “corporate/first
level long-run strategies” and “business strategic level”, related to man-
agerial choices concerning revenue enhancement policies, cost efficiency
objectives and risk management processes. The first level of analysis
should consider a time span long enough to take into account changes in
the economic cycle, which are intertwined with and determine strategic
corporate choices, such as growth, internationalisation and diversification
before the crisis, and deleveraging, capital saving and different sources of
funding after the crisis. This approach could help to strengthen a
forward-looking approach, preventing the static assessment of corporate
strategies in terms of profitability and risk assumption. One example is
the different negative impact of the two phases of the crisis on bank
performance in Europe: in 2008–2009 notably higher for wholesale and
investment than for retail banks; the opposite after 2012 with lower
profitability for retail banks mainly serving the SME segment, worst
affected by the economic recession. Moreover, this latter effect was more
serious in peripheral countries than in core countries with better eco-
nomic trends. This last finding is also relevant for country-specific factors
that should be taken into account when explaining lower bank prof-
itability. Since macroeconomic conditions seem to be the main driver of
current differences in profitability across the country bank binomial
156 S. Cosma et al.
(Bonaccorsi et al. 2016), a key question arises concerning the effective-

ness of peer analysis for taking these factors into account.
A further reflection arising from the literature review concerns the
different behaviour of banks in relation to their size and diversification
and the different analysis required by regulators. This distinction is evi-
dent from the BMA developed by the ECB for significant banks and that
adopted by the National Competent Authorities (NCAs) for less sig-
nificant banks. In the first case, a high degree of granularity of infor-
mation at the second strategic level is a key element in verifying the
sustainability of corporate strategies, especially for systemic banks dis-
playing high complexity in the BMA conceptual framework. Peer anal-
ysis must carefully consider banks’ different dimensions in terms of
complexity and diversification: the granularity required is a strong
argument in favour of one bank-one BM. For these banks, given the
variety of their strategic choices, a rough definition of clusters, based on
too few overly generic variables, could lead to a misallocation of banks
and/or an inadequate framework for their evaluation (Gualandri 2016).
On the other hand, a less complex BMA approach is envisaged by the
NCAs, with lower granularity, for the large amount of small banks with a
higher degree of homogeneity of strategies and asset mix.
Another consideration is that capital adequacy is a key element con-
sidered by supervisory authorities, bank management and also the market
from slightly different but complementary perspectives. Supervisors
evaluate capital strength for its effect on idiosyncratic and systemic risks,
managers as a basis for strategic decisions such as growth and diversifi-
cation. In the new regulatory context, the level of capital and its allocation
become a more long-term strategic variable, strictly interrelated with
corporate strategies. As a consequence, from the supervisory point of view
the forward-looking perspective adopted in BMA should be reinforced
with a further analysis aimed at evaluating the market’s capacity to absorb
capital-intensive strategies (Calomiris and Nissim 2014). Analysis based
on market value measures may be used to highlight the nexus and dis-
tinction between BM variables and measures of individual and systemic
bank risk. In particular, recent research has identified a stronger rela-
tionship between large banks and systemic risk, with a positive association
between size and sensitivity to severe shocks in the financial system.
The capital, corporate strategy nexus clearly reveals how regulation

may influence strategic choices, especially those regarding portfolio mix
and risk assumption and management, determining an important linkage
between diversification and regulatory arbitrage. In the literature, this
theme is exemplified by capital arbitrage behaviour affecting levels of risk
density, which is determined at a twofold level: at the corporate strategy
level, the focus is on strategic areas requiring less capital absorption than
others, while at the business strategies level the management of risk
weights depends on the BM adopted. Peer analysis may give supervisors
some indicators of possible regulatory arbitrage across BMs via Basel risk
weights manipulation. This kind of analysis could help SSM in the
targeted review of internal models (TRIM) to assess the reliability and
comparability of internal rating systems and models. The project is
scheduled by 2019.
Finally, the two perspectives of micro- and macroregulations should be
deepened and cross-analysed to appreciate, at the microlevel, the viability
and sustainability of a bank’s BM and, at the macrolevel, each bank’s
contribution to systemic risk. To this end, in a supervisory perspective,
appropriate BM diversification within the system is an important factor
in reducing risk arising from external shocks. A key point is the definition
of the characteristics and composition of peer groups as already under-
lined, where more in-depth analysis is required on systemic banks.
Information on BM variety is fairly significant, especially in the case of
systemic banks where variety helps to reduce systemic risk.
7.6 Conclusions
Evolving market conditions, technological innovations, regulatory
changes and current monetary policy stances challenge the sustainability
of banks’ BMs. The “business model question” is increasingly grabbing
the attention of bank managers, regulators, investors and financial ana-
lysts. The need to use the BM concept as a tool for analysing a bank’s
performance and assessing its viability requires, first of all, a clear
understanding of what “business model” means, since the existing lit-
erature does not offer a uniform picture.
158 S. Cosma et al.
We start by drawing on strategic management studies to deepen and

specify the concepts of corporate strategy, business strategy and BM.
Three different strands of the literature deal with BMs: IT and
e-business, strategy and strategic groups. In the first, the BM symbolises
how a firm creates, distributes and captures value; in this holistic per-
spective, BM and business strategy often overlap. In the second field of
studies, strategy, the foundation of competitive advantage and value
creation, is implemented at two levels: corporate (what) and business
(how). Corporate strategy delineates the breadth and diversification of
the company’s business portfolio in terms of SBAs; it is the set of
high-order (first level) long-term choices such as growth, size, governance
structures, diversification and internationalisation. Business strategy
(second level) identifies how to achieve competitive advantage in each
SBA. Some scholars see a clear distinction between strategy and BM:
strategy focuses on the market and external competition, while BM has a
more operative nature, focusing on the internal consistency of strategic
choices (operative approach). For others, BM and strategy are different
but strongly related, since BM is the direct result of a firm’s implemented
strategy (systemic approach). A systemic approach is detectable in the
latter strand as well, but with reference to groups of companies with
similar strategies (strategic groups) instead of single firms. Cluster analysis
of data at the firm level is adopted to identify strategic groups; since the
input data are usually the result of both strategic and operative choices,
the overlap between strategy and BM is amplified.
This theoretical framework guides our review of the BBM literature.
In banking, corporate strategy (what) leads to SBA choices reflected in
the business mix (asset & liability composition and income composi-
tion), while business strategy (how) relates to the management of rev-
enues, efficiency and risk in each SBA. Business mix and business strategy
are the components of the BM and the factors affecting performance
indicators (profitability, risk levels, market value).
The first group of studies we review can be traced to the strategic
group literature, in its attempt to classify banks in a small number of BM
archetypes with different performances. However, this approach is based
on a distinction of what the bank is doing (proxied by asset/liabilities
and/or income composition), from how the bank is doing it (revenue

enhancement, efficiency and risk management strategies), so that the BM
concept overlaps with the business mix, whereas other strategic variables
(revenues, efficiency and risk strategies) are implicitly regarded and
evaluated as outcomes of portfolio choices. This BM definition can lead
to the attribution of performance results to the business mix and obscure
the role played by the bank’s skill in managing the individual business
area.
A second stream of the BM literature in banking relies on a wide
definition of BM that combines corporate and business strategies with
context variables, but fails to make a clear distinction between long-term
strategies, business mix and business strategies. The primary aim of this
approach is to identify which BM variables affect banks’ vulnerability,
reflecting the supervisors’ concerns for the consequences of bank
strategies on default events. Most studies deal with the risk of individual
banks. When the analysis extends to systemic risk, important implica-
tions for regulators emerge: some BM variables have opposite effects on a
bank’s tail risk and its exposure to severe shocks in the financial system,
signalling the need for the right balance between micro- and macro-
prudential objectives.
The BBM literature is clearly linked to that exploring diversification
strategies. On the one hand, both consider the business mix, particularly
the distinction between interest and non-interest income. On the other
hand, some control variables (i.e. efficiency and risk profile) popular in
diversification studies may represent the business strategy level. As a
consequence, our review extends to the literature on diversification in
banking to gain additional insights into the BM debate by considering
the pros and cons of diversification.
BM analysis has recently become a supervisory tool. After tracing the
birth and evolution of BMA as a proactive supervisory instrument, we
propose some reflections on the scheme adopted by the supervisory
authorities to evaluate the viability, sustainability and key vulnerabilities
of banks’ current BMs.
Overall, the main lessons stemming from our review of the literature
and the supervisory viewpoint are the following.
160 S. Cosma et al.
• The literature on bank BMs applies the concepts developed in the

strategic management field with some difficulties. The BM is often
restricted to the business mix (cluster approach). When the definition
of BM is more holistic, there are flaws in the recognition of the
different strategic levels (corporate and business).
• Empirically, the degree of reliance on retail deposits and their con-
tribution to the funding of loans are the most significant elements of
the business mix in defining strategic groups.
• In our perspective, the shift towards retail funding as a way of dealing
with the shrinking of the wholesale market due to the financial crisis is
more a change in growth strategy and risk appetite than in business
mix (SBAs tend to be quite stable).
• BBM sustainability has to be evaluated over a time span long enough
to cover entire business and financial cycles. The BM most resilient in
the first wave of the financial crisis (retail banks with loans oriented to
SMEs) has been the worst performer in the subsequent economic
slowdown.
• In the banking system, as in industry, dichotomous strategies are
emerging: to respond to the crisis some banks are restricting their size
and scope of activities, others are growing through M&A. The cluster
approach does not seem sufficiently effective in capturing this trend.
• The availability of equity capital heavily influences banks’ corporate
strategy. From this point of view, listed banks and banks with capital
market access have a competitive advantage. At the same time, the
market evaluation of banks’ securities (debt and equity) is important
both in influencing banks’ strategies and for bank supervisors.
• BM complexity depends on size. Small banks usually have homoge-
nous strategic behaviours and business mix, while big banks tend to
adopt specific BMs that require more detailed information to be
analysed.
• In a large proportion of the BBM literature and in diversification
studies, business mix is proxied by income composition. In general,
greater diversification means greater risk, not always rewarded with
higher profitability. However, diversification in terms of non-interest
income share seems to be beneficial when commissions and fees come
from traditional banking services and detrimental if they derive from
asset-based non-traditional activities such as venture capital, invest-

ment banking and asset securitisation. Therefore, the granularity of
data on the income from services is essential for analysing the influ-
ence of BM on bank risk.
BMA perspectives rely on the intersection between the literature on

BMs and the literature on diversification to get a more integrated rep-
resentation of banking, able to explore both corporate and business
strategies and their connections. At the same time, more detailed data are
needed to allow a deeper understanding of the different elements that
define banks’ BMs.
This holistic and systemic approach to BM valuation is only partially
detectable in the SREP guidelines, where the BM is one of the four areas
under assessment (alongside internal governance and institution-wide
control arrangements, capital adequacy and adequacy of liquidity), rather
than representing the framework for the risk profile analysis of corporate
and business strategic choices, with a limited view on how business mix
and risk and resource management interact.
On the subject of better BM disclosure, it is worthwhile mentioning
the recent (July 2016) statement of the Financial Reporting Council:
“We encourage clear disclosure of a company’s business model as part of the
strategic report, including a description of the main markets in which the
company operates and its value chain”. This form of transparency helps
both academic research and the judgements of analysts and investors,
with positive effects on the information and signalling content of stock
and bond market prices, useful for supervisory authorities in preventing
crisis and for bank managers in acquiring market expectations.
Notes
1. In the literature (for a review see Rossi et al. 2009), diversification is
analysed using two main parameters linked to income sources and geo-
graphical areas. The term functional diversification usually refers to the
profile of the diversification between interest and non-interest bearing
activities.
2. See Stiroh (2009) for a review of the literature.
162 S. Cosma et al.
References
Altunbas, Y., S. Manganelli, and D. Marques-Ibanez. 2011. Bank risk during
the financial crisis—do business models matter? ECB, Working Paper Series,
N. 1394, November.
Amel, D.F., and S.A. Rhoades. 1988. Strategic groups in banking. The Review of
Economics and Statistics 70 (4): 685–689.
Amit, R., and C. Zott. 2001. Value creation in e-business. Strategic Management
Journal, 22, 493–520.
Ansoff, H.I. 1965. Corporate Strategy. McGraw-Hill.
Ayadi, R.E., and W.P. de Groen. 2014. Banking business models monitor 2014:
Europe. Montreal, Joint Centre for European Policy Studies (CEPS) and
International Observatory on Financial Service Cooperatives (IOFSC)
publication.
Ayadi, R.E., W.P. de Groen, I. Sassi, W. Mathlouthi, H. Rey, and O. Aubry.
2016a. Banking business models monitor 2015: Europe, Alphonse and
Dorimène Desjardins International Institute for Cooperatives and
International Observatory on Financial Service Cooperatives (IOFSC).
Ayadi, R., V. Pesic., and, G. Ferri. 2016b. Regulatory arbitrage in EU banking:
Do business models matter? IRCCF Working Paper, Montreal.
Baden-Fuller, C., and S. Haefliger. 2013. Business models and technological
innovation. Long Range Planning 46 (6): 419–426.
Baele, L., O. De Jonghe., and R. Vander Vennet. 2007. Does the stock market
value bank diversification? Journal of Banking & Finance, 31(7), 1999–2023.
Bank of England and FSA. 2012. The Bank of England, prudential regulation
authority. The PRA’s approach to banking supervision, October, London.
Bonaccorsi di Patti, E., R. Felici., and F.M. Signoretti. 2016. Euro area
significant banks: Main differences and recent performance. Bank of Italy,
Questioni di economia e finanza, Occasional Papers, n. 306.
Brighi, P., and V. Venturelli. 2014. How do income diversification, firm size
and capital ratio affect performance? Evidence for bank holding companies.
Applied Financial Economics 24 (21): 1375–1392.
Brighi, P., and V. Venturelli. 2016. How functional and geographic
diversification affect bank profitability during the crisis. Finance Research
Letters 16: 1–10.
Calomiris, C.W., and D. Nissim. 2014. Crisis-related shifts in the market
valuation of banking activities. Journal of Financial Intermediation 23 (3):
400–435.
Campa, J.M., and S. Kedia. 2002. Explaining the diversification discount. The
Journal of Finance 57 (4): 1731–1762.
Casadesus-Masanell, R., and J.E. Ricart. 2010. From strategy to business models
and onto tactics. Long Range Planning 43: 195–215.
Chen, L., J. Danbolt, and J. Holland. 2014. Rethinking bank business models:
The role of intangibles. Accounting, Auditing & Accountability Journal 27:
563–589.
Cotugno, M., and V. Stefanelli. 2012. Geographical and product diversification
during instability financial period: Good or bad for banks? Available at
SSRN: http://ssrn.com/abstract=1989919 or http://dx.doi.org/10.2139/ssrn.
1989919.
DaSilva, C.M., and P. Trkman. 2014. Business model: What it is and what it is
not. Long Range Planning 47 (6): 379–389.
De Jonghe, O., M. Diepstraten, and G. Schepens. 2015. Banks’ size, scope and
systemic risk: What role for conflicts of interest? Journal of Banking &
Finance 61: S3–S13.
De Meo, E., A. De Nicola, G. Lusignani., and L. Zicchino. 2016. European
banks in the XXI century: Are their business models sustainable? (forthcoming).
Demirgüc-Kunt, A., and H. Huizinga. 2010. Bank activity and funding
strategies: the impact on risk and returns. Journal of Financial Economics, 98
(3), 626–650.
Demsetz, R.S., and P.E. Strahan. 1997. Diversification, size, and risk at bank
holding companies. Journal of Money Credit and Banking 29: 300–313.
DeYoung, R., and K.P. Roland. 2001. Product mix and earnings volatility at
commercial banks: Evidence from a degree of total leverage model. Journal of
Financial Intermediation 10 (1): 54–84.
DeYoung, R., and T. Rice. 2004. Noninterest income and financial
performance at US commercial banks. Financial Review 39 (1): 101–127.
DeYoung, R., and G. Torna. 2013. Nontraditional banking activities and bank
failures during the financial crisis. Journal of Financial Intermediation 22 (3):
397–421.
Diamond, D.W. 1984. Financial intermediation and delegated monitoring.
Review of Economic Studies 51: 393–414.
Diamond, D.W. 1991. Monitoring and reputation: The choice between bank
loans and directly placed debt. Journal of Political Economy 99: 689–721.
EBA. 2014. Guidelines on common procedures and methodologies for the supervisory
review and evaluation process (SREP), EBA/GL/2014/13, 19 December.
164 S. Cosma et al.
ECB Banking Supervision. 2015. SSM SREP methodology booklet level playing
field—High standards of supervision—Sound risk assessment.
ECB Banking Supervision. 2016. SSM priorities 2016, January.
ECB. 2016. Recent trends in euro area banks’ business models and implication
for banking sector stability. Financial Stability Review—Special Features,
May.
Elsas, R., A. Hackethal, and M. Holzhäuser. 2010. The anatomy of bank
diversification. Journal of Banking & Finance 34 (6): 1274–1287.
FSA. 2009. The Turner review. A regulatory response to the global banking crisis.
March, London.
Gallo, J.G., V.P. Apilado, and J.W. Kolari. 1996. Commercial bank mutual
fund activities: Implications for bank risk and profitability. Journal of Banking
& Finance 20 (10): 1775–1791.
George, G., and A.J. Bock. 2011. The business model in practice and its
implications for entrepreneurship research. Entrepreneurship Theory and
Practice 35: 83–111. doi:10.1111/j.1540-6520.2010.00424.x.
Gertner, R., D. Scharfstein, and J. Stein. 1994. Internal vs. external capital
markets. Quarterly Journal of Economics 109: 1211–1230.
Goddard, J., D. McKillop, and O.J. Wilson. 2008. The diversification and
financial performance of US credit unions. Journal of Banking & Finance 32
(9): 1836–1849.
Graham, J.R., M. L. Lemmon., and Wolf, J.G. 2002. Does corporate
diversification destroy value? The Journal of Finance, 57 (2), 695–720.
Gualandri, E. 2016. Vigilanza unica: traguardi raggiunti e cantieri aperti.
Bancaria 5: 2–18.
Herring, R.J., and A.M. Santomero. 1990. The corporate structure of financial
conglomerates. Journal of Financial Services Research 4: 471–97.
Klang, D., M. Wallnöfer, and F. Hacklin. 2014. The business model paradox: A
systematic review and exploration of antecedents. International Journal of
Management Reviews 16: 454–478.
Klein, P.G., and M.R. Saidenberg. 1997. Diversification, organization and
efficiency: Evidence from bank holding companies, mimeo.
Köhler, M. 2014. Does non-interest income make banks more risky?
Retail-versus investment-oriented banks. Review of Financial Economics 23
(4): 182–193.
Köhler, M. 2015. Which banks are more risky? The impact of business models
on bank stability. Journal of Financial Stability 16: 195–212.
Laeven, L., and R. Levine. 2007. Is there a diversification discount in financial

conglomerates? Journal of Financial Economics 85 (2): 331–367.
Lautenschlager, S. 2016. European banking supervision business model analysis,
CEO/CFO/CRO-Roundtable 7 July.
Leask, G. 2004. Is there still value in strategic group research? Aston Business
School research papers, no. RP0404, Aston University, Birmingham (UK).
Lepetit, L., E. Nys, P. Rous, and A. Tarazi. 2008. Bank income structure and
risk: An empirical analysis of European banks. Journal of Banking & Finance
32 (8): 1452–1467.
Llewellyn, D.T. 1996. Universal banking and the public interest: A British
perspective. Universal Banking: Financial System Design Reconsidered, 161–204.
Mercieca, S., K. Schaeck, and S. Wolfe. 2007. Small European banks: Benefits
from diversification? Journal of Banking & Finance 31 (7): 1975–1998.
Mergaerts, F., and R. Vander Vennet, R. 2016. Business models and bank
performance: A long-term perspective. Journal of Financial Stability, 22, 57–75.
Moloney, N. 2012. Supervision in the wake of the financial crisis. Achieving
effective ‘Law in Action’—A challenge for the EU. In Financial Regulation and
Supervision. A Post-Crisis Analysis, ed. E. Wymeersch, K.J. Hopt, and G.
Ferrarini, Oxford University Press, Oxford.
Morris, M., M. Schindehutte, and J. Allen. 2005. The entrepreneur’s business
model: Toward a unified perspective. Journal of Business Research 58: 726–735.
Mottura, P. 2011. Banche; strategie, organizzazione e concentrazioni. Milano:
Egea.
Osterwalder, A. 2004. The business model ontology-a proposition in a design
science approach (2004), These Présentée à l’Ecole des Hautes Etudes
Commerciales, de l’Université de Lausanne Universite de Lausanne.
Osterwalder, A., and Y. Pigneur. 2012. Business model generation. New York:
Hoboken, John Wiley.
Passmore, S.W. 1985. Strategic groups and the profitability of banking.
Research Paper 8501, Federal Reserve Bank of New York.
Penrose, E. 1959. The theory of the growth of the firm. Oxford: Oxford University
Press.
Porter, M.E. 1996. What is strategy? Harvard Business Review 74 (6): 61–78.
Rajan, R., H. Servaes, and L. Zingales. 2000. The cost of diversity: The
diversification discount and inefficient investment. Journal of Finance 55: 35–80.
Rajan, R.G. 1992. Insiders and outsiders: The choice between informed and
arm’s length debt. The Journal of Finance XLVII 4: 1367–1400.
166 S. Cosma et al.
Roengpitya, R., N.A. Tarashev, and K. Tsatsaronis. 2014. Bank business models,
55–65. Quarterly Review December: BIS.
Rossi, S.P., M.S. Schwaiger, and G. Winkler. 2009. How loan portfolio
diversification affects risk, efficiency and capitalization: A managerial behavior
model for Austrian banks. Journal of Banking & Finance 33 (12): 2218–
2226.
Shafer, S.M., H.J. Smith, and J.C. Linder. 2005. The power of business models.
Business Horizons 48 (3): 199–207.
Short, J.C., D.J. Ketchen, T.B. Palmer, and G.T. Hult. 2007. Firm, strategic
group and industry influences on performance. Strategic Management Journal
28: 147–167.
Stein, J.C. 1997. Internal capital markets and the competition for corporate
resources. Journal of Finance 52 (1): 111–133.
Stein, J.C. 2002. Information production and capital allocation: Decentralized
versus hierarchical firms. The Journal of Finance 57: 1891–1921.
Stiroh, K.J. 2004. Diversification in banking: Is noninterest income the answer?
Journal of Money, Credit, and Banking 36 (5): 853–882.
Stiroh, K.J. 2006. A portfolio view of banking with interest and noninterest
activities. Journal of Money, Credit, and Banking 38 (5): 1351–1361.
Stiroh, K.J. 2009. Volatility accounting: A production perspective on increased
economic stability. Journal of the European Economic Association 7: 671–696.
Stiroh, K., and A. Rumble. 2006. The dark side of diversification: The case of
US financial holding companies. Journal of Banking & Finance 30 (8): 2131–
2432.
Stulz, R. 1990. Managerial discretion and optimal financial policies. Journal of
Financial Economics 26: 3–27.
Teece, D.J. 1982. Towards an economic theory of the multiproduct firm.
Journal of Economic Behavior & Organization 3 (1): 39–63
Teece, D.J. 2010. Business models, business strategy and innovation. Long
Range Planning 43: 172–194.
Vallascas, F., F. Crespi, F., and J. Hagendorff. 2012. Income diversification and
bank performance during the financial crisis. Available at SSRN: http://
ssrn.com/abstract=1793232 or http://dx.doi.org/10.2139/ssrn.1793232.
Van Oordt, M., and C. Zhou. 2014. Systemic risk and bank business models.
WP No. 442, De Nederlandishe.
Williams, B. 2016. The impact of non interest income on bank risk in Australia.
Journal of Banking & Finance 73: 16–37.
Wirtz, B.W., A. Pistoia, S. Ullrich, and V. Göttel. 2016. Business models:

Origin, development and future research perspectives. Long Range Planning
49: 36–54.
Wirtz, B.W. 2011. Business model management: Design instrument-success factor.
Wiesbaden: Gabler.
Zott, C., R. Amit, and L. Massa. 2011. The business model: Recent
developments and future research. Journal of Management 37 (4): 1019–
1042.
8
On European Deposit Protection Scheme(s)
Milena Migliavacca
8.1 Introduction
The stability of the banking sector has proven to be a crucial requisite for
the overall stability of a nation’s economy, particularly in recent years
(Wheelock and Wilson 1995; Acharya 2009; Lambert et al. 2015).
Financial safety nets are systems of legislative measures put in place to
guarantee banks’ stability (Demirgüç-Kunt and Huizinga 1999, 2004);
one of the most relevant actors of a national safety net, along with the
prudential regulator, the lender of last resort and the supervisory
authority is the deposit protection scheme (DPS) (Laeven 2002; Schich
2008). DPSs are public authorities designed to reimburse depositors in
place of their financial institution, when the latter is insolvent; within the
European Union, they are in force in all Member States since 1994
(Directive 94/19/EC). This analysis focuses specifically on the DPSs
across the EU28, as they have been recognised among the most effective
M. Migliavacca (&)
Catholic University, Milan, Italy
e-mail: milena.migliavacca@unicatt.it
170 M. Migliavacca
tools to temper investors’ panic during the financial turmoil following

both the global financial crisis and the European sovereign crisis and they
are undergoing a major revision process (Demirgüç-Kunt et al. 2015).
The banking industry is particularly vulnerable to instabilities because
of an intrinsic characteristic of its core business: the maturity transfor-
mation that leads to fractional-reserve banking. There is a physiological
mismatching of volumes, interest rates and maturities between assets and
liabilities in a bank’s balance sheet, so only a fraction of bank deposits are
available for immediate withdrawal. If the demand for deposits exceeds
the fraction hold in a bank’s account (as, for instance during a bank run),
financial institutions cannot refund their depositors and the intervention
of DPSs becomes necessary to prevent the depositors from losing the
totality of their savings. DSPs safeguard the banking stability by acting
contemporarily on the depositors’ and the financial institutions’ side. On
one hand, they protect depositors against the unavailability of their
deposits, and on the other hand, they protect banks from contagious
runs; they may also provide additional liquidity to credit institutions
during financial crises. These objectives contribute to the DPSs’ stabili-
sation ultimate purpose. However, one of the costs of deposit insurance
provision is the potential for moral hazard, as depositors have less
incentive to monitor their bank managers’ behaviour, knowing that their
deposits are guaranteed by national funds (Gropp and Vesala 2004;
Hovakimiam et al. 2003). Banks as well can engage in imprudent
practices, secure in the knowledge that if the high-risk loans do not pay
off, deposit insurance will shelter their principal (Demirgüç-Kunt and
Kane 2002; Pennacchi 2006). Another subtle drawback of DPSs lays in
the specific characteristics of the deposit insurance funds, which are
largely left to the Member States’ discretion, despite the constant har-
monisation effort by the European legislator over the years (Carbo‐
Valverde et al. 2012), because such asymmetries may interfere with
cross-country fair competition in the banking sector (Cordella and Yeyati
2002; Engineer et al. 2013). DPSs have been in the spotlight for both
policy makers and scholars for decades, but particularly so after the global
financial crisis and the following European sovereign crises. In this
period, they have been the core topic of several empirical and theoretical
contributions and went through significant legislative changes. There is a
rich literature that analyses benefits and costs of explicit deposit insurance
and looks for their optimal design (e.g. Diamond and Dybvig 1983;
Pennacchi 1987; Kane 1995; and Pennacchi 2006) and two main
streams stem from it: one that maintains that the stabilisation effect of
DPSs prevails on moral hazard incentives (e.g. Diamond and Dybvig
1983) and another, which supports the opposite view (e.g. Pennacchi
2006). A handful of contributions (among others, Anginer et al. 2014),
moreover, tries to refine the analysis by disentangling the prevailing effect
in particular circumstances, such as the financial crisis. Anginer et al.
(2014) provide evidence that the overall effect of an explicit deposit
insurance mechanism is detrimental to the bank stability, but highlight a
prevailing stabilising effect during financial turmoil as well.
Finally, among the most recent qualitative contributions, Cariboni
et al. (2010) and Arnaboldi (2014) shed light on the effectiveness of the
European DPSs during the crisis and their legal evolution.
This contribution provides an analytical overview of the DPSs set-up
in the European Union Member States and analyses the main features of
their design; it investigates the level of harmonisation reached by the
DPSs within the EU28. According to the author, the way the features of
the deposit insurance fund are combined may influence its prevailing
effect (stabilising or risk-seeking); for instance, some DPS’s characteris-
tics, such as risk-based premia or ex-ante contribution may indeed mit-
igate the moral hazard that too generous DPS may trigger (Duan et al.
1992; Forssbaeck 2011).
The two main data sources exploited in the chapter are the 94/19/EC
Directive and its development and the World Bank’s Bank Regulation
and Supervision Surveys. This chapter is structured as follows. In the
second section, an overview of the reference Directives is given, along
with a brief excursus on the DIS’s features analysed. The third section
details the status quo of the deposit insurance designs of the EU Member
States and provides some insight into the most common features com-
binations and how they changed after the financial crisis. The final
section draws some final remarks.
172 M. Migliavacca
8.2 The Legal Framework

This paper aims at drawing a detailed overview of the DPSs in the EU28
by taking into account a number of features that characterise the national
deposit insurance fund/s of the Member States. Information on the
deposit insurance design has been mainly taken from the Bank
Regulation and Supervision surveys. The Surveys, carried out by the
World Bank, are a source of world-wide data on how banks are regulated
and supervised, and they comprise a section dedicated to deposits pro-
tection schemes. The surveys have been carried out in 2001, 2003, 2007
and 2012 giving the chance to closely monitor the evolution of the DPSs
over the years. For the sake of this analysis, nine features of the EU28
Member States’ DPSs are taken into account (see Table 8.1).
Coverage Limit_ indicates the extent of the investors’ deposits cover-
age. Between 1994 and 2008, the coverage limit in the EU28 spanned
from a minimum of €3400 (Romania) to a maximum of €103,291
(Italy). Some Member States, furthermore, granted blanket guarantees to
depositors during the period 2008–2010. From 2009 on, the national
coverage limits were harmonised to the common level of €50,000, raised
to €100,000 in 2010, according to the EU-Directive 2009/14.
In each Member States, the deposit coverage limit is considered per
account, in Euros.
Coinsurance_ With coinsured deposits, account holders are explicitly
insured for less than 100% of their savings, according to the national
legislation. There has been much debate on this feature, designed as an
incentive for depositors’ monitoring. It might be questionable, though,
in the first place, whether depositors have the necessary competences to
carry out such an activity; moreover, as in the deposit insurance context,
the banks themselves are the risk creators, high level of coinsurance may
spur risk-seeking choices by the banks’ management, so the net effect of
this feature on the banks’ moral hazard incentives remains unclear. After
the financial crisis, though, Directive 2009/14/EC banned Coinsurance
from the possible characteristics of deposit insurance funds.
Fund Management_ EU Member States may choose whether their
deposit insurance scheme is managed by the private sector, by the public
Table 8.1 Deposit protection scheme (DPS) features

Feature Survey question Comments
Coverage
Coverage limit What is the deposit insurance The maximum amount to
limit per account (in US$ which depositors can be
and local currency)? reimbursed. Yearly values,
in Euros
Coinsurance Is there formal coinsurance? Coinsurance means that the
Yes?/No? depositor has to accept part
of the risk for his/her own
account, the presence of
coinsurance increases the
banks’ MH incentives
Funding and management
Fund The insurance fund is The moral hazard incentive is
management managed by considered higher if the
(a) The private sector alone; insurance fund is
(b) Jointly by private/public administered solely by
officials banks
(c) The public sector alone
Source of Funding is provided by According to the literature,
funding a) Government; the moral hazard incentive
b) Banks is considered higher when
c) Combination the funds are provided by
the government or jointly
by banks and government.
Levies can typically be
imposed upon banks, e.g.
based on volume of insured
deposits, and/or on the basis
the risk profile of the bank
Ex-ante Is there an ex-ante “Ex-ante funds” implies the
contribution fund/reserve to cover presence of an upfront
deposit insurance claims in amount of funds available
the event of the failure of a to be used to compensate
member bank? depositors, so it is expected
Yes?/No? to dampen the moral
hazard incentives
Risk-adjustment Do deposit insurance According to the literature,
fees/premiums charged to Risk-based premia dampen
banks vary, based on some the banks’ MH incentives
assessment of risk?
Yes?/No
(continued)
174 M. Migliavacca

Intervention
Intervention Does the deposit insurance This authority allows the
power agency/fund administrator agency to, e.g. issue cease
have the following powers and desist orders, impose a
as part of its mandate? form of temporary
Bank intervention authority administration which
overrides management,
replace managers, etc
Power to cancel Does the deposit insurance The presence of this power
authority by itself have the reduces banks’ MH
legal power to cancel or incentives
revoke deposit insurance
for any participating bank?
Legal power Can the deposit insurance Legal action implies
agency/fund take legal enforceable corrective
action for violations of laws, action based on powers
regulations, and by-laws granted in law or
against bank directors or regulation, providing direct
other bank officials? powers to order action
Yes?/No? against individuals in the
bank or its Board
dampening the MH
incentives
Notes This table summarises nine characteristics of DPSs analysed, divided up into
three buckets: Coverage, Funding & Management and Intervention
Source Adapted from Bank regulation and Supervision surveys
sector or jointly by both of them. Pure private funding is expected to

relax the discipline on DIS management and to leave room for possible
banks’ moral hazard incentives, which may be tempered if both private
and public officials jointly manage the deposit insurance fund and are
even more reduced if the public sector alone administers it.
Source of Funding_ indicates whether the government, the banks or the
two actors together fund the national DPS. The possible banks’ moral
hazard incentives decrease significantly if the source of funding is the
banks themselves; in this case, in effect, taxpayers are not involved in
bearing part or the whole risk arising from the deposits insurance funds’
management decisions.
Ex-ante Contribution_ the contributions to the deposit insurance fund

can be collected on a regular basis to meet potential future obligations
(ex-ante contribution) or only after a bank’s failure (ex-post). Before
Directive 2014/49/EU, 19 countries had ex-ante contributions (Belgium,
Bulgaria, Croatia, Cyprus, Denmark, Finland, France, Germany, Greece,
Hungary, Ireland, Latvia, Lithuania, Malta, Poland, Portugal, Romania,
Slovakia and Spain). This option is considered among both scholars and
practitioners to enhance the stabilising effect of DPSs and has, in effect,
always been encouraged by the European Supervisory authorities (FSB
2012; Directive 2009/14/EC; IADI 2015), as the presence of readily
accessible funds supports the clients’ confidence. The difference between
ex-ante and ex-post contribution, though, might not be stark, as in some
Member States with ex-post contribution, there might be “virtual funds”,
not officially collected by the DPSs but privately set by the member
banks. Another possible element of heterogeneity among the Member
States with an ex-ante contribution system is the amount of the periodical
contribution.
Risk-based Contribution_ The contributions to the DPS are based on the
degree of riskiness of each member, measured by a set of ad hoc indicators.
The methodologies that assess the banks’ risk may take into account the
banks’ business model, the level of total deposits or total liabilities, the
different types of covered deposits or eligible deposits1 and may vary across
countries. Despite there might be room for unequal treatment, as the
risk-weighting criteria are not homogeneous among the Member States,
this feature is an effective way to minimise the moral hazard incentives on
the banks’ side and has been introduced as a mandatory feature of
European DPSs since 2014 (Directive 2014/49/EU).
The three Intervention Powers reported in Table 8.1 refer to the
supervisory agency’s right to take actions against the management to
restore the bank’s solvency; these interventions comprise temporary
administration, which overrides the bank’s management, the exclusion of
a bank from the national DPS and the possibility to take legal actions
against the bank’s management. The presence of these features serves as a
rather powerful deterrent against banks’ moral hazard, as confirmed by
IADI (2014).
176 M. Migliavacca
All the EU28 deposit insurance schemes are explicit, but may differ in
the way the features described above are combined. DPSs’ features have
been designed to protect deposits, enhance depositor confidence and
minimise bank-runs risk; with too generous schemes, though, the sta-
bilising effect may be outweighed by banks’ moral hazard incentives
(Demirgüç-Kunt and Detragiache 2002; Barth et al. 2013; Angkinand
and Wihlborg 2010; IADI 2013a, b, c, 2014).
In order to have a better understanding of the DPSs’ design within the
EU28, the features described by the World Bank’s Bank Regulation and
Supervision Surveys need to be considered within the legal framework in
force. The first attempt to homogenise the deposit insurance regulation
in the EU was the Directive 94/19/EC in 1994. The Directive ensured
that the DPSs of the Member States had at least three main character-
istics in common. All Member States needed to have at least one explicit
DPS. If one bank went bankrupt, the national DGS would reimburse the
account holders, and the minimum protection level was initially set to
ECU 15,000 and then raised at ECU 20,000 on 01/01/2000. The third
common feature among the EU Members’ DPSs was payout time limits:
21 days for the deposit insurance authority to make the determination
that a credit institution is unable to repay the depositors and three
months for the actual reimbursement, with the possibility of extending
this deadline up to nine months. Regarding this last point, however, the
Directive did not specify the period in which depositors could submit the
claim necessary to trigger the reimbursement process. The implementa-
tion of more stringent rules was left to the Member States, which were
free to choose the categories of deposits/depositors to include in the DPS,
how to finance and how to manage the fund/s and whether or not to
resort to coinsurance; even among the Member States that apply coin-
surance, the coinsured percentage of the deposits could significantly vary.
The characteristics of the contributions themselves could vary, as well:
members could choose between ex-ante and ex-post contributions and
even among those countries, which chose the ex-ante method, the
amount of resources periodically set aside could be heterogeneous,
depending on the structure of the national banking industry and the legal
role of the DPS, mainly. The payouts could be risk based or not, but
once again, even among the Member States with risk-based contribution,
the weighting coefficients were not harmonised. Each Member State,

furthermore, could decide the intervention powers to leave to the
supervisory authority. Even the coverage limit used to span within a wide
range from the minimum guarantee of €20,000 in Romania to €103,291
in Italy. As in some countries the coinsurance was in force, the actual
coverage could be even more heterogeneous, because the amount guar-
anteed could be lower than the coverage level, depending on the per-
centage of coinsurance. Finally, the Member States could exclude from
the guarantee mechanism or lower the relative level of coverage of a
number of typologies of deposits listed in the Directive. The Directive
allowed considerable scope for Member States to design the DPSs, which
were best in line with their national market conditions, but such freedom
left room to distortions within those banking systems in which
cross-country financial institutions coexisted. To address this issue, the
Directive enforced the topping-up principle: the first attempt to homo-
genise the DPSs around Europe. According to this principle, the host
country’s conditions prevailed at national level, irrespectively of the
origin of the hosted credit institutions and branches. The Directive on
deposit guarantee schemes adopted in 1994 was substantially unchanged
for 14 years, leaving the bulk of the DPSs’ design to the single Member
State’s autonomous decision. This excessive discretionary power turned
out to be detrimental to the financial stability, and the outburst of the
global financial crisis forced the legislator to introduce the Amending
Directive 2009/17/EC. In such a delicate moment, three main changes
were made in order to boost depositors’ confidence and limit contagious
bank runs. The coverage limit was increased immediately to €50,000 and
the Member States were asked to further raise the coverage for the
aggregate deposits of each depositor to €100,000 by the end of
December 2010. The Directive banned the use of coinsurance, as well.
The third issue addressed is the payout time; the deadline for the
determination of a bank’s inability to repay depositors was reduced from
21 to 5 working days, and the repayment of claimed deposits was
shortened from three months, extendible to nine months, to 20 working
days; only under exceptional circumstances and after formal approval by
the competent authorities, this deadline could be postponed by 10 more
working days.
178 M. Migliavacca
The Amending Directive represented an emergency measure; however,

by comparing the information of World Bank’s Bank Regulation and
Supervision Surveys of 2003 and 2012, no major changes in the most
qualifying features of the European DPSs emerge, as outlined in
Table 8.2.
From a preliminary analysis, the substantial static of the DPSs’ designs
might surprise, given the shortcomings some Member States’ DPSs faced
during the most turbulent periods or the crisis. As a matter of fact, from
the debate immediately following the outburst of the financial crisis,
which led to the Amending Directive 2009/14/EC, the need for the
deposit insurance regulation to be deeply renewed emerged clearly.
Because of the financial crisis, the need for timely negotiations
impeded addressing the numerous open issues the European DPS
framework suffered from, but at least three main areas of intervention
were identified. Despite the need for a pan-European DPS began to be
tangible, the proposal was rejected by nearly all stakeholders, so a detail
cost/benefit analysis on the topic was encouraged. Ex-ante, risk-based
contribution systems were promoted as well, even though the pro-
cyclical effect of too heavy ex-ante contributions during times when
banks were already in difficulty prompted further debate on this feature.
On 16 April 2014, the Directive 2014/49/EU answered to these
unaddressed issues by introducing several changes; the most important
one involved the contributions to the deposit insurance fund. Yearly ex-
ante premia collection was introduced, and the target level of 0.8% for
covered deposits was set to be reached by 2024. The Directive imposes,
moreover, the contributions to DPSs to be based on the amount of
covered deposits and the member’s degree of risk, measured by a set of
specific indicators assessed according to EBA’s guidelines (2015).
Member States are allowed to use their own risk-weighting method;
however, these methods shall be approved by the national competent
authority, and EBA should be informed, as well. Furthermore, the
weighting techniques shall be revised 3 years after the entry into force of
the Directive and every five years afterwards; the weights anyhow should
reflect the risk profiles of the member credit institutions, including their
different business models. Furthermore, a number of changes were
introduced to simplify and speed up the depositors’ reimbursement.
Table 8.2 Design features distribution across European countries
Country AT BE BG CY CZ DK EE FI FR
Feature/pre/post-crisis pre Post pre post pre post pre post pre post pre post pre post pre post pre post
Coinsurance X X X X X
DIS source of funding
Banks X X X X X X X X X X X X X X X X X
Government or both X
Ex-ante premiums X X X X X X X X X X X X X X
Risk-adjusted premiums X X X X X X
DIS power of intervention on bank X
8
members
DIS power to cancel members X X X X X
DIS legal power on bank X X X X X
management
DIS administration
Private sector X X X X X X X
Public sector or jointly X X X X X X X X X X X
Country DE GR HR HU IE IT LV LT LU MT
Feature/pre/post-crisis pre post pre post pre post pre post pre post pre Post pre post pre post pre post Pre post
Coinsurance X X X X X X X X
Banks X X X X X X X X X X X X X X X X X X
Government or both X X
Ex-ante premiums X X X X X X X X X X X X X X X
Risk-adjusted premiums X X X X X X X X X
DIS power of X X X X
intervention on bank
members
DIS power to cancel X X X X X X X X X X X X
members
On European Deposit Protection Scheme(s)
DIS legal power on bank X X X X X X X X

management
DIS administration
Private sector X X X X
Public sector or jointly X X X X X X X X X X X X X X X X
179
Country NL PL PT RO SK SL ES SW UK
180
Feature/pre/post-crisis pre post pre post pre post pre post pre post pre post pre post pre post pre Post
Coinsurance X X X X X X X
Banks X X X X X X X X X X X X – X X
Government or both X X X –
Ex-ante premiums X X X X X X X X X X
Risk-adjusted premiums X X X X
DIS power of intervention on bank X X X
members
DIS power to cancel members X X
M. Migliavacca
DIS legal power on bank X X X X X X X

management
DIS administration
Private sector X X
Public sector or jointly X X X X X X X X X X X X X X X X
Source World Bank’s Bank Regulation and Supervision Surveys (2003, 2012)
The payout time was sensibly shortened from 20 working days to 15 by

the end of 2019, to 10 by the end of 2023 and to 7 from 2024 on;
furthermore, during the transitional period, DGSs should ensure
depositors to be able to have access to part of their deposits in order to
cover their cost of living. Moreover, the scheme repays the depositors on
its own initiative; no official application from depositors is needed.
The Member States should, then introduce periodical stress tests on
their national DPSs, have in place adequate systems to determine
potential criticalities and produce annual reports on their activities. EBA
is supposed to carry out peer reviews in order to guarantee the resilience
of national DPSs every five years, as well. If the deposit insurance fund’s
resources are not enough to cover claimed deposits, the Member States
should furthermore ensure that their DPSs have in place adequate
alternatives to have access to alternative short-term funding. In this case,
extraordinary additional contributions may be imposed to banks or
governments may step in the scene, whether they are officially among the
DPS funders or not. The DPS, alternatively, can borrow financial
resources from the market, usually from the European Central Bank.
One further possible alternative funding source is for DPSs to borrow
from one another, in order to create an informal network, that does not
require changes in the national legal frameworks, but that might still lead
the way towards a fully fledged European DPS. Even though the
European DPSs harmonisation process began over twenty years ago and
the Directive 2014/49/EU introduced much more stringent rules, there
are still stark differences across the EU.
8.3 The EU Status Quo

In the EU28, there are almost forty different DPSs (see Table 8.3);
usually each country has one DPS, which covers the deposits of all the
banks in a country, however, more than one DPS may operate in a single
state. This is the case of Austria (five DPSs), Germany (four DPSs), the
Czech Republic and Spain (three DPSs) and Cyprus, Italy and Portugal
(two DPSs). In these cases, a single deposit insurance fund is established
for each typology of DPS. Usually, specific DPSs are set according to
182 M. Migliavacca
Table 8.3 Deposit protection schemes (DPSs) in the EU28

Country EU Deposit insurance organisation
membership
Austria 1995 Deposit Protection Company of the Austrian
Commercial Banks Ltd
O¨ sterreichische Raiffeisen-Einlagensicherung
reg.Gen.m.b.H. (O¨ RE)
AT3 Austria 3 Sparkassen-Haftungs
Aktiengesellschaft
AT4 Austria 4 Schulze-Delitzsch
Haftungsgenossenschaft reg.Gen.m.b.H.
AT5 Austria 5 Hypo-Haftungs-Gesellschaft m.b.H.
Belgium 1958 Fonds de Protection
Bulgaria 2007 Bulgarian Deposit Insurance Fund
Croatia 2013 State agency for deposit insurance and bank
resolution
Cyprus 2004 Deposit Protection Scheme
Deposit Protection Scheme for cooperative
societies
Czech 2004 Deposit Insurance Fund
Republic
Denmark 1973 The Guarantee Fund for Depositors and Investors
Estonia 2004 Deposit Guarantee Sectoral Fund
Finland 1995 Deposit Guarantee Fund
France 1958 Fonds de Garatie des Depots (FDG)
Germany 1958 Entschädigungseinrichtung deutscher Banken
GmbH (EdB)
VÖB-Entschädigungseinrichtung GmbH
Entschädigungseinrichtung der
Wertpapierhandelsunternehmen (EdW)
Greece 1981 Hellenic Deposit Guarantee Fund
Hungary 2004 National Deposit Insurance Fund (NDIF)
Ireland 1973 The Deposit Guarantee Scheme (DGS)
Italy 1958 Fondo Interbancario di Tutela dei Depositi (FITD)
Fondo di Garanzia dei Depositanti del Credito
Cooperativo
Latvia 2004 Deposit Guarantee Fund
Lithuania 2004 Valstybės įmonė “Indėlių ir investicijų draudimas”
Luxemburg 1958 Luxembourg Deposit Guarantee Association
Malta 2004 Depositor Compensation Scheme
Netherlands 1958 Depositogarantiestelsel
Poland 2004 Bankowy Fundusz Gwarancyjny (BFG)
Portugal 1986 Fundo de Garantia de Depósitos
(continued)

Fundo de Garantia do Crédito Agrícola Mútuo
Romania 2007 Deposit Guarantee Fund in the banking system
Slovakia 2004 Deposit Protection Fund
Slovenia 2004 The central bank of the Republic of Slovenia
Spain 1986 Fondos de Garantía de Depósitos (FGD)
Sweden 1995 Swedish National Debt Office
UK 1973 Financial Services Compensation Scheme
Notes This table presents the national DPSs in each EU28 Member State
banks’ specialisation and operate autonomously; the obligation to pay

contributions and refund depositors applies primarily to the DPS to
which the affected bank belongs to. In case the DPS in question is not
able to refund the insolvent banks’ depositors in full, the other national
DPSs get involved in the reimbursement process. Only in the unlikely
event that all the national DPSs are jointly not able to refund the claimed
deposits, the firstly affected DPS should turn to alternative external
sources of funding as, for instance, loans from public or private third
parties. Finally, besides the official deposit insurance funds, additional
so-called virtual funds can be established, as well on a voluntary base
among groups of banks; this possibility increases significantly the com-
plexity and homogeneity of DPS across the EU.
Over the past decades, DPSs have developed in quite heterogeneous
ways in Europe, as most of the decisions regarding their characteristics
were left to single Member States, hence the numerous different DPSs
across the EU. The following brief overview of the current DPSs,
according to the latest World Bank’s Bank Regulation and Supervision
Survey (2012), sheds light on the different design the EU282 Member
States chose for their national DPSs.
Austria: In Austria, there is a legally autonomous deposit insurance
agency, which does not have by itself any intervention authority, nor the
possibility to revoke deposit insurance for any participating bank or take
legal action against banks’ directors and officials. The insurance funding
is jointly provided by public and private institutions, according to a
three-stage mechanism. At first, the DPS members are required to pay
proportionate contributions; if a DPS is unable to reimburse the insured
184 M. Migliavacca
deposits in full, the other DPSs are obliged to make proportionate con-
tributions in order to cover the shortfall. In case all the DPSs together are
unable to pay out the claimed insured deposits in full, the primarily con-
cerned DPS must issue debt securities in order to meet the remaining
payment obligations. The Federal Minister of Finance may assume liabil-
ities on behalf of the federal government, according to a special govern-
mental authorisation and tops up the difference to the maximum insured
amount. Despite the source of funding is both public and private, the
private sector alone manages the fund. Foreign currency deposits, inter-
bank deposits and deposits of the foreign subsidiaries of domestic banks are
excluded from the coverage. The contributions to the deposit insurance
fund, before the Directive 2014/49/EU was in force, were ex-post and not
based on some assessment of the members’ risk. After the financial crisis,
Austria increased the amount of covered deposits, and during the most
critical part of the crisis, the Austrian Government guaranteed in full the
account holders’ deposits and banks’ debt.
Belgium: The Belgian banks alone contribute to the deposit insurance
fund, which, though, is managed by both private and public officials.
Even before the Directive 2014/49/EU was in force, the contributions to
the insurance fund were periodical, but not risk based. The deposit
insurance agency has both intervention authority and the power to take
legal actions for the violation of its regulations and by-laws against bank
directors and other officials. It cannot, though, cancel or revoke a bank’s
the membership to the DPS. The deposit insurance coverage is per
depositors per institutions, and interbank deposits and deposits of foreign
subsidiaries of domestic banks are excluded by the insurance mechanism.
Bulgaria: The DPS in Bulgaria is funded by banks only, but managed by
both private and public actors. Even prior to Directive 2014/49/EU, the
contributions to the insurance fund were periodical, but not risk based. The
insurance authority has limited powers, and it cannot intervene against
banks’ officials or directors nor impose temporary administration; it is not
allowed to revoke a membership from the respective DPS, either. From
2008 to 2010, the percentage of the total deposits of commercial banks
covered by the DPS rose from 54.31 to 65.39% covering savings per
depositor per institution, but excluding interbank deposits and deposits of
both foreign subsidiaries and branches of national banks. In accordance
with the Directive 2009/14, as a result of the global financial crisis, Bulgaria
raised its coverage limit to €100,000.
Croatia: The claim for payment by the deposit insurance system is
triggered by a court-declared bank’s bankruptcy. The peculiarity of
Croatian DPS is that the insurance fund is entirely funded by banks but
exclusively managed by the public sector. The contributions to the fund
have always been ex-ante, but not risk based until the Directive 2009/14
entered into force. The Croatian insurance authority can take legal action
directly against banks’ directors and officials in case they violate its reg-
ulations or by-laws; it can replace management and impose some form of
temporary administration if necessary, as well.
Cyprus: The Cypriot DPS is funded by banks but jointly managed by
banks and the public sector. Both a bank’s bankruptcy and the banking
supervisor’s decision can trigger the deposits repayment claim. The
Cypriot deposit insurance authority, which is a legally independent
entity, is particularly week, regarding its intervention power: its authority
in effect is limited to banks’ examination. The participation in the
deposit insurance system is not only compulsory for domestic banks, but
for foreign bank subsidiaries and branches, as well; deposits of foreign
subsidiaries of domestic banks, though, are excluded from the DPS.
Banks had to periodically contribute to the deposit insurance fund, even
before the Directive 2009/14 made it compulsory, but before the entry
into force of the Directive, the periodical contributions did not depend
on the members’ riskiness. As a result of the global financial crisis, the
contributions to the Cypriot DPS were increased, as well as the amount
of guaranteed deposits.
Denmark: The Danish DPS guarantees deposits per depositor per
institution up to a maximum of €40,000 increased to €100,000 in com-
pliance to the Directive 2009/14. Only banks fund the DPS, which is,
though, jointly managed with public officials. As in a number of EU28
Member States, the intervention power left to the deposit insurance
authority is rather limited; they do not have intervention power, nor legal
power against the banks’ officials; they cannot exclude banks from the DPS
or have access to information collected by the banking supervision, either.
Domestic and foreign banks subsidiaries are included in the DPS, but
foreign bank branches are not and the deposits of foreign branches of
186 M. Migliavacca
domestic banks are excluded from the Danish DPS, as well. Even before
respectively Directive 2009/14/EC and 2014/49/EU, there was no coin-
surance and the contributions to the fund were periodical (ex-ante regime).
Estonia: The Estonian DPS is jointly administered by banks and the
public sector. The deposit insurance authority is an autonomous entity,
but it only has the power to access to information collected by the
banking supervisor and cannot directly intervene against the banks
belonging to the national DPS. In Estonia, foreign banks’ branches do
not necessarily have to take part in the DPS; the participation is required
to foreign bank subsidiaries, instead. The DPS is designed per deposit per
institution, and before the Directive 2014/49/EU, the contributions to
the fund were collected ex-post and were not risk adjusted.
Finland: The Finnish DPS is both funded and managed by banks. The
intervention powers of the deposit insurance authority, moreover, are
particularly limited; it only controls the reimbursement process. The
possible moral hazard originating from these characteristics of the DPS is
tempered by the lack of coinsurance, the ex-ante funding and the
risk-based contributions that were in place even before Directive 2014/
49/EU. The risk-adjusted premia are calibrated according to the par-
ticipating bank’s capital adequacy. In case the deposit insurance fund is
not large enough to compensate the guaranteed deposits, it has to borrow
money from the market, as no public intervention is due.
France: The insurance agency is a legally independent entity, in
France; it does not have the power to take strong action to bring a bank
back to solvency, nor revoke the DPS membership to a bank, but it can
take legal action against banks’ directors and other officials. This possi-
bility tempers the possible moral hazard incentives arising from the fact
that banks alone fund and administer the national DPS. Other than
domestic banks, both foreign banks subsidiaries and branches are
required to participate in the French DPS. Even before the respective
directives were in force, the French DPS had no formal coinsurance, the
deposit insurance reserve was funded ex-ante and the contributions were
adjusted according to the members’ riskiness, assessed on the partici-
pating banks’ insured deposits level. Over and above the post-crisis
measures put in place by the Directive 2009/14, the French DPS
enlarged the typologies of exposures and depositors covered.
Germany: In Germany, there are four different typologies of deposit

protection schemes. The statutory protection schemes are supervised by
the Deposit Guarantee and Investor Compensation Act, which imple-
ments the respective EU-Directive. Two statutory protection systems are
devoted to German commercial banks: the Entschädigungseinrichtung
deutscher Banken (EdB) for private commercial banks, and the
Entschädigungseinrichtung Öffentlicher Banken (VÖB) for public
commercial banks. There are also voluntary protection schemes that
supplement the statutorily required protection; for private commercial
banks, a voluntary deposit protection fund is set up at the Association of
German Banks, whereas for public commercial banks the corresponding
voluntary deposit protection fund is set up at the Federal Association of
Public Banks. Despite they are kept separated, both associations govern
the respective statutory deposit funds and the voluntary deposit funds.
Saving banks and cooperative banks are excluded from statutory deposit
protection, but are covered by the protection schemes of the savings bank
(DSGV) and cooperative bank sectors (BVR); therefore, claims of
depositors from saving deposits, time deposits and sight deposits are fully
covered. In this fragmented framework, the banks themselves fund and
administer the deposit insurance funds; the contributions have always
been periodical and based on the members’ risk, assessed according to the
amount of deposits of non-bank institutions. The deposit insurance
authorities, moreover, have the power to cancel the membership to the
DPS of any participating bank.
Greece: Despite the source of the DPS funding is private, banks and
public officials together manage the insurance fund in Greece. The
deposit insurance authority cannot take major action to restore a bank’s
soundness, as for instance imposing a form of temporary administration,
nor take legal action against banks director or officials, but it can revoke a
bank’s participation to the DPS. The participations to the DPS are not
only compulsory for domestic banks but also for foreign banks’ sub-
sidiaries and branches. Deposits of foreign subsidiaries of domestic banks
are, together with interbank deposits, the only form of deposit excluded
from the Greek DPS. Even before the Directive 2014/49/EU made these
features compulsory, the Greek DPS had ex-ante funding, risk-based
contributions and no coinsurance. In order to tackle the severe financial
188 M. Migliavacca
crisis that hit Greece, the amount of covered deposits was increased, the
Government guaranteed both deposits and banks’ debt, the bank’s
contributions were increased and the reimbursement period was short-
ened. As the sovereign crisis was particularly severe in Greece, a drastic
capital control was imposed in order to avoid deposits flight.
Hungary: The Hungarian DPS funding is provided solely by banks,
but the fund is jointly administered by public officers. The contributions
to the deposit insurance funds have always been regularly collected and
risk based, even before the Directive 2009/14 entered into force. The
deposit insurance authority has the legal power to revoke a bank’s
membership to the national DPS, but cannot directly intervene against
banks’ officials and directors, nor replace part of the management or
impose a form of temporary administration. The participation at the
national DPS is compulsory only for domestic banks, but deposits in
foreign currency and deposits of both foreign branches or subsidiaries of
domestic banks are included in the deposit insurance coverage.
Ireland: In Ireland, the deposit insurance agency is not a stand-alone
authority, but it is included within the national Central Bank. Another
peculiarity of the Irish DPS is that the funding is entirely provided by
banks, but the deposit insurance fund is administered by the public
sector alone. The deposit insurance authority, moreover, has the power
to heavily intervene in the banks’ management, by imposing a form of
administration, which temporarily overrides the management, by sus-
pending or replacing part of the management team and by taking legal
action against banks’ directors or officials. The DPS authority is also
allowed to revoke the deposit guarantee for any participating bank. The
participation to the national DPS is compulsory for both domestic banks
and foreign banks’ subsidiaries and branches, whereas the deposits of the
foreign branches and subsidiaries of domestic banks are excluded from
the insurance coverage, just like interbank deposits. Before respectively
Amending Directive 2009/14/EC and Directive 2014/49/EU, the
deposits were coinsured and the contributions were ex-ante but not risk
based. In case the deposit insurance fund cannot cover in full the claimed
deposits, the Government would intervene, but the deposit insurance
fund is supposed to refund it in subsequent years.
Italy: In Italy, the national DPS is both funded and managed by the
private sector alone. The agency authority has very limited powers; it is in
charge of managing the payout of the insured portions of the deposits
and has access to information collected by the banking supervisor, but it
does not have direct intervention power. The national DPS, moreover, is
not only used for depositor protection purposes, but can also provide
liquidity to banks. Even before respectively Amending Directive 2009/
14/EC and Directive 2014/49/EU, the deposits were not coinsured and
the contributions were risk based, but the Italian DPS was not prefunded
(premia were collected ex-post, when needed). Only the banking super-
visor can start the depositors’ reimbursement process.
Participation in the Italian DPS is compulsory for Italian banks and
foreign bank subsidiaries and branches, as well; foreign currency deposits
are covered by the Italian DPS, as well as deposits of foreign branches of
domestic banks, but interbank deposits and deposits of foreign sub-
sidiaries of domestic banks are not included in the guarantee scheme.
The DPS guarantees a maximum coverage of €100,000 per depositor per
institution, but before 2014, the coverage limit was the highest in the EU
€103,291.38.
Latvia: The deposit insurance agency is not a legally separate
authority, but part of the banking supervision agency. The Financial and
Capital Market Commission ensures the collection of funds, the com-
pensations’ payment and indirectly manages the Deposit Guarantee
Fund. By the end of 2010, 79.50% of the total deposit of participating
commercial banks was actually covered by the national DPS, one of the
largest percentages in the EU28. Even before the 2014/49/EU, the
collection of the contributions was ex-ante and risk based; in determining
the applicable rate, the deposit insurance authority takes into account
capital adequacy, liquidity ratio and large exposure ratios of deposit takers
as well as the quality of the loan portfolio.
Lithuania: The Lithuanian DPS is funded by banks, but administered
by the public sector only. The deposit insurance authority does not have
extensive intervention power towards the banks’ management; it only has
the possibility to have access to information collected by the banking
supervision. The Lithuanian DPS includes domestic banks and foreign
banks subsidiaries and branches and any typologies of deposits, but the
190 M. Migliavacca
interbank ones. The deposit insurance coverage is per investors per

institution, and coinsurance has never been applied. Before Directive
Directive 2014/49/EU, the contributions to the deposit insurance
schemes were not risk based, but were collected ex-ante. A court-declared
bankruptcy sentence triggers immediately the claim for payments by the
DPS, which in general are fully reimbursed within 30 working days.
Luxemburg: The private sector both funds and manages the DPS in
Luxemburg; the evident moral hazard incentive stemming from these
characteristics is partially tempered by the intervention power exerted by
the insurance agency, which can cancel or revoke a bank’s membership
to the national DPS and take legal action against banks directors or
officials. The Directive 2014/49/EU introduced a deep change, as
beforehand the collection of the fund contribution was ex-post and did
not depend on some assessment of risk. National banks and foreign
banks subsidiaries are part of the DPS in Luxemburg, which does not
include, though, foreign banks branches. The deposit insurance coverage,
moreover, does not include the deposits of the foreign subsidiaries of
domestic banks, but covers deposits of the foreign branches of domestic
banks. Finally, the deposit insurance coverage is the most generous
possible: per depositors per institution without any form of coinsurance,
even before the Amending Directive 2009/14/EC banned it.
Malta: The DPS in Malta is a legally separate entity, which is funded
by banks but managed by the public sector only. It does not have the
authority to revoke a bank’s participation to the DPS, but can take direct
legal actions against banks’ management, replace it or impose temporary
external administration, which overrides its power. Before the Directive
2014/49/EU, the Maltese DPS’s premia were not risk-based, but their
collection has always been ex-ante. All deposits from domestic banks,
foreign banks’ subsidiaries and branches are included in the national
DPS, as well as foreign currency deposits and deposits of foreign branches
of domestic banks. The only deposits that are not covered by the
Maltese DPS are the interbank deposits and the deposits of foreign
subsidiaries of domestic banks.
Netherlands: Is one of the rare cases in which the deposit insurance
authority is not a legally independent institution, but it is part of the
central bank. It is furthermore the only Member State within the EU28
in which the DPS fund is jointly funded by private and public actors, but
only public officials administer it; in this way, the possibility of moral
hazard by the banking sector’s side is virtually nullified. This may explain
why the deposit insurance authority does not need the power to directly
intervene in the banks’ management, cannot take legal actions against the
banks directors or officials and does not have the authority to revoke a
bank membership to the insurance scheme. Before 2014, the collection
of the banks’ contributions to the DPS was only ex-post and not based on
some assessment of the banks’ riskiness. Foreign banks branches are
excluded from the DPS, as well as deposits of foreign subsidiaries of
domestic banks. Whereas, foreign banks subsidiaries’ deposits, foreign
currency deposits and deposits of foreign branches of domestic banks are
covered by the Dutch DPS.
Poland: The Polish DPS is funded by banks only, but the financial
resources are jointly managed with public officials; this form of public
control is particularly significant, as the Polish DPS also provides liq-
uidity to banks when needed. The deposit insurance agency power is
rather weak, it only has on-site examination authority, it cannot directly
intervene against the banks’ management, nor take legal actions against
banks directors or officials and cannot exclude a bank from the DPS. The
Polish DPS has always had an ex-ante contribution mechanism, but prior
to the Directive 2014/49/EU, the single contributions weren’t calibrated
on the participating banks’ riskiness.
Portugal: The Portuguese DPS is funded by the participating banks;
although there has been an initial endowment by the national central
bank, the fund is managed jointly by the private and public sector. The
deposit insurance authority has access to the information collected by
the banking supervisor and is in charge of managing the payout of the
claimed deposits funds to depositors; only a banking supervisor’s state-
ment, though, triggers the depositors’ refund process. It has no direct
intervention authority. Both Amending Directive 2009/14/EC and
Directive 2014/49/EU did not dramatically change the Portuguese DPS,
as it has never had coinsurance and the banks’ contributions have always
been collected ex-ante and weighted by some assessment of the con-
tributors’ riskiness. Participation to the DPS is compulsory not only for
192 M. Migliavacca
domestic banks, but also for foreign banks’ subsidiaries and branches; the
deposit insurance coverage includes foreign currency deposits, but
interbank deposits and deposits of both foreign branches and subsidiaries
of domestic banks are left uncovered. In order to cope with the global
financial crisis, the Portuguese DPS was promptly adjusted to adapt to
the Amending Directive and the Government guaranteed new issuance
of bank debt upon request.
Romania: Only the private sector contributes to the Romanian DPS,
which, though, is jointly managed by both private and public officials.
The monitoring activity of the deposit insurance authority does not
include strong intervention powers, as it cannot exclude a bank from the
DPS, nor take legal actions against banks’ officials or directors and nei-
ther impose a temporary administration. It can, though, decide which
kind of resolution option best fits a bank’s failure, and it has access to the
information collected by the banking supervisors. Both foreign banks’
subsidiaries and branches are excluded from the Romanian DPS, which
does not cover deposits of the foreign subsidiaries of domestic banks,
either. Even before the Amending 2009/14/EC, Romanian DPS exclu-
ded any form of coinsurance. Directive Directive 2014/49/EU intro-
duced risk-based contributions but did not change the contributions’
collection procedure, which has always been ex-ante. The reimbursement
process, in the end, is triggered by both a court-declared bank bank-
ruptcy and a banking supervisor’s statement.
Slovakia: The source offunding of the DPS in Slovakia is entirely private,
but managed by both private and public officials. The intervention
authority of the deposit insurance agency is rather limited, it can only have
access to the information collected by the banking supervisor and is
responsible for the organisation of the payout mechanism of claimed
deposits, but it cannot actively intervene towards participating banks’
management nor exclude some banks from the DPS. Domestic banks and
foreign banks’ subsidiaries are required to participate to the DPS, which,
though, is not compulsory for foreign banks’ branches. Interbank deposits
and deposits of the foreign subsidiaries of domestic banks are not covered
by the Slovak DPS. There has never been formal coinsurance in Slovakia,
and the participating banks’ premia collection has always been ex-ante, even
before Directive 2014/49/EU, which introduced risk-based contribution.
Slovenia: The Slovenian supervisory design is rather centralised: DPS is

part of the national central bank, which is also the supervisory authority,
and there is no separation of the deposit insurance functions from the
other functions of the central bank. The deposit insurance fund, though,
is solely used for deposit insurance purposes. The participation in the
deposit insurance system is compulsory for domestic banks and foreign
banks’ subsidiaries, but foreign banks’ branches are excluded from the
DPS. Apart from domestic deposits and foreign currency deposits, all
other typologies of deposits (e.g. interbank deposits and deposits of the
foreign branches/subsidiaries of domestic banks) are excluded from the
insurance coverage. Before Directive 2014/49/EU was implemented,
the Slovenian contributions’ collection was ex-post and not subject to the
banks’ risk assessment. From the outburst of the global financial crisis to
31 December 2010, the Slovenian Government guaranteed unlimited
coverage for the insured deposits.
Spain: The Spanish deposit protection agency is a legally independent
authority; it coordinates the deposit insurance fund, which is not only
used for deposit protection purposes, but also for liquidity provision to
banks, when needed. The deposit insurance agency has access to the
information collected by the banking supervisor and can take legal action
against banks’ directors and officials, in case of violation of the deposit
insurance agency by-laws or regulations. It is also in charge of the claimed
deposits’ reimbursement process supervision and of the selection of
the most suitable failure resolution typology, in case of bank failure. The
deposit insurance authority cannot, though, assess the quality of the
banks’ business and balance sheets and put in place corrective measures,
such as raising premia, request improvement in the banks’ practices, nor
withdraw the insurance coverage of a bank. The deposit insurance
authority in Spain does not have intervention authority either, so it
cannot, for instance, impose a temporary external administration to a
bank or replace managers. Domestic and both foreign banks’ subsidiaries
and branches are covered by the Spanish DPS, which only exclude
interbank deposits and deposits of the foreign subsidiaries of domestic
banks from the protection system. The Spanish DPS has never had
coinsurance and ex-post contribution system, but before Directive 2014/
49/EU, the single contributions were not based on the participating
194 M. Migliavacca
banks’ riskiness. After the global financial crisis, in 2009, the

Spanish DPS was adapted according to the Amending Directive 2009/
14/EC, and a new Fund for the Orderly Restructuring with resolution
functions was established.
Sweden3: Swedish deposit insurance fund is managed solely by public
officers, excluding most possibilities of moral hazard behaviours by the
banks’ management at their roots. This might be the reason why the
deposit insurance authority has limited intervention powers; it cannot
withdraw a banks’ participation to the DPS, nor take legal actions against
banks’ management. The contributions to the deposit insurance fund
have always been ex-ante, but were not based on the participants’ riski-
ness, until Directive 2014/49/EU entered into force. The coverage is per
person and is extended to foreign currency deposits, as well.
United Kingdom: The deposit insurance agency is a legally separate
entity, whose only purpose is to protect deposits and is both funded and
administered by banks only, without any public intervention. The deposit
insurance agency powers are particularly narrow, as well: it does not have
access to the information collected by the banking supervisors, nor is able to
ask for improvements in the banks’ management in case of excessive
risk-taking. It cannot intervene directly in the banks’ management either,
nor can replace management or take legal actions against it. It cannot
impose a form of administration not even temporary, nor choose the res-
olution methodology to resolve a bank failure. It has, though, the ultimate
power to revoke deposit insurance for a participating bank. Foreign bank
branches are excluded from the DPS and so are deposits of the foreign
subsidiaries of domestic banks. Deposits of domestic banks’ foreign
branches, interbank deposits and foreign currency deposits are treated as
regular ones, so covered up to the legislative limit.
Just few common trends can be identified among the heterogeneous
DPS’ designs described in the above overview. In order to be able to
somehow compare the DPSs’ designs around EU28, the DPSs’ charac-
teristics examined have been grouped into three major clusters: the first
one comprises the coverage limit and the contributions’ characteristics
and is the one that has been mostly targeted by the legislative inter-
ventions, so it is the most homogeneous across the 28 Member States.
The coverage limit has been reasonably the first feature to be
homogenised with Directive 94/19/EC, and twenty years later, even the
contributions to the DPSs are collected in the same way across Europe
(ex-ante), and they are all currently anchored to the contributors’
riskiness.
There are other characteristics, on the contrary, that are still largely left
to the Member States’ autonomy. There are a number of different activities
that in some jurisdictions the deposit insurance agency can carry out
autonomously, such as the possibility to take major actions to bring a bank
back to solvency (e.g. to impose a form of temporary administration), the
possibility to take legal actions against a bank’s management and the
possibility to revoke a bank’s participation to the DPS; these features have
been grouped in the second cluster. Intervention powers, broadly speak-
ing, are particularly relevant in those Member States where the private
sector plays an active role in the deposit insurance fund management,
because it is an indirect form of supervision on the banks’ possible moral
hazard incentives (Barth et al. 2004). Only the deposit insurance authority
in Ireland, though, has all of the three major intervention powers.
Finally, the funding and management decisions can be grouped
together, as their combination significantly influences the possible moral
hazard incentives of the participating banks. It is reasonable to expect
that the higher the presence of public funds within the deposit insurance
funding, the higher the moral hazard incentives on the banks’ side,
especially if the deposit insurance fund is managed by the banks under a
private sector arrangement. On the contrary, the higher the private
intervention in the management of the deposit insurance fund, the
higher the possible consequent moral hazard incentive for banks’ officials.
The “fund management” feature is the characteristic that may leave more
room to moral hazard, because it defines which actor is in charge to use
the funds devoted to the deposits coverage, especially if the contribution
is ex-ante. In accordance with this interpretation (see Barth et al. 2013),
the different possible combinations of public and private intervention in
the DPS management and funding leave room to different degrees of
moral hazard incentives. Table 8.4 illustrates four simplified combina-
tions of actors that fund and manage the deposit insurance fund, iden-
tified with letters D to A, with increasing moral hazard incentives on the
banks’ side.
196 M. Migliavacca
Table 8.4 Funding and management design

Funding actors Management actors Room to potential Moral
Hazard
A Government and & Banks
jointly
B Banks & Banks
C Government and & Government and
jointly jointly
D Banks & Government and
jointly
The most critical matching is banks’ management of governmental or

mixed funds (combination A, see Table 8.4), even though from 1994 on
the typologies of possible investments shrank significantly. On the
extreme opposite situation, when banks are the only source of the DPS
funding but they either do not manage their contributions or have the
possibility to manage them only in cooperation with public officials
(combination D), the moral hazard incentives on the banks’ side is
minimum. Among the EU28 Member States, there is a pronounced
heterogeneity regarding the combination of sources of funding and
management of the deposit insurance fund. There is no DPS entirely
funded and administered by the public sector, which in a number of
cases,4 though, participates in the fund management, together with banks.
In only six cases (Finland, France, Germany, Italy, Luxemburg and UK),
banks both fund and administer the DPS; furthermore, in none of these
countries, the deposit insurance authority has strong intervention power;
only in Luxemburg, it can both take legal action against a banks’ man-
agement and withdraw a bank’s participation to the PDS.
As the characteristics comprised in the first cluster (i.e. the absence of
coinsurance, the presence of an ex-ante contribution system and of
risk-based premia) have been standardised respectively by the Amending
Directive 2009/14/EC and by the Directive 2014/49/EU, the analysis of
the differences among European DPSs focuses on the management and
funding characteristics on one hand and the extension of the intervention
power on the other one.
Figure 8.1 presents these two clusters on the vertical and horizontal
axis respectively and plots the EU28 Member States.5 On the vertical
Management and Funding
PL HR
RO EL IE
D BG CY HU BELV
LT ES
PTSKDK MT LU
C
NL
B
FI IT FRDE
UK
A AT SI
0/3 1/3 2/3 3/3 Intervention

Notes: this graph shows the combination of Intervention and Management & Funding features adopted by the current EU28 Member States
Fig. 8.1 EU28 Member States DPSs’ design.
axis, the four combinations of DPS funding and management are ordered
from A to D, according to the combinations displayed in Table 8.3; the
horizontal axis shows the number of intervention powers delegated to the
deposit insurance agency. The strictest DPS should be on the top right
corner as this quarter presents, on the vertical axis, the combination of
funding and management that leaves least space to the banks’ moral
hazard and, on the horizontal axis, the deposit insurance authorities with
the strongest intervention powers.
The countries are concentrated in the upper left quarter of the graph.
In most Member States, the DPS provides the combination of man-
agement and funding actors that should most reduce possible imprudent
behaviours from the banks’ side (private funding, managed by either
public officials alone or in a joint effort with private officials). In the vast
majority of cases, though, the intervention power left to the deposit
198 M. Migliavacca
insurance authority is particularly limited and the number of countries

gradually decreases as the number of intervention powers increases
(moving from left to right on the Figure).
The extreme cases are the most surprising ones; the combination that
leaves most freedom to banks is in Austria, where the deposit insurance
authority has no intervention power, but only supervises the claimed
deposits’ payout process and the DPS is mix funded but administered by
the banks only. In this perspective, moreover, it is quite meaningful to
point out that the deposit insurance fund can be used for other purposes
than depositors’ protection, such as liquidity provision to banks, if
needed. On the other extreme, the Irish deposit insurance authority is
included within the national Central Bank and has therefore a rather
extensive intervention power against the supervised banks, which con-
tribute ex-ante to the deposit insurance fund, but do not play any role in
the fund management, which is administered by public officials only.
From the qualitative analysis above, it can be concluded that the EU28
Member States seem to prefer to control potential moral hazard incen-
tives by leaving room to public intervention on either the funding or the
management of the deposit insurance funds, whereas the insurance
agency does not usually have major direct intervention powers, which in
most cases are left to the banking supervisor authority.
8.4 Conclusive Remarks

The European legislative framework regulating DPSs has historically left
some discretion to the Member States, so that they could adjust their
DPSs to the national peculiarities of the banking industry and supervi-
sion. In this way, though, the differences among DPSs even within the
EU28 became so stark to be perceived as unfitting in the banking union
framework. The Directive 94/19/EC and its subsequent modifications
have gradually homogenised the DPSs across the EU, by acting firstly on
the coverage limit and then on increasingly more specific features, such as
the repayment time, the coinsurance and finally on the contributions
collection methodology and on risk-based premia, with Directive 2014/
49/EU. Despite the remarkable effort spent in the last two decades to
harmonise the DPSs around Europe, though, there is still scope for
improvement, even because the urge for a unified European Deposit
Insurance Scheme (EDIS) is becoming increasingly more pressing.
Immediately after the outburst of the global financial crisis, national
DGSs in effect proved to be rather vulnerable to large local shocks.
Moreover, significant differences among national DGSs can contribute to
market fragmentation by affecting the ability and willingness of national
banks to expand their businesses cross border.
A strong heterogeneity emerges from the analysis of the single DPSs
within the EU28. At least three major groups of features shape each
national DPS design: there are characteristics strictly related to the
coverage, such as the amount, the risk weighting of the contributions and
the collection methodology, which have been homogenised across
Europe between 2009 and 2014. A second group of features regards the
typology of actors who fund the DPS and manage its financial assets;
whereas a third group of features controls for the intervention powers left
to the deposit insurance authority. As the first cluster has been legally
homogenised, the analysis focuses on the combination of the other two
and their possible role in leaving space to potential moral hazard on the
banks’ side. From a qualitative analysis, the EU28 Member States seem
to prefer to control these potential moral hazard incentives by leaving
room to public intervention on either the funding or the management of
the deposit insurance funds, whereas the insurance agency does not
usually have major direct intervention powers.
In order to reduce the deep diversities the national DPSs present,
initial incentives towards a unified DPS were included in the Directive
2014/49/EU, as the Legislator opens up the way to the possibility to
“merge the DGSs of different Member States or to create separate
cross-border schemes on a voluntary basis”, as a tangible answer to the de
facto integration of the European banking industry. Even the possibility
for national DPSs to borrow from one another, as an extreme funding
source, encourages the creation of networks that do not require formal
changes in the national legal frameworks, but still contribute to put in
place a unified European DPS. The EDIS would provide a more uniform
degree of insurance coverage for all European depositors, ensuring that
their level of confidence in a bank’s ability to refund their savings would
200 M. Migliavacca
not depend on mere geographical reasons. Furthermore, EDIS would

increase the resilience of the banking sector against systemic and national
crises within the EU, leading the way towards an increasingly more
interconnected Banking Union.
Notes
1. Eligible deposits: deposits repayable by the guarantee scheme under
national laws, before the level of coverage are applied.
Covered deposits: deposits obtained from eligible deposits when applying
the level of coverage provided for in your national legislation.
2. Sweden did not participate to the latest World Bank’s Bank Regulation
and Supervision Survey.
3. Sweden did not take part in the latest Bank Regulation and Supervision
survey, so the information provided are taken from the World Bank’s
Bank Regulation and Supervision survey (2007).
4. Belgium, Bulgaria, Cyprus, Denmark, Estonia, Greece, Hungary, Poland,
Portugal, Romania, Slovakia and Spain.
5. Estonia and Sweden are excluded from the graph, because they did not
disclose enough information in the World Bank’s Bank Regulation and
Supervision Surveys, 2012.
References
Acharya, V.V. 2009. A theory of systemic risk and design of prudential bank
regulation. Journal of Banking & Finance 5 (3): 224–255.
Anginer, D., A. Demirgüc-Kunt, and M. Zhu. 2014. How does deposit
insurance affect bank risk? Evidence from the recent crisis. Journal of Banking
and Finance 48: 312–321.
Angkinand, A., and C. Wihlborg. 2010. Deposit insurance coverage, ownership,
and banks’ risk-taking in emerging markets. Journal of International Money
and Finance 29 (2): 252–274.
Arnaboldi, Francesca. 2014. Deposit Guarantee Schemes. In Deposit Guarantee
Schemes: A European Perspective, 50–85. UK: Palgrave Macmillan.
Barth, J.R., G. Caprio, and R. Levine. 2004. Bank regulation and supervision:
What works best? Journal of Financial intermediation 13 (2): 205–248.
Barth, J.R., G. Caprio Jr., and R. Levine. 2013. Bank regulation and supervision
in 180 countries from 1999 to 2011. Journal of Financial Economic Policy
5 (2): 111–219.
Cariboni, Jessica, Elisabeth Joossens, and Adamo Uboldi. 2010. The prompt-
ness of european deposit protection schemes to face banking failures. Journal
of Banking Regulation 11 (3): 191–209.
Carbo-Valverde, S., E.J. Kane, and F. Rodriguez-Fernandez. 2012. Regulatory
arbitrage in cross-border banking mergers within the EU. Journal of Money,
Credit and Banking 44 (8): 1609–1629.
Cordella, Tito, and Eduardo Levy Yeyati. 2002. Financial opening, deposit
insurance, and risk in a model of banking competition. European Economic
Review 46 (3): 471–485.
Demirgüç-Kunt, A., and H. Huizinga. 1999. Market discipline and financial
safety net design. World Bank Policy Research Paper 2183.
Demirgüç-Kunt, Asli, and Enrica Detragiache. 2002. Does deposit insurance
increase banking system stability? An empirical investigation. Journal of
Monetary Economics 49: 1373–1406.
Demirgüç-Kunt, Asli, and Edward Kane. 2002. Deposit insurance around the
globe: Where does it work? Journal of Economic Perspectives 16 (2): 175–195.
Demirgüç-Kunt, A., and H. Huizinga. 2004. Market discipline and deposit
insurance. Journal of Monetary Economics 51 (2): 375–399.
Demirgüç-Kunt, A., E. Kane, and L. Laeven. 2015. Deposit insurance around
the world: A comprehensive analysis and database. Journal of Financial
Stability 20: 155–183.
Diamond, Douglas W., and Philip H. Dybvig. 1983. Bank runs, deposit
insurance, and liquidity. The Journal of Political Economy: 401–419.
Directive 2009/14/EC of the European Parliament and of the Council of
11 March 2009 [2009] L68/3.
Directive 2014/49/EU of the European Parliament and of the Council of
16 April 2014 on deposit guarantee schemes [2014] L 173/149.
Duan, J.C., A.F. Moreau, and C.W. Sealey. 1992. Fixed-rate deposit insurance
and risk-shifting behavior at commercial banks. Journal of Banking and
Finance 16: 715–742.
EBA. 2015. Guidelines on methods for calculating contributions to deposit
guarantee schemes, EBA/GL/2015/10, 22.09.2015.
Engineer, M.H., P. Schure, and Mark Gillis. 2013. A positive analysis of deposit
insurance provision: Regulatory competition among European Union
countries. Journal of Financial Stability 9 (4): 530–544.
202 M. Migliavacca
Forssbaeck, J. 2011. Ownership structure, market discipline, and banks’

risk-taking incentives under deposit insurance. Journal of Banking & Finance
35 (10): 2666–2678.
Gropp, R., and J. Vesala. 2004. Deposit insurance, moral hazard and market
monitoring. Review of Finance 8: 571–602.
Hovakimiam, A., E.J. Kane, and L. Laeven. 2003. How country and safety-net
characteristics affect bank risk-shifting. Journal of Financial Services Research
23 (3): 177–204.
IADI. 2013a. Enhanced guidance for effective deposit insurance systems:
Deposit insurance coverage, Basel.
IADI. 2013b. Enhanced guidance for effective deposit insurance systems:
Mitigating Moral Hazard, Basel.
IADI. 2013c. General guidance on early detection and timely intervention for
deposit insurance systems, Basel, 2013.
IADI. 2014. IADI Core principles for effective deposit insurance systems, Basel.
IADI. 2015. Enhanced guidance for effective deposit insurance systems: Ex Ante
funding, Basel.
Kane, Edward J. 1995. Three paradigms for the role of capitalization
requirements in insured financial institutions. Journal of Banking &
Finance 19 (3): 431–459.
Laeven, L. 2002. Bank risk and deposit insurance. World Bank Economic Review
16: 109–137.
Lambert, C., F. Noth, and U. Schüwer. 2015. How do insured deposits affect
bank risk? Evidence from the 2008 Emergency economic stabilization act.
Journal of Financial Intermediation.
Pennacchi, G. 1987. A Reexamination of the over- (or under-) pricing of
deposit insurance. Journal of Money, Credit and Banking 19: 340–360.
Pennacchi, G. 2006. Deposit insurance, bank regulation, and financial system
risks. Journal of Monetary Economics 53: 1–30.
Schich, S. 2008. Financial crisis: Deposit insurance and related financial safety
net aspects. Financial Market trends 2: 76–121.
Wheelock, D.C., and P.W. Wilson. 1995. Explaining bank failures: Deposit
insurance, regulation, and efficiency. The Review of Economics and Statistics:
689–700.
9
A Technical Approach to Deposit
Guarantee Schemes
Francesca Arnaboldi
9.1 Introduction
Progress towards a common European financial framework has been a
constant trend over the past forty years, with ongoing harmonization of
national legislation and practices. The financial sector has played a key
role in the integration of the European countries. Indeed, financial
integration has been enhanced by the introduction of a single currency.
Despite the positive achievements in the integration of European
financial markets and economies, the financial crisis confirms that closer
coordination of prudential policies and safety nets is required. The
European financial system has revealed more fragile than expected. The
crisis meant a serious setback for financial integration and the possibility
of the break-up of the single currency.
F. Arnaboldi (&)
University of Milan, Milan, Italy
e-mail: farnaboldi@unimi.it
204 F. Arnaboldi
As regards the European retail banking markets, the financial crisis

illustrated once more how banks are susceptible to the risk of bank runs
and the need of a coordinated supervision at European level.
Deposit guarantee schemes help preventing such risk, through the
reimbursement of a limited amount of deposits to depositors whose bank
has failed.
Directive 2014/49/EU set a uniform level of protection for depositors
throughout the European Union (EU), thanks to a broadened and
clarified scope of coverage, faster repayment periods, improved infor-
mation and robust funding requirements. However, it did not establish
the third pillar of the Banking Union, a European deposit insurance
scheme (EDIS). In the first moment, it was decided to delay its creation
and to opt instead for a harmonized network of national deposit guar-
antee schemes (DGSs).1
In 2015, progress towards the EDIS accelerated. The Five Presidents’
Report (President of the European Commission, in close cooperation with
the President of the Euro Summit, the President of the Eurogroup, the
President of the European Central Bank and the President of the European
Parliament) was published in July 2015 (Juncker et al., 2015). It sets out an
ambitious programme of measures to underpin the economic and mone-
tary Union, among which is the European Deposit Insurance
Scheme (EDIS). It will be applied alongside the Single Supervisory
Mechanism (SSM) and the Single Resolution Mechanism (SRM) and
funded by risk-based contributions from banks operating in the Banking
Union countries (FITD 2016).
In May 2015, in order to ensure consistent application of Directive
2014/49/EU and to provide incentives to banks to operate under a less
risky business model, the European Banking Authority (EBA) issued
guidelines to specify methods for calculating the contributions to DGS.
In a context where many member states did not have pre-financed DGS,
EBA set out principles for technically sound methods for calculating
contributions to ensure that costs of deposit insurance are borne pri-
marily by the banking sector (EBA 2015).
The European Commission, in fulfilling a commitment, published in
November 2015 a proposal for legislation, which sets out a euro
area-wide deposit insurance scheme for bank deposits and further
measures to reduce remaining risks in the banking sector in parallel

(European Commission 2015a). The legislative proposal proceeds
through three successive stages: a reinsurance scheme for participating
national DGSs in the first period of 3 years, a co-insurance scheme for
participating national DGSs in the second period of 4 years, and full
insurance for participating national DGSs in the steady state, which starts
in 2024 (European Commission 2015b).
Within this framework, EBA guidelines offer a basis on which to assess
progress in the convergence of national practices in calculating contri-
butions to DGSs.
In this chapter, we take advantage of the EBA guidelines and study
whether Italian banks would be negatively affected by their implemen-
tation, fuelling systemic risk, as opined by some member countries.
“Germany, the EU’s biggest economy, does not want its depositors to be
liable for payouts in the event of bank failures elsewhere. It insists the EU
must first take steps to minimise risks before starting talks on shared
responsibility. Berlin insisted that any reference to setting up such a
deposit scheme be removed at the EU summit in October, and has
succeeded in doing so again at the December meeting” (Reuters 2015).
Specifically, this chapter investigates the system of calculating risk-based
contributions to DGS currently in use in the Italian banking system and
compares this to the system promoted by EBA, using a sample of 172 out of
202 member banks, 85% of the population of the Fondo interbancario di
tutela dei depositi (FITD). Using Bankscope data from 2012, when the single
supervisory mechanism was established, to 2014, we examine the impact of
the EBA system on the classification of Italian banks among risk categories
and, subsequently, on the contributions banks have to pay to DGS.
We find that EBA proposal would increase the number of banks in the
lower-risk classes, where contribution quota to the DGS would remain
unchanged or would decrease.
This chapter contributes to the literature on banking supervision by
investigating the third pillar of the Banking Union, that is, deposit
guarantee schemes, a matter of which the use of information has been
limited in order to prevent such use from affecting the stability of the
banking system or depositor confidence (Directive 2014/49/EU art.16
c.5).2 In particular, the main contribution lies in the comparison of the
206 F. Arnaboldi
two methodologies mentioned above and in the prediction of the EBA

algorithm’s effect on Italian banks contributions. The analysis may have
significant policy implications, as it forecasts the future contributions of
Italian banks providing an empirical evidence that should reassure about
the possible Italian banks’ moral hazard.
This investigation shows some caveats: in principle, the FITD uses
semi-annual or quarterly data, whereas Bankscope reports annual data.
Secondly, for some ratios, it is not possible to match data as described by
the FITD documents to data in Bankscope.
However, uncertainty about the real exposure of depositors to bank
failures impairs the relationship with clients and with other member
states. Therefore, we believe that the FITD, which is the only institution
with access to real data, should provide additional information on this
relevant topic.
The rest of the chapter is organized as follows. Section 9.2 provides
the framework for deposit guarantee schemes in Italy. Section 9.3
analyses the system of calculating risk-based contributions established by
the FITD. Section 9.4 applies EBA guidelines to the same sample of
domestic banks, using both core and additional ratios and the buckets
method. Section 9.5 compares the two systems, and Sect. 9.6 concludes.
9.2 Deposit Guarantee Schemes in Italy

The Fondo interbancario di tutela dei depositi (Interbank Deposit
Protection Fund) is a private-law consortium established in 1987 on a
voluntary basis, which has since become a mandatory Fund (FITD
2016). Bank participation in a deposit guarantee scheme became
mandatory in 1996 with the transposition of the first Directive on
Deposit Guarantee Systems, 94/19/EEC, in the Italian legislation. The
second DGS was created in Italy in 1997, the Fondo di garanzia dei
depositanti del credito cooperativo, which covers mutual banks and
replaced the Fondo centrale di garanzia, created in 1978 to guarantee
deposits in rural and cooperative banks (Senato 2015). Thus, all Italian
banks are members of the FITD, except for mutual banks and branches
of non-EU banks authorized in Italy if they already participated in an
equivalent scheme in their home country. Italian branches of EU banks

also may adhere to FITD, in certain cases, to top-up their home guar-
antee coverage.
FITD guarantees the deposits in the member banks, which provide
the financial resources for FITD to accomplish its mission. The Fund
conducts a variety of interventions in favour of member banks when they
are under compulsory administrative liquidation, in resolution or in
special administration. Pursuant to art. 96-ter of the Legislative Decree
385/1993 (Italian banking Law), the Bank of Italy exercises specific
powers of oversight on the deposit guarantee systems.
Today, FITD is regulated by Directive 2014/49/EU and, as a result, it
undergoes many changes. These include, among others: (1) the passage
from an ex post to an ex ante system of payment of contributions to the
scheme; (2) the investment of available financial resources; (3) the
reduction to seven working days of the deposit payout time, presently
established within 20 working days from the date the compulsory
administrative liquidation takes effect, by the end of the year 2023;
(4) calculation of banks’ risk-based contributions, following EBA
guidelines; and (5) use of the Fund’s resources for a wide variety of
measures, alternative to direct reimbursement (FITD 2016). In this
context, the FITD began raising ex ante contributions in December
2015 to avoid an excessive burden in the following financial years given
the obligation to reach the target level by the year 2024.
9.3 The FITD’s Monitoring System of Bank

Riskiness
9.3.1 Balance Sheet Indicators
The Fund has in place a monitoring system to measure and control member
banks’ riskiness. This system works through balance sheet indicators on
four different risk profiles: asset quality, solvency, liquidity and profitability
(FITD 2012). The reporting frequency is semi-annual or quarterly,
depending on the specific source of data of the Bank of Italy.
208 F. Arnaboldi
Five ratios are computed to measure the four risk profiles: A1, P, L,
D1 and D2.
According to FITD (2016), at the end of 2015, member banks were
202. Fourteen banks have been dropped since they do not report data on
Bankscope, a Bureau Van Dijk database, and sixteen do not report
enough data to compute any ratios over the 2012–2015 sample period.
The final sample is, thus, formed by 172 member banks, 85% of the
population of member banks to the FITD.3 The sample period starts in
2012 when it was decided to establish a single supervisory mechanism
(SSM) and ends in 2014, because of the paucity if data in 2015.
As previously mentioned, this analysis shows some caveats: in prin-
ciple, the Fund uses semi-annual or quarterly data, whereas Bankscope
reports annual data. Secondly, for some ratios, it is not possible to match
data from the Bank of Italy to data in Bankscope. To avoid confusion,
the rest of the chapter uses the ratio definitions provided by the Fund
(FITD 2012).
The first ratio (A1) measures the capacity of a bank to absorb potential
losses without risk of insolvency, and it is given by the ratio of bad debts
to supervisory capital (FITD 2012). To compute the asset quality ratio
A1, total impaired loans are used. According to Bankscope, total
impaired loans are the total value of the loans that have a specific
impairment against them. The Fund uses bad loans, that is, loans which
will be never repaid, even if this status has not been proved yet in court
(Bank of Italy 2016). The computed ratio is, therefore, higher, overes-
timating the risk of the bank.
P provides a measure of bank’s capital: according to the Fund, it is the
ratio of supervisory capital (including tier 3) minus total capital
requirements to risk-weighted assets. The solvency ratio P is not com-
puted since the FITD does not provide clear information on
risk-weighted assets, preventing a match to Bankscope data.
The liquidity ratio L measures the structural liquidity of the bank
dividing receivables from clients by an aggregate given by the sum of
payables from clients, circulating bonds and structured payables from
clients and bonds at fair value. The Fund does not specify whether
receivables from clients include impairments or not, so both specifica-
tions have been computed. Furthermore, the denominator is an
aggregate, and it does not have a match in Bankscope; the Fund does not
provide a list of the components and of their maturities. Therefore, two
components have been used: (1) total deposits, money market and
short-term funding which includes total customer deposits, deposits from
banks and other deposits and short-term borrowings; (2) trading liabil-
ities, that is, short positions, repos, short-term notes and other liabilities
classified at fair value. The computed ratio could under- or overestimate
the Fund ratio, which however cannot be estimated.
The fourth risk profile has two ratios: D1 is given by operating
expenses to gross income, and it shows whether gross income covers the
cost of core banking activity and/or the ability of the bank to meet
extraordinary expenses. It does not show critical issues.
D2 measures loan losses on profit before tax. It is computed only if
both numerator and denominator are positive; otherwise, it takes the
value of zero or four (only if numerator is positive and denominator is
negative) (FITD 2012). D2 is computed using total impaired loans and
pre-tax profit; once more the estimate is larger than the value provided by
the Fund, thus underestimating the true member bank’s efficiency.
To better appreciate the pros and cons of the present analysis, the balance
sheet ratios measured by the Fund are now compared to the estimated
values for years 2012 and 2013 (FITD 2012). Differences can be explained
by: (1) the use of proxies, since not all data used by the Fund is publicly
available; (2) the frequency of data, semi-annual for the Fund measures,
annual in the present estimation. As a consequence, the Fund ratios are the
median values of three observations (June and December 2012, June
2013), while the present estimation uses year-end data.
Comparing June 2012 with June 2013, there was a slight worsening in
A1 (+18%, from 18.01 to 21.18%), and in D2 (from 37.22 to 50.67%),
similar to the change computed for A1 (+21%, from 93 to 119%) and
for D2 (from −957 to 1002%), reported in Table 9.1.
Over the same period, there was a slight improvement in the median
value of the liquidity ratio (−7.45% points, from 91.73 to 84.18%) and
in the profitability ratio D1 (−1.64% points, from 68.17 to 66.53%)
(FITD 2012). As for the liquidity ratio, Table 9.1 shows similar trends
using both gross and net receivables, but net ratio is preferred since it
provides closer estimates (from 92 to 83%, versus from 97 to 88%).
210 F. Arnaboldi
Table 9.1 Balance sheet ratio computed from Bankscope

2012 2013 2014
Ratio # observations Mean # observations Mean # observations Mean
A1 149 0.93 148 1.19 148 1.99
D1 165 0.66 164 0.70 161 0.43
D2 151 −9.57 152 10.02 155 29.69
L 167 0.97 164 0.88 162 0.90
Lnet 167 0.92 164 0.83 162 0.81
Source Own computation on Bankscope’s data
D1 is the only ratio which is different from the Fund data, and it
increases from 66 to 70% (Table 9.1).
9.3.2 Thresholds, Classes and Coefficients
To assess bank’s risk, the Fund sets four thresholds per each ratio, which
correspond to five classes. FITD assigns a coefficient to each class
(Table 9.2).
According to the Fund, the sum of the coefficients of each ratio defines
an aggregate indicator (AI) ranging from 0 to 24 (Table 9.3). Since this
chapter does not compute P ratio, the aggregate indicator varies from 0 to
20. The aggregate indicator is grouped in clusters, and each cluster
corresponds to a statutory position. If the AI is lower than 3.5, the
corresponding statutory position for the bank is “low risk”, that is the
bank is classified as a low-risk bank according to the Fund rules. To avoid
distortions due to the fact that AI ranges from 0 to 20 and not from 0 to
24 as stated by the Fund, in this chapter the scale of AI has been changed
proportionally.
Figure 9.1 shows the distribution of the sample banks and of the
coefficients for each ratio (A1, L, D1 and D2) over the 2012–2014
period. Looking at A1 and D2, 72 and 89.6% of banks, respectively,
show the highest coefficient (which equals to eight for A1 and to four for
D2) and thus belong to the riskiest class. Conversely, investigating L and
D1, 0.72 and 8.32% of banks, respectively, belong to the riskiest class.
9
Table 9.2 Thresholds, classes and coefficients

A1 L D1 D2
Risk classes Thresholds Coefficients Thresholds Coeff. Thresholds Coeff. Thresholds (%) Coeff.
(%) (%)
Low risk <10 0 <90 0 <60 o 0 0–20% o 0
numerator = 0 numerator <=0
Medium-low 10–20 1 90–100 0.5 60–70 0.5 20–40 0.5
Medium 20–30 2 100–130 1 70–80 1 40–50 1
Medium-high 30–50 4 130–200 2 80–90 2 50–60 2
High risk >50 8 >200 4 >90 o 4 >60 o 4
denominator < 0 denominator < 0
Source FITD (2012)
A Technical Approach to Deposit Guarantee Schemes
211
212 F. Arnaboldi
Table 9.3 Statutory position, aggregate indicator and scaled aggregate indicator
Statutory position Aggregate indicator (*) Scaled aggregate indicator(**)
Low risk 0–3.5 0–2.9
Medium-low 3.5–6.5 2.9–5.4
Medium 6.5–8 5.4–6.7
Medium-high 8–10.5 6.7–8.8
High risk 10.5–14.5 8.8–12
Expulsion >14.5 >12
Source FITD (2012)
Note (*) Upper bounds are included
(**) The aggregate indicator has been scaled to take into account that AI ranges
from 0 to 20 rather than from 0 to 24
180
160
140
120
Banks (%)
100 Liq
80
D2
60
D1
40
20 A1
0
0
0.5
1
2
4
Coefficients
Fig. 9.1 Banks (%) and coefficients Source Own computation on Bankscope’s
data. Note As far as ratio A1 is concerned, coefficients are 0, 1, 2, 4 and 8
In fact, 47 and 39% of banks score a coefficient equal to zero, and the
lowest risk. Sample banks seem less risky under the liquidity and prof-
itability profiles.
The scaled aggregate indicator is computed summing up the coeffi-
cients for each of the four ratios. Then, according to Table 9.3, the
statutory position is assigned to each bank of the sample in each year.
Table 9.4 investigates the year-to-year statutory position, showing a
migration of banks from the highest to the lowest risk position.
Table 9.4 Year-to-year statutory position

Year 2012 2013 2014 2012/2014
Statutory # % # % # % change (%)
position banks banks banks
Low risk 13 8.9 14 9.46 15 10.2
Medium-low 4 2.74 6 4.05 9 6.12 40
Medium 10 6.85 10 6.76 5 3.4
Medium-high 10 6.85 9 6.08 8 5.44 −35
High risk 15 10.27 17 11.49 28 19.05 85
Expulsion 94 64.38 92 62.16 82 55.78 −13
Total 146 100 148 100 147 100
Over 2012–2014, banks classified in the low and medium-low statutory

position increased by 40%; conversely, banks classified in the medium
and medium-high risk decreased by 35%.4 Likewise, banks belonging to
the high-risk statutory position increased by 85%, whereas banks clas-
sified in the expulsion position decreased by 13%.
As previously mentioned, two caveats apply to the analysis: (1) the
statutory position computed in this chapter does not include the P ratio
and it is, therefore, incomplete; (2) some proxies have been used to
compute the ratios, since actual data are not publicly available.
9.3.3 Contribution Quotas
After calculating the statutory position, the Fund computes the pro-
portional quota of the contribution base which is given by the individual
contribution base over the total reimbursable funds. Two correction
methods, the regressive mechanism and the weighted average aggregate
indicator (WAAI), that may increase or decrease the proportional quota,
are then applied (FITD 2012).
The regressive correction method modifies the proportional quota
according to the size of the bank: bigger banks get a reduction in the
proportional quota, while the smaller ones get an increase.5
The second correction method is related to the value of the aggregate
indicator, linking contributions to bank riskiness. The WAAI is
214 F. Arnaboldi
computed on the last three semi-annual ratios submitted by the bank to

the Fund: each semi-annual ratio receives a weight, which is larger the
closer in time the ratio is. The weight is four for the closest ratio, two for
the middle one and one for the ratio, which refers to the earliest time.
The WAAI is given by the following formula:
X3
AIt weightt
WAAI ¼ P ð9:1Þ
t¼1
weight
where
t 1, 2 and 3; semi-annual reports

AI semi-annual aggregate indicator
Weight 1, 2 and 4 if the AI refers to semester 1, 2 or 3, respectively
According to the Fund, when the WAAI is greater than 3.5, the bank’s
contribution quota shall be increased, proportionally to the WAAI value;
when it is greater than zero and less than or equal to 3.5, the bank shall
retain its contribution quota unchanged; if the WAAI is equal to zero,
the bank shall benefit from a reduction in its contribution quota, linked
to the total amount of increases. To account for the change of scale of AI,
this chapter uses 2.9 rather than 3.5 as threshold.
9.3.4 The Weighted Average Aggregate Indicator

and the Sample Banks
WAAI is computed applying (1) to the sample banks with the goal to
determine any changes in the statutory position of banks due to risk.
Since Bankscope reports annual data, only two observations are used.
The weights are, therefore, equal to one, for the AI of the previous year,
and to four for the most recent AI.
The denominator in (1) is equal to seven, given by the sum of the
weights, whereas in (2) is equal to five, since the weights are now only
two. As a consequence, in this investigation, the WAAI is given by:
1 AIt1 þ 4 AIt
WAAI computed in t ¼ ð9:2Þ
5
Table 9.5 part (a) shows banks with WAAI equal to zero, that would
benefit from a reduction in contribution quota. Over 2012–2014 period,
five banks would benefit from such reduction. One bank shows a WAAI
equal to zero in two years and another one in one single year. Table 9.5
part (b) shows those banks with WAAI greater than zero and less than or
equal to 2.9. Those banks would retain their contribution quota
unchanged. Once more, the same banks recur over years: six banks
belong to the group for 3 years out of three and one bank recurs twice.
When WAAI is greater than 2.9, banks’ contribution quota shall be
increased. Over 2012–2014, the number of sample banks with WAAI
over the threshold remains almost stable (92%); within this group, the
average WAAI decreases from 12 to 11.7, the minimum value of WAAI
Table 9.5 Weighted average aggregate indicator

Bank identification number\year 2012 2013 2014 Total
(a) Banks benefitting from a reduction in contribution quota
1 1 1 1 3
50 1 1 0 2
152 0 0 1 1
155 1 1 1 3
166 1 1 1 3
Total 4 4 4 12
(b) Banks retaining contribution quota unchanged
8 1 1 1 3
21 1 1 1 3
37 0 0 1 1
42 1 1 1 3
43 1 1 1 3
48 0 0 1 1
55 1 1 1 3
89 1 0 0 1
98 1 0 0 1
161 0 1 1 2
184 1 1 1 3
Total 8 7 9 24
216 F. Arnaboldi
Table 9.6 Changes in statutory position after risk adjustment—all years

SP/RASP 1 2 3 4 5 6 Total
Low risk—1 24 1 0 0 0 0 25
Medium-low—2 0 12 1 0 0 0 13
Medium—3 0 1 13 1 0 0 15
Medium-high—4 0 0 0 13 1 0 14
High risk—5 0 0 0 2 24 18 44
Expulsion—6 0 0 0 0 2 167 169
Total 24 14 14 16 27 185 280
Changes SP—RASP −1 1 −1 2 −17 16
remains stable at 4, and the maximum value decreases from 18 to 17

(−6%)6. Investigating the adjustment to bank riskiness, less banks would
have their contribution quota increased, and the average increase in
contribution quota would be lower, thus showing a safer risk profile for
the banks under scrutiny.
Applying thresholds in Table 9.3 to the weighted average aggregate
indicator, a risk-adjusted statutory position (RASP) can be computed.
Table 9.6 investigates the changes in the statutory position when the risk
adjustment is applied. Minor changes involve five banks. In particular,
one bank moves from the low-risk statutory position (SP equal to one) to
medium-low risk-adjusted scaled statutory position (RASP equal to two);
one from SP equal to two to three when risk adjustment is applied;
medium risk statutory position (SP equal to three) receives this bank but
looses one bank which moves to RASP equal to four. The medium-high
RASP is now formed by 16 rather than 14 banks. Major change involves
the riskiest clusters: 16 banks move from high to expulsion RASP.
9.4 The EBA’s Monitoring System of Bank

Riskiness
9.4.1 Risk Indicators
This section is based on EBA guidelines on methods for calculating

contributions to DGSs (EBA 2015). EBA defines core and additional
indicators, as they belong to one of the following risk categories: capital,

liquidity and funding, asset quality, business model and management,
potential losses for the DGS. This chapter describes only the indicators
used in the empirical investigation and refers to EBA (2015) for further
details on all risk categories and indicators.
EBA guidelines have been applied to a sample of 171 Italian banks,
member of the FITD. From the initial 202 member banks, 14 banks
have been dropped because of the lack of data on Bankscope, as we did in
the previous analysis on the FITD system. In addition, the year 2015 has
been dropped, because only 22 banks out of the remaining 188 (12% of
the sample) report data to compute EBA indicators. Seventeen more
banks have been excluded because they do not report enough data to
compute any EBA indicator over the 2012–2014 sample period.
Table 9.7 shows descriptive statistics on core and additional indicators
for the final sample of 171 banks.
Table 9.7 Descriptive statistics—EBA core and additional indicators

Indicators Number of Mean Standard Minimum Maximum
observations deviation
Core
Leverage ratio 413 0.09 0.07 0.01 0.95
CET1 470 0.19 0.18 0.01 2.98
Capital 472 3.52 3.65 0.11 66.22
coverage ratio
(%)
Liquidity ratio 494 0.17 0.19 0.00 0.97
NPL ratio 468 0.06 0.05 0.00 0.64
Return on asset 318 0.01 1.25 –7.37 4.65
(%)
RWA to total 470 0.55 0.19 0.07 1.33
asset
Additional
Return on 318 0.04 17.30 −115.48 49.34
equity (%)
Total asset 321 0.08 0.34 −0.89 3.80
growth
Cost income (%) 482 65.78 44.26 9.05 895.25
218 F. Arnaboldi
For the first risk category (capital), EBA proposes two core indicators:
leverage ratio, defined as tier 1 capital to total asset ratio, and capital
coverage ratio (actual to required CET1 ratio) or common equity tier 1
ratio (common equity tier 1 capital to risk-weighted assets). Capital
indicators reflect the level of loss-absorbing capacity of the bank. Higher
amounts of capital show that the bank has a better ability to absorb losses
internally, thus decreasing its likelihood of failure. Therefore, banks with
higher values of capital indicators should contribute less to the DGS
(EBA 2015). In the sample, the leverage ratio is on average 9% and
CET1 19%. Due to available information, the numerator of CET1 is tier
1 capital and not common equity tier 1 capital. Thus, the computed
leverage ratio overestimates the EBA ratio, underestimating the level of
risk.7 Similar considerations can be drawn on the capital coverage ratio,
which average is equal to 3.52%.8
For the liquidity and funding category, the two core indicators sug-
gested by the authority (liquidity coverage ratio—LCR—and net stable
funding ratio—NSFR) cannot be applied until their definition as
determined in Regulation (EU) No 575/2013 is fully operational. As a
transitional indicator, the liquidity ratio (LR) defined as liquid assets to
total assets is computed. It measures the bank’s ability to meet its
short-term debt obligations as they become due. The higher the ratio, the
larger the safety margin to meet obligations and unforeseen liquidity
shortfalls. Indeed, low liquidity levels indicate the risk that the institution
may be unable to meet its current and future, expected or unexpected,
cash-flow obligations and collateral needs. Liquid assets cover 17% of
total assets on average. In 6 banks, LR is close to zero indicating possible
future liquidity tensions (LR below 0.010 for 1 bank in 2012, for 4
banks in 2013 and for 1 bank in 2014).
The asset quality category shows the extent to which the bank is likely
to experience credit losses. Large credit losses may cause financial
problems that increase the likelihood of failure, therefore justifying higher
contributions to the DGSs. This category includes the non-performing
loan (NPL) ratio, given by non-performing loans to total assets. It pro-
vides an indication of the type of lending the bank engages in. A high
degree of credit losses in the loan portfolio indicates lending to high-risk
customers. The NPL ratio is on average 6%. Twenty banks out of 171
(12% of the sample) show a ratio higher than 15% in one or more years
(Two banks in 2012, four banks in 2013—of which one already over the
threshold in 2012—and 15 banks in 2014). Among those banks, 1 bank
has a NPL ratio larger than 50% in 2014 and another one larger than
50% in the same year. These two latter banks have a high degree of credit
losses in the loan portfolio, which increases the likelihood of failure.
Business model and management takes into account the risk related to
the bank’s current business model and strategic plans, and reflects the
quality of internal governance and controls. Business model indicators
can, for instance, include indicators related to profitability, balance sheet
development and exposure concentration. The first core indicator pro-
posed by EBA is risk-weighted assets to total assets ratio, which indicates
the kind of risky activities a bank engages in. A higher value indicates
higher risk. The second core indicator is return on asset (ROA).
A business model which is able to generate high and stable returns
indicates lower risk. However, unsustainably high levels of ROA also
indicate higher risk (EBA 2015). In the sample, RWA to total assets ratio
is 55% on average, but it is larger than 100% for three banks in 2012
and in 2013, raising doubts about the sustainability of the business
model. ROA is on average equal to 0.01%.9 Fifty three banks have a
negative value of ROA in 2013 and 54 banks in 2014 (about 32% of the
sample). The maximum value of ROA in the sample is 4.65%, and it
does not seem unsustainably high.
The last risk category is potential losses for the DGS. EBA (2015)
suggests one core indicator (unencumbered assets to covered deposits)
which measures the degree of expected recoveries from the bankruptcy
estate of the bank, which was resolved or put into normal insolvency
proceedings. A bank with a low ratio exposes the DGS to higher expected
loss. However, the proposed definition of unencumbered asset does not
allow to compute the ratio.10
In addition to the core risk indicators, DGSs may include additional
risk indicators that are relevant for determining the risk profile of
member banks. The additional risk indicators should be classified into
the above-listed risk categories. EBA proposes indicators for the asset
quality, business model and management and potential losses for the
DGS categories. In this chapter, three additional indicators belonging to
220 F. Arnaboldi
the business model and management category are applied: (1) excessive
balance sheet growth ratio (TAG) that measures the growth rate of the
bank’s balance sheet. Unsustainably, high growth might indicate higher
risk; (2) return on equity (ROE), which measures the ability to generate
profits to shareholders from the capital these have invested in the bank.
A business model which is able to generate high and stable returns
indicates reduced likelihood of failure. However, unsustainably, high
levels of ROE indicate higher risk; (3) cost to income ratio (CI) which
measures cost efficiency. An unusually high ratio may indicate that the
institution’s costs are out of control, especially if represented by the fixed
costs (i.e. higher risk). A very low ratio may indicate that operating costs
are too low for the institution to have the required risk and control
functions in place, also indicating higher risk (EBA 2015).
The mean of the sample for total asset growth is 8%. However, 99 banks
over 171 (58% of the sample) have a negative asset growth at least in 1 year
(72 banks in 2012 and in 2013—of which 45 banks are common to both
years); four banks have a TAG ratio larger than 100% (two banks in 2013
and in 2014). Among those four banks, one has a ratio larger than 200% in
2014 and one larger than 300% in 2013. These banks show an unsus-
tainable high growth which indicates higher risk.
On average, ROE is equal to 0.04%, and it is negative for 53 and 54
banks in 2012 and 2013, respectively. EBA (2015) states that unsus-
tainably high levels of profitability ratios also indicate higher risk. The
maximum value of ROE in the sample is 49%, and the ratio is larger
than 20% for 18 banks (10 banks in 2013 and 11 in 2014). This
numbers may suggest some problems of the sustainability of the business
model in the long term.
On the efficiency side, the average cost to income ratio is 66%.
Nineteen banks have a ratio larger than 100% at least in one year: in
particular 7 banks in 2012, 9 in 2013 and 8 in 2014; among them, two
banks have a ratio larger than 200%. The unusually high ratio indicates
that the bank’s costs are out of control. A very low ratio may indicate that
operating costs are too low for the bank to have the required risk and
control functions in place, also indicating higher risk, but this is not the
case for the sample under scrutiny since only seven banks have a CI ratio
smaller than 20% (EBA 2015).
Overall, this chapter examines seven over nine core indicators and
three over 13 additional indicators, which are enough to perform a sig-
nificant analysis, in the author’s point of view.
9.4.2 Individual Risk Score
As the FITD, also EBA proposes thresholds, classes and weights to

compute individual bank risk scores (IRS). Unlike the Fund, however,
EBA allows two methods to assign banks to risk classes: the bucket
method and the sliding method. The first one uses a fixed number of
buckets defined for each risk indicator by setting upper and lower
boundaries for each bucket. The number of buckets for each risk indi-
cator should be at least two. The buckets should reflect different levels of
risk posed by the member banks (e.g. high, medium, low risk) assessed
on the basis of particular indicators (EBA 2015).
Where the calculation method follows the sliding scale approach
instead of a fixed number of risk classes, the upper and lower limits are
set by the DGS on the basis of regulatory requirements or historical data
on the particular indicator. Since the sliding method is based on infor-
mation available only to the national DGS, this chapter uses the bucket
method, which is also closer to the FITD system, thus allowing easier
comparison between the two.
9.4.3 Bucket Method
In the bucket method, an individual risk score is assigned to each bucket.

The buckets’ boundaries should be determined either on a relative or
absolute basis. When using the relative basis, the IRSs of banks depend
on their relative risk position vis-à-vis other institutions; in this case,
institutions are distributed evenly between risk buckets, meaning that
institutions with similar risk profiles may end up in different buckets. In
the absolute basis, the buckets’ boundaries are determined to reflect the
riskiness of a specific indicator; in this case, all banks may end up in the
same bucket if they all have a similar level of riskiness.
222 F. Arnaboldi
Table 9.8 Buckets, boundaries and individual risk score

Buckets Boundaries (%) IRS
1 <2 0
2 ¼< 2–7 < 50
3 >=7 100
Source EBA (2015)
Note Risk indicator for which higher values indicate higher risk (NPL ratio)
For each risk indicator, the IRSs assigned to buckets should range
from 0 to 100, where zero indicates the lowest risk and 100 the highest
risk.
Table 9.8 shows an example of bucket-scoring by type of risk indi-
cator, where higher values of the risk indicator mean higher risk (for
example, NPL ratio).
To compute the IRS of the sample banks, buckets and boundaries
provided by EBA have been used for LR, NPL ratio, ROA, ROE and
total asset growth. EBA does not provide specific examples for the
leverage ratio, CET1, RWA/TA and cost to income ratio, thus relative
boundaries, which correspond to the 20, 40 and 60th‰ of the sample
banks distribution year to year, have been used for those indicators. The
percentiles and corresponding IRS have been fixed according to EBA
guidelines. Relative boundaries imply an even distribution of banks
among risk buckets, and Table 9.9 shows an example of buckets, relative
boundaries and individual risk scores.
Table 9.9 Buckets, relative boundaries and individual risk score

Bucket Boundaries IRS
1 >60° ‰ 0
2 <40°–60° = < 33
3 <20°–40° = < 66
4 ¼< 20° ‰ 100
Source own computation on EBA (2015)
Note Risk indicator for which higher values indicate lower risk (liquidity ratio)
9.4.4 Aggregate Risk Score
EBA (2015) multiplies each IRS by an indicator weight (IW) which

should be the same for all banks and calibrated by using supervisory
assessment and/or historical data on failures of institutions (EBA 2015).
The sum of weights assigned to all risk indicators is equal to 100%.
When assigning weights to particular risk indicators, the minimum
weights for the risk categories and core risk indicators, which sum up to
75%, should be preserved.
When only core indicators are computed and NSFR is not yet avail-
able, EBA (2015) states that the minimum IW assigned to NSFR is
assigned to LR, which belongs to the same risk category. One of the
possible allocation of weights suggested by EBA, when both core and
additional indicators are computed, allows five additional indicators in
four different categories. These indicators can be freely chosen by the
DGS.
The aggregate risk score (ARS) is the weighted average of the IRS,
according to the following formula:
X
n
ARSi ¼ IWj IRSj ð9:3Þ
j¼1
P
n
where: IWj ¼ 100% and IRSj ¼ IRSxj when X in fA; B; . . .; M g,
j¼1
that is the bucket corresponding to indicator Aj .
Following the guidelines, since NSFR is not computed during the
transition period, the IW explained above is applied to core indicators. In
addition, as previously mentioned, the ratio of unencumbered assets to
covered deposits has not been computed because data on unencumbered
assets for the sample banks were not available. Thus, the weight (17%)
originally assigned by EBA to this ratio is equally allocated among all
other computed indicators.
Consequently, when only core indicators are investigated, the ARS is
computed according to:
224 F. Arnaboldi
ARScore ¼ 0:15 leverage ratio þ 0:15 CET 1 þ 0:25 LR þ 0:21

NPL ratio þ 0:12 RWA=TA þ 0:12 ROA
ð9:4Þ
When core and additional indicators are considered, weights are

applied to each risk category, except for the business model and man-
agement. All three additional indicators belong to this category; thus, its
weight is given by the sum of the weights of business model and man-
agement and of potential losses for the DGS.
ARScore þ additional ¼ 0:115 leverage ratio þ 0:115 CET1 þ 0:18

LR þ 0:18 NPL ratio þ 0:085
RWA=TA þ 0:085 ROA þ 0:08
ROE þ 0:08 TAG þ 0:08 CI
ð9:5Þ
Descriptive statistics of ARS are computed by applying formula

(4) and (5) to the sample banks. Over 2012–2014, the averages of
ARScore and ARScore + additional are almost the same, but ARScore +
additional standard deviation is lower and the minimum higher than
ARScore. This may suggest a lower volatility when additional indicators
are taken into consideration.
According to EBA (2015), every ARS has a corresponding aggregate
risk weight (ARW), which should be used to calculate the contribution
of an individual member bank to the DGS (Table 9.10). When ARW is
75%, the member bank gets a discount on contribution to be paid
because it is considered as a low-risk bank. When ARW is 100%, con-
tribution does not change. When ARW is higher than 100% (either 125
or 150%), the member bank is considered as a high-risk bank and has to
pay higher contributions.
The average ARS of the sample banks is about 60, which assigns the
sample to the ARW of 125%. Overall, banks should pay higher con-
tributions to the national DGS. Of course, ARS is assigned to each
member bank year by year: additional information is reported in
Table 9.10 Aggregate risk weight

Risk classes ARS boundaries ARW (%)
1 <40 75
2 ¼< 40–55 < 100
3 ¼< 55–70< 125
4 > = 70 150
Source EBA (2015)
Table 9.11. Table 9.11 part (a) lists the number of banks in each risk
class using only core indicators in the year 2013 and year 2014; part
(b) shows the distribution of banks considering both core and additional
indicators.
Looking at core indicators [Table 9.11 part (a)], from 2013 to 2014,
the number of banks in risk class 1 decreases by 20%, whereas the
number of banks in class 2 (ARW = 100%) increases by 45%. Changes
in the other two risk classes are negligible. Thus, it seems that the sample
banks became more risky in 1-year time, and their contributions to the
Table 9.11 Number of banks, risk classes, ARWcore and ARWcore+additional (2013 and
2014)
2013 2014
Risk ARWcore Number Percentage Number Percentage Change
classes (%) of banks of banks 2013–2014
(%)
(a)
1 75 27 20 16 15.53 −20
2 100 27 20 29 28.16 45
3 125 39 28.89 28 27.18 −3
4 150 42 31.11 30 29.13 −4
Total 135 100 103 100
(b)
1 75 18 13.53 13 12.62 −4
2 100 35 26.32 31 30.1 18
3 125 46 34.59 30 29.13 −13
4 150 34 25.56 29 28.16 13
Total 133 100 103 100
226 F. Arnaboldi
Fund would not be further discounted. Table 9.11 part (b) confirms this
scenario, applying core and additional indicators to sample banks in the
year 2013 and year 2014. The number of banks in class 1 diminishes by
4%, and the number of banks in class 2 increases by 18%. The change in
the number of banks in class 3 (−13%) is perfectly matched by the
change in class 4.
9.5 Comparison Between FITD and EBA

Monitoring System of Bank Riskiness
The comparison between the monitoring systems applied by the FITD
and proposed by EBA is not straightforward for many motives. First, risk
categories are different, for instance the FITD does not consider the
potential losses for DGS. Second, within the same category, indicators
are computed differently, as, for example, indicator A1 of the FITD and
the NPL ratio proposed by EBA, or liquidity ratios, which have different
numerator and denominator. Third, the number of indicators is signif-
icantly different between the two systems: five indicators for the Italian
Fund versus nine core indicators proposed by EBA. Furthermore, EBA
suggests to use thirteen additional indicators. Fourth, indicator weights
are the same for all ratios except A1 in the case of the FITD, whereas
many different scenarios are proposed by EBA, with the only prescription
of minimum weights for core indicators. Last but not least, risk classes
cannot be easily compared since EBA proposes four classes (75% lowest
risk, 100% average risk, 125% risky and 150% most risky), whereas
FITD assigns banks to six different classes.
However, notwithstanding all differences, the core of the two systems
is the same, since both works on the assessment of member bank’s risk
and the result of the assessment increase or decrease contributions to be
paid to the DGS. Assuming to modify the thresholds in Table 9.4 in
order to fit the four risk classes proposed by EBA in Table 9.12, this
chapter suggests the match among risk classes reported in Table 9.12.
The match reported in Table 9.12 is based on the level of contribu-
tions to be paid to the DGS according to FITD (2012): when WAAI is
Table 9.12 Risk classes (FITD versus EBA)

WAAI ARW
0 75% lowest risk
0–2.9 100% average risk
2.9–8.8 125% risky
>8.8 150% most risky
Source Own computation on FITD (2012) and EBA (2015)
larger than 3.5, the bank’s contribution quota shall be increased as it

happens when ARW is greater than 100%; when WAAI is greater than
zero and less than or equal to 3.5, the bank shall retain its contribution
quota unchanged, as when ARW is equal to 100%; if the WAAI is equal
to zero, the bank shall benefit from a reduction in its contribution quota,
as it happens when ARW is 75%. The additional threshold (10.5) for
WAAI has been identified on the basis of Table 9.3: when WAAI is
above 10.5, the member bank is considered at high or expulsion risk.
This category is matched with ARW equal to 150%, which means a
substantial increase in contribution. Since in this chapter thresholds have
been scaled to avoid distortions, the Funds 3.5 and 10.5 are scaled to 2.9
and 8.8.
Table 9.13 summarizes the changes in member banks’ classification in
2013 and 2014 when EBA core indicators (panel a) or core and addi-
tional indicators (panel b) are applied instead of FITD indicators. When
indicators proposed by EBA are applied to Italian banks, the distribution
of those banks among risk classes improves.
Table 9.13 panel (a) shows the changes in risk classes when only core
indicators are applied. The number of banks belonging to the low-risk
class increases by 500% (20 banks) in 2013 and by 367% (11 banks) in
2014. Those banks experience a discount in contribution quota to the
Fund. The number of banks in class 2, which retains their contribution
quota unchanged, increases by 500% (+20 banks) in 2013 and by 700%
(+21 banks) in 2014. Conversely, the number of banks in the highest
risk class (class 4) decreases by 59% in 2013 and by 61% in 2014 (−58
and −46 banks, respectively).
When core and additional indicators are applied (Table 9.13 panel b),
the number of banks in class 1 would increase by 275% (11 banks) in
228 F. Arnaboldi
Table 9.13 Changes in risk classes

Risk EBA 2013 Changes in risk classes from FITD
classes core to EBA
FITD 0.75 1 1.25 1.5 Total Number of Percentage
banks (%)
(a)
1 4 0 0 0 4 20 500
2 3 0 1 0 4 20 500
3 12 7 1 1 21 18 86
4 5 17 37 40 99 −58 −59
Total 24 24 39 41 128
2014
0.75 1 1.25 1.5 Total
1 3 0 0 0 3 11 367
2 2 1 0 0 3 21 700
3 6 7 0 0 13 14 108
4 3 16 27 29 75 −46 −61
Total 14 24 27 29 94
(b)
1 3 1 0 0 4 11 275
2 1 2 1 0 4 28 700
3 8 10 1 1 20 26 130
4 3 19 44 32 98 −65 −66
Total 15 32 46 33 126
2014
0.75 1 1.25 1.5 Total
1 3 0 0 0 3 8 267
2 2 1 0 0 3 25 833
3 4 9 0 0 13 15 115
4 2 18 28 27 75 −48 −64
Total 11 28 28 27 94
2013 and by 267% (8 banks) in 2014. In class 2, the number increases

by 700% (+28 banks) in 2013 and by 833% (+25 banks) in 2014,
whereas the number of banks in class 4 decreases by 66% in 2013
(65 banks) and by 64% in 2014 (48 banks).
9.6 Conclusion
Deposit guarantee schemes are an essential element in the completion of
the internal market and an indispensable complement to the system of
supervision of banks.
The set-up of a European Deposit Insurance Scheme was mildly wel-
comed by some member states. They were concerned that sharing the
responsibility of backstopping deposits without tackling the remaining
risks in banking systems would increase moral hazard. This concern is based
on the assumption that EDIS would increase the level of contributions
banks of some member states have to pay according to their riskiness.
To test this hypothesis, this chapter analyses monitoring systems of
bank riskiness currently applied by the FITD and proposed by EBA on a
sample of Italian banks members of the FITD. The conclusion is
twofold.
First, the change of indicators, thresholds, weights and risk classes is
applied to years 2013 and 2014 and shows that EBA proposal would
increase the number of banks in the lower-risk classes, where contribu-
tion quota to the DGS would remain unchanged or would decrease. This
outcome points out that, on average, sample banks would pay less
contributions to the DGS when EBA guidelines are applied. This should
reassure member states concerned about Italian banks’ moral hazard in
the event of the set-up of a common backstop for deposits.
Unfortunately, the results in this chapter are approximate because of
the lack of data and information. Since the issue is relevant, the FITD,
which is the only one that has the full set of data and information, might
consider disclosing the real situation. Uncertainty undermines
bank-client relationship and obstacles a trustfully relationship with other
member states.
Second, on the effectiveness of EBA implementation of Directive
2014/49/EU goals, this chapter suggests some caution. Carefully ana-
lysing the monitoring system proposed by the European regulator, it
emerges that EBA proposes many indicators, which composition is not
always clear, and allows national DGSs great flexibility in line with the
principle of proportionality. The choices of how many and which
230 F. Arnaboldi
additional indicators to use, of the weights to assign to each risk category

and, within each category, to each indicator, of the bucket or of the
sliding method to fix boundaries, and, depending on the chosen method,
of boundaries themselves to compute individual risk scores, are just some
of the decisions EBA allows national DGSs to take. While it is clear that
EBA guidelines will contribute to providing incentives to banks to
operate under a less risky business model and to speed up the conver-
gence process, it is not so clear whether such discretionary power allowed
to national DGSs on many relevant features would benefit the system.
The outcome of the risk assessment can vary strongly, depending on the
choices made. This negatively affects the harmonization and compara-
bility of the national schemes, fuelling, once again, concerns among
member states about the true riskiness of other member states’ banking
systems. The goals of Directive 2014/49/EU seem postponed to the near
future and rely on the adoption of EDIS.
Notes
1. Further information can be found in “Germany warns on eurozone
bank deposit plan” (Financial Times 2015) and in the Deutsche
Bundesbank’s Monthly Report of December 2015, p. 58–60.
2. D. Lgs. 659/96, art. 2, c. 1, which transposes directive 94/19/CE, states
that all information, news or data related to FITD are privileged
communications.
3. Six banks out of 172 (3% of the sample) report data to compute only
one ratio.
4. The year 2015 is not included in the analysis, since the number of
observations is less than 50% compared to the previous years.
5. “This procedure consists of a set of steps made for determining the point
of equilibrium quota” which could be only performed by the Fund
(FITD 2012, p. 23).
6. Data are available upon request.
7. A higher CET1 indicates a better risk mitigation. Tier 1 capital is given
by the sum of common equity tier 1 and of additional tier 1 (BIS 2012).
8. Required tier 1 ratio is 4.5% and 5.5% for the year 2013 and 2014,
respectively (BIS 2012).
9. To avoid including one-off events and avoid pro-cyclicality in contri-

butions, an average of 2 years data is used (EBA 2015).
10. EBA defines unencumbered and encumbered asset as the following: “an
asset should be treated as encumbered if it has been pledged or it is
subject to any form of arrangement to secure, collaterise or
credit-enhance any on-balance sheet or off-balance sheet transaction
from which it cannot be freely withdrawn (for instance, to be pledged
for funding purposes)” (EBA 2015, p. 22).
Appendix
Table A.1 shows the test for equality of means between top- and
bottom-performing banks in terms of return on equity (ROE) over
2012–2014. As far as asset quality is concerned, most profitable banks
are significantly less risky (A1 equals 50% versus 178%, respectively).
Top-performing banks are also less exposed to liquidity risk (L takes the
value of 65 versus 97%). Top- and bottom-performing banks do not
show any significant difference in means for profitability ratios D1 and
D2.
Table A.2 shows the results of the test of the difference in means
among banks belonging to the top and bottom quartile in terms of ROE.
Top quartile banks have a smaller leverage ratio (66% vs. 96%), a higher
liquidity ratio (27% vs. 14%), a higher quality of loan portfolio (NPL
ratio equal to 4 and 8%, respectively), lower RWA to total assets
Table A.1 Test for difference in means—ROE

Ratio Bottom quartile Top quartile Difference in means
# Mean # Mean (p-value)
observations observations
A1 87 1.78 78 0.5 0.0000***
D1 96 0.52 93 0.6 0.3941
D2 87 −16.95 84 −16.43 0.4927
L 96 0.97 94 0.65 0.0000***
Note Top-performing banks belong to the first quartile of yearly distribution,
bottom-performing ones to the fourth quartile
232 F. Arnaboldi
Table A.2 Top and bottom quartile—ROE

Indicators Bottom quartile Top quartile Difference in
Core Number of Mean Number of Mean means (p-value)
observations observations
Leverage ratio 82 0.96 75 0.66 0.0134**
CET1 92 0.19 88 0.22 0.2391
Capital 93 3.8 88 3.78 0.4871
coverage
ratio (%)
Liquidity ratio 97 0.14 96 0.27 0.0001***
NPL ratio 91 0.08 86 0.04 0.0000***
Return on 53 −0.84 55 0.94 0.0000***
asset (%)
RWA to total 92 0.59 88 0.43 0.0000***
asset
Additional
Total asset 57 0.1 55 0.12 0.3966
growth
Cost income 92 70.64 95 59.23 0.0021***
(%)
(43 vs 59%) and lower CI (59 vs. 71%) than bottom quartile banks.
All EBA indicators suggest that top-performing banks in terms of ROE
have a lower risk than the worst-performing ones.
References
Bank of Italy. 2016. FAQ—La centrale dei rischi. Available at https://www.
bancaditalia.it/servizi-cittadino/servizi/accesso-cr/faq-cr/faq-cr.
html#faq8761-9.
BIS. 2012. Basel III. Necessary, but not sufficient, Remarks by W. Byres,
Secretary General of the Basel Committee on Banking Superivision to the
Financial Stability Institute’s 6th Biennial Conference on Risk Management
and Supervision, Basel.
Deutsche Bundesbank. 2015. Monthly Report December 2015.
Directive 2014/49/EU of the European Parliament and of the Council of 16
April 2014 on Deposit Guarantee Schemes. Official Journal of the European
Union L 173/149.
European Commission. 2015a. Proposal for a Regulation of the European

Parliament and of the Council amending Regulation (EU) 806/2014 in order
to establish a European Deposit Insurance Scheme, Strasbourg, 24.11.2015,
COM(2015) 586 final, 2015/0270(COD).
European Commission. 2015b. Factsheet: A stronger banking union, 24 November
2015. Available at: http://ec.europa.eu/finance/general-policy/docs/banking-
union/european-deposit-insurance-scheme/151124-factsheets_en.pdf.
EBA. 2015. Guidelines on Methods for Calculating Contributions to Deposit
Guarantee Schemes, EBA/GL/2015/10, 28 May 2015.
Financial Times. 2015. Germany Warns on Eurozone Bank Deposit Plan,
8 December.
FITD. 2012. The FITD’s Monitoring System of Bank Riskiness and Risk-Based
Contribution, November.
FITD. 2016. Annual Report 2015.
Juncker J-C, Tusk D., Dijsselbloem J., Draghi M., and Schulz M. 2015.
Completing Europe’s Economic and Monetary Union, July. Available at: http://
ec.europa.eu/priorities/sites/beta-political/files/5-presidents-report_en.pdf.
Reuters. 2015. EU Deposit Insurance’ Vanishes from EU Leaders’ Draft
Conclusions, 18 December 2015. Available at: http://uk.reuters.com/article/
uk-eurozone-banks-deposits-idUKKBN0U11KP20151218.
Senato. 2015. Dossier del Servizio Studi sull’A.S. n. 1758 “Delega al Governo
per il recepimento delle direttive europee e l’attuazione di altri atti
dell’Unione europea —Legge di delegazione europea 2014” XVII legislatura,
February 2015 n. 197.
Index
A Bankruptcy, 50, 190

Administration, 185, 188, 193, 194, Bankscope, 64, 121, 206, 208–209,
207 214, 217
Aliano, M., 55, 79 Basel II, 112
Arbitrage, 144, 157 Basel III, 112
Arnaboldi, F., 203 Belgium, 175, 184
Asset quality, 58, 60, 63, 207–208, Bloomberg, 90, 93
217–218, 231 Bond, 36–37, 39–42, 148, 161, 208
Australia, 50 Borrower, 57–58
Austria, 183–184, 198 Branch, 120, 177, 184–194, 208,
207
Bulgaria, 121, 175, 184–185
B
Balance sheet, 11, 36–37, 56, 59,
63–65, 75–76, 90, 105, 132, C
142–143, 170, 193, 207, 209, Canada, 38, 50
219–220 Capital market, 117, 150, 189
Bank fund, 50, 52 Capital ratio, 51, 61, 145, 147
Bank risk, 12, 15, 80, 88, 144–146, Cash flow, 87, 93
151, 156, 161, 213, 216, 221,
226, 229
© The Editor(s) (if applicable) and The Author(s) 2017 235

in Banking and Financial Institutions, DOI 10.1007/978-3-319-54894-4
236 Index
Central bank, 5, 7, 22, 26, 29, 36, E

42, 93, 143, 181, 188, 190, Earnings per share (EPS), 93, 99,
193, 198, 204 102, 105
Chesini, G., 1 EBIT, 86, 93
Cluster analysis, 138, 141, 188 E-business, 134–135, 158
Coinsurance, 172, 177, 186, Efficiency, 22
190–193, 196, 198, 205 Elliot, V., 35
Commercial bank, 148, 187, 189 Equity, 52, 68, 71, 73, 82, 85, 89,
Company, 85, 88, 113–114, 93, 99, 115–116, 133, 142,
119–120, 126, 135–137, 175, 160, 218
181 Estonia, 186
Corporate governance, 114, 119 EU28, 168, 171–172, 176, 181, 183,
Cosma, A., 131 185, 189–190, 194, 196,
Credit quality, 56–59, 61, 67, 70–73, 198–199
76, 143 Europe, 38, 110, 155, 177, 183, 195,
Credit risk, 8, 12, 14–16, 52, 73 199
Crespi, F., 55 European Banking Authority (EBA),
Croatia, 121, 175, 185 42, 118, 142–143, 153–154,
Cruz-Garcìa, 5 178, 181, 204–206, 216, 231
Currency, 51, 184, 190–194, 203 European banks, 63, 141–143, 149
Customer, 45, 52, 60, 64, 111, European Central Bank (ECB), 7, 30,
113–114, 116–118, 120, 76, 142–143, 146, 153–154,
134–137, 139–140, 148, 209, 156, 181, 204
218 Eurozone, 23, 27, 29, 30
Cyprus, 175, 181, 185
Czech Republic, 121, 181
F
Fernàndez de Guevara, J., 5
D Ferretti, R., 131
Default, 9, 15, 37–38, 80, 144, 146, Financial crisis, 79
159 Finland, 175, 196
Denmark, 61 Framework, 8–9, 38, 71, 110, 114,
Depositor, 115, 170, 172, 176–177, 119, 121, 126, 132, 152, 156,
181, 183–187, 189–191, 199, 161, 172, 176, 178, 181, 187,
204–205 198–199, 205, 206
DGS, 176, 181, 199, 204–206, France, 175, 186, 196
216–219, 223–224, 226, 229,
230
Index 237
G 153, 155, 161, 169, 170,

Galloppo, G., 79 176–178, 183–187, 189–190,
GDP, 16–17, 24, 27, 58, 76, 125 206, 218, 220–221, 223
Germany, 61, 175, 181, 187, 196, Insurance, 39, 52, 110, 115, 118,
205 150, 170–172, 174, 176, 178,
Giaretta, E., 1 181, 183–199, 204, 205, 229
Governance, 112, 114–115, Interbank, 10, 12, 15, 22, 113, 184,
118–120, 126, 136–137, 153, 187–194
154, 158, 161, 219 Interest income, 9, 11, 16, 26, 68,
Greece, 62, 175, 187–188 70, 71, 72, 73, 75, 145, 149,
Gualandri, E., 131 150, 151, 159
Guarentees, 76, 172, 185, 189, 207 Interest margin, 5–12, 16–17,
22–24, 26, 29–30
Interest rate, 5–12, 15, 17, 22–24,
H 26–27, 29–30, 35–36, 41–42,
Herfindahl Index, 25 44, 47–51, 58, 64, 71, 73, 132
Household, 6, 35–36, 58–59 Investor, 36, 39–40, 50, 79, 84, 103,
Hungary, 175, 188 113, 115, 119, 122, 133, 148,
157, 161, 170, 172, 187, 190
Ireland, 175, 188, 195
I IT, 134, 158
Income, 9, 11–13, 16–17, 26, 35, Italy, 56, 61, 65, 68, 75–76, 172,
58–59, 62, 68, 70–73, 75, 177, 181, 189, 196, 206–208
88–89, 93, 99, 104, 123,
141–147, 149–151, 158–161,
209, 217, 220, 222 J
Indebtedness, 35, 38, 51, 58 Japan, 23, 27, 29
Index, 9, 12–14, 22, 24, 63, 85,
89–90, 93, 96, 119, 122–125,
147 K
Industry, 61, 63, 90, 116–117, Kil, K., 109
137–138, 140, 149, 160, 170,
176, 198, 199
Insolvency, 15, 39, 50, 64, 67, 85, L
149, 150, 208, 219 Landi, A., 131
Institution, 12, 37, 39, 40, 93, 103, Latvia, 189
109, 111–113, 115–116, 118, Legislation, 36, 39, 172, 203, 204,
120–121, 144–146, 149, 150, 206
238 Index
Lending bank, 143–144 N

Lending long, 48–49 Net interest margin, 5–6, 7–12, 16,
Lerner Index, 9, 12–14, 22 17, 22–24, 26, 29, 30
Leverage, 80, 86, 88, 93, 94, 96, 99, Network, 133–134, 137, 181, 199,
104, 149, 218, 222 204
Liability, 51, 115, 141–143, 147, Non-performing Loan (NPL), 7, 15,
158 55–65, 67, 75–76, 123–124,
Lindblom, T., 35 218, 219, 222, 226
Liquidity, 9, 11, 12, 50, 80–82, 84,
88–89, 93–94, 96, 99, 102,
104, 112, 115, 132, 153, 161, O
170, 189, 193, 198, 207–209, OECD, 12, 30
212, 217, 218, 226
Listed bank, 110, 125–126, 142,
145, 149, 160 P
Lithuania, 121, 175, 189 Paimanova, V., 79
Luxemburg, 190, 196 Paltrinieri, A., 1
Panel, 12–13, 22, 61–62, 67, 72, 74,
82–84, 110–111, 120–121,
M 123–124, 126, 150–151, 227
Macroeconomic variable, 10, 12, Payment, 16–17, 24, 70, 120, 184,
56–57, 59, 63, 125, 143 189–190, 198
Malta, 175, 190 Payout, 176, 177, 181, 189,
Management, 16, 17, 55–57, 59–64, 191–192, 198, 205, 207
70, 71, 72, 75, 110–113, 119, Performance, 62, 82, 93, 110, 116,
120, 133–136, 138–140, 119–125, 132, 134, 138,
143–144, 148, 152–156, 158, 140–143, 147–151, 155, 157,
161, 174–175, 188, 191–192, 158–159
194–198, 217, 220 Poisson distribution, 10
Maudos, J., 5 Poland, 121, 175, 191
Migliavacca, M., 169 Policy, 96, 99, 102, 104–105, 170
Miklaszewska, E., 109 Portfolio, 10, 47, 60–61, 139–140,
Monetary policy, 5, 22, 26, 30, 76, 143–144, 147–148, 151,
78 157–159, 189, 219
Mortgage rate, 36–37, 42–45, 48 Portugal, 175, 181, 191
Mortgage loan, 36, 47–48 Post-crisis, 39, 109, 113, 116, 189
Index 239
Pre-crisis, 109, 144–146 Survey, 53, 65, 81, 116–119, 126,

Previati, D.A., 79 171–172, 176, 178, 183
Prudential Regulation Authority Sweden, 36–37, 39, 42, 50, 51, 194
(PRA), 153 Swedish Bankers’ Association (SBA),
35, 39, 42, 45
Swedish Financial Supervisory
R Authority (SFSA), 35, 39, 42
Rating, 10, 41, 120–122, 128, 157 Systematic risk, 80–81, 89–90, 99,
Regression, 72, 77, 81, 84, 110 102, 150
Regulation, 39, 114, 132, 147, 176,
178
Return on assets (ROA), 93–94, 96, T
102, 104, 219, 222, 224 Test, 8, 59, 73, 81, 93, 105,
Return on equity (ROE), 62–63, 76, 121–122, 181
123–125, 220, 222, 224, 232 Total Assets, 13, 17, 24, 56, 65, 68,
Revenue, 61, 140, 148, 151 70, 72–73, 75, 88, 93, 145,
Review, 63, 81, 112, 131–133, 142, 218, 219, 220, 222
147, 153, 156–159, 181
Romania, 121, 172, 175, 192
U
UK, 23, 27, 29, 61, 116, 152, 196
S USA, 8, 23, 27, 29, 56, 80, 90, 112,
Saving, 5, 12, 42, 49, 145, 148, 170, 116, 121
172, 184, 187, 199
Security, 90
Shareholder, 111–115, 120–121, 220 V
Slovakia, 121, 175, 192 Venturelli, V., 131
Slovan Republic, 121, 193 Volatility, 12, 15, 79–84, 103–105,
Solvency, 15, 39, 50, 64, 67, 85, 149, 133, 141–142, 149
150, 175, 186, 195, 207, 208
Spain, 61, 175, 181, 183, 193
Strategy, 36, 113, 116, 118, 125, Z
134–138, 148, 149, 153–155, Z-score, 123–124, 146
157–160

2017 Book TheBusinessOfBanking

Uploaded by

Copyright:

Available Formats

2017 Book TheBusinessOfBanking

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2017 Book TheBusinessOfBanking

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What topics are covered in the series?

What topics are covered in the series?

PA LG R AV E M AC M I L L A N S T U D I E S I N

BANKING AND FINANCIAL INSTITUTIONS

Edited by Giusy Chesini,

More information about this series at

Palgrave Macmillan Studies in Banking and Financial Institutions

© The Editor(s) (if applicable) and The Author(s) 2017

Cover image: © Maciej Bledowski/Alamy Stock Photo

Printed on acid-free paper

This Palgrave Macmillan imprint is published by Springer Nature

Finally, as conference organizers, we would like to thank Steven

2 Interest Rates and Net Interest Margins: The Impact

3 The Swedish Mortgage Market: Bank Funding, Margins,

4 Incapability or Bad Luck? Testing the “Bad

5 Why Do US Banks React Differently to Short Selling

6 Reputational Risk in Banking: Important to Whom? 109

7 The Business Model of Banks: A Review of the

8 On European Deposit Protection Scheme(s) 169

9 A Technical Approach to Deposit Guarantee Schemes 203

Mauro Aliano is an Assistant Professor of Banking and Finance at the

Giusy Chesini is Associate Professor in Banking and Finance at the

Administration, New York University. He is a member of CEFIN

Journal of International Financial Markets, Institutions and Money,

years, with particular focus on bank management, strategy, and organi-

Fig. 2.1 Intervention interest rates by the main Central Banks 6

Fig. 3.6 Average discounts on the banks’ ofﬁcially offered

Table 2.1 Descriptive statistics (2003–2014 averages) 18

Table 5.5 Stock price change 97

G. Chesini (&) E. Giaretta

Deyoung (1997), which suggests that poor managerial practice causes an

P. Cruz-García (&) J. Fernández de Guevara J. Maudos

3.00% Bank of England (BoE)

2.00% Federal Reserve (Fed)

depression in consumption which this entails), a lower rate of techno-

proﬁtability, as a result of the fall in ﬁnancial revenues due to the

existence of a non-linear relationship between interest rates and the net

2.2 Theoretical Framework and Testable

There are various theoretical frameworks in which the behaviour of net

With these additions to the original model by Ho and Saunders

indicators and variables that approximate the characteristics of each bank

2.2.2 Testable Hypothesis

2.3 Data, Deﬁnition of Variables

and the income statement of a representative sample of banks from

Level of Interest Rates

We use the three-month interbank market interest rate (Short-term

As an approximation of market power, two alternative measures are used.

where Pi is the average price of banking products, which is approximated

The logarithm of loan volumes (log-loans) is included as a proxy for bank

Given the possibility of non-payment or default on loans, banks include

Interest Rate Risk

Money market uncertainty is approximated by using the coefﬁcient of

Interaction Between Credit Risk and Market Risk (Risk

Interaction between credit risk and market risk (Risk covariance) is

Average Cost of Transactions (Average Cost)

Liquid Reserves (Reserves)

A high volume of liquid reserves has a positive effect on the bank

Implicit Interest Payments