Irek Akhunianov

Transparency and Firm Value: Evidence from the Financial Crisis

2007-2009

Working Paper # BSP/2008/101E

This paper is based on the Master Thesis prepared at NES in 2009 in the framework of the
research project “Politics and Finance” under the supervision of professors A. Goriaev, R.
Enikolopov and M. Petrova (NES).
The research was supported by Ford Foundation, World Bank, John D. and Catherine T.
MacArthur Foundation.
The author would like to thank his academic advisors, Art Durnev and all the participants of
XXV NES research conference.

Moscow
2009
 Akhunianov Irek. Transparency and Firm Value: Evidence from the Financial Crisis
2007-2009 / Working Paper # BSP/2008/101E – Moscow, New Economic School, 2009. –
31 p. (Engl.)

During the current financial crisis stock markets all over the world experienced a sharp decline.
The magnitude of the fall, however, varied across countries and companies. There are a number of
factors that can explain this variation. In this work I focus on the effect of institutions and corporate
governance on the magnitude of the fall of stock prices during the crisis. I consider both the effect
of country-level institutions and firm-level characteristics. In cross-country analysis using data from
53 countries I show that the fall of the stock market was stronger in countries with worse corporate
governance and lower quality of government institutions. In the firm-level analysis using the data
on over 850 firms from 43 countries I find that the fall in the stock prices was stronger from firms
with lower level of disclosure and transparency. These results are consistent with the finding in the
literature regarding the effect of the Asian financial crisis of 1997-1998. In addition I find that during
the crisis the effect of firm-level transparency is more important in countries with better corporate
governance and government institutions, which is contrary to the findings in the literature for the
non-crisis periods.

Key words: transparency, corporate governance, stock market, government institutions, financial
crisis

Ахуньянов И.Х. Уровень прозрачности компаний и финансовый кризис 2007 -

2009/ Препринт # BSP/2008/101E - М.: Российская Экономическая Школа, 2009. – 31
с. (Англ.)

Во время текущего финансового кризиса наблюдается значительное падение многих

фондовых рынков. Однако размах падения индексов разных стран и цен акций разных
компаний сильно различается. Существует множество факторов, которые потенциально
объясняют эти различия, но в данной работе исследуется влияние качества институтов и
корпоративного управления на поведение цен акций во время кризиса. Рассматривается
эффект качества государственных институтов страны на поведение фондового рынка и
эффект корпоративного управления на динамику котировок компании. Анализ на выборке
из 53 стран показал, что в странах со слабыми государственными институтами, величина
падения индекса фондового рынка страны более значительна. В данной работе показано,
что на выборке из более 850 фирм из 43 стран компании с низким уровнем корпоративного
управления испытывают большее падение в цене акции. Ключевым результатом является
то, что высокий уровень корпоративного управления играет значительную роль во время
кризиса в странах с хорошими государственными институтами, что противоречит
предыдущим результатам работ, исследующих некризисные периоды.

Ключевые слова: корпоративное управление, прозрачность, фондовый рынок, финансовый

кризис, государственные институты

ISBN
© Ахуньянов И.Х.., 2009 г.
© Российская экономическая школа, 2009 г.

2
Contents
1. Introduction ...................................................................................................................................4
2. Data and Methodology ..................................................................................................................5
3. Results ...........................................................................................................................................7
3.1. Main results ...........................................................................................................................7
3.2. Firm-level interactions.........................................................................................................10
3.3. Country-level interactions ...................................................................................................11
3.4. Robustness check.................................................................................................................14
3.5. Pre-crisis analysis ................................................................................................................17
4. Country-level analysis .................................................................................................................18
4.1. Variable descriptions and econometric issues .....................................................................18
4.1.1. Measuring the crisis .......................................................................................................18
4.1.2. Measuring institutions....................................................................................................19
4.1.3. Control variables ............................................................................................................21
4.1.4. Final sample and econometric models ...........................................................................22
4.2. County-level results.............................................................................................................23
5. Conclusion ...................................................................................................................................28
6. References ...................................................................................................................................30

3
1. Introduction
The current world economic crisis affected almost all the countries. The crisis started with
the mortgage crisis in the USA, the first signs of which appeared in 2006 in the form of decrease in
number of sales of houses. By mid-2007 it had outgrown in the full-scale crisis of subprime
mortgages. In 2008 the crisis affected almost all the countries and from financial crisis it has
gradually transformed into an economic crisis which have manifested itself in decrease in volumes
of output, reduction in demand, falling of stock prices and growth of unemployment. The world
economic crisis significantly affected the value of almost all firms around the world. The magnitude
of the fall, however, was different across both countries and firms. Among many factors which
potentially had impact on firm and country value I focus on the quality of corporate governance,
which becomes especially important during the crisis.

The previous works considering corporate governance and financial markets, show positive
dependence between the level of investor rights protection and stock market prices, between market
capitalization and level of concentration of shareholders (La Porta, Lopez-de-Silanes and Shleifer,
2005, Levine, 1998). LLSV (1997) shows that in the sample of 49 countries weak protection of
creditors and poor quality of execution of laws leads to delay of development of financial sector.
Research by Nicola, Laeven and Ueda (2007) has shown corporate governance improvement in the
majority of the countries in recent years, convergence of quality of a corporate governance and
positive influence of corporate governance on traditional measures of economic activity of the
country. The work by Bekaert, Harvey, Lundblad and Siegel (2008) shows that measures of
political risks, financial openness and development of the financial market explain about half of
variation of the stock market.

The corporate governance can be especially important during the financial crisis for two
reasons: first, crisis can lead to great volumes of benefit expropriation by managers of the company
as they expect reduction of profit and the future investments. Second, the level of corporate
governance serves as a signal to investors about protection their investments.

Analyzing the Asian financial crisis of 1997-1998 Johnson et al (2000) show that
developing countries with better corporate governance experienced a smaller exchange rate
depreciation and reduction of stock market index. Mitton (2002) provides empirical evidence that
during the Asia crisis firms with better disclosure quality showed better stock price performance.

It is not entirely clear, however, if these results can be generalized to the current crisis.
Asian crisis of 1997-1998 was to a large extent caused by the problems with corporate governance
in the non-financial sector, whereas in the current crisis other factors played major role. Also, these

4
papers do not look how the relationship between a firm’s governance and a firm’s value vary in
different countries and different firms. In particular, it is important to know if country- and firm-
level governance practices work as complements or as substitutes during the crisis. Durnev and Kim
(2005) suggest that corporate governance is more important for firm value in countries with weaker
legal system. I do not know, however, if this relationship remains the same during the crisis, and
which other country- or firm-level variables affect this relationship.

In this paper, I look at the relationship between corporate governance and the decline in firm
value during the current crisis. I analyze both the effect of country-level institutions on stock market
decline in a country and the effect of country-level institutions and firm-level disclosure practices
on the decline in the value of particular stocks.

Our basic hypothesis is that during the current crisis corporate governance plays the same
role as in the Asian crisis and countries and firms with better corporate governance experience
smaller decline in the value of their stocks. I find that data is consistent with this hypothesis, even if
country and industry fixed effects are included. In addition, I find that firm-level corporate
transparence is more important in countries with more developed political and financial institutions.
All our regression results are robust to the inclusion of country and industry fixed effects, thus
reducing a potential omitting variable bias.

I also show that the effect of corporate transparence is smaller for larger firms, probably
because there is more information about their performance available for investors. I also find firms
with larger levels of debt to be less sensitive to corporate governance1.

The next section describes data and methodology. Section 3 reports the main results, results
of robustness tests and pre-crisis study. Section 4 presents country-level analysis and section 5
concludes.

2. Data and Methodology

In this paper, I use data on 885 firms from 43 countries and 60 industries. The ongoing financial
crisis affected many countries all over the world, thus I decided to include firms from as many
countries as possible. All firms in our sample meet following three criteria. First, each firm must
have financial data reported in the OSIRIS2 database, which is the primary data source used in this
study. Second, I exclude firms for which the primary business segment is financial services. Finally,

1
We propose that it is the case as heavily indebted firms are struggling to stay solvent, so their corporate governance
becomes relatively unimportant, e.g. compared with the probability of bankruptcy.
2
Osiris Database include financial information on 38,000 listed and major unlisted/delisted companies worldwide
(30,000 are non-US companies). The information includes: standardized and “as reported” financials (including restated
reports), SEC filings, detailed earnings estimates including recommendations, ownership, stock data, news and ratings.

5
I include firms which have a Transparency & Disclosure (T&D) score which is measured by
Standard & Poor’s3 agency. Industry classification used in this paper is 2-digit SIC4 code which is
available at OSIRIS database.
Figure 1 Distribution of firms by country, totally 885 firms from 43 countries

To measure the impact of the financial crisis on a firm’s performance I use minimum stock
price during the crisis period, from June 2007 through March 2009 with June 2007 equal 100. For
example, if firm’s stock value at 1st June 2007 is 2000 and during the crisis its price fell down to a
minimum point 1200, then the firm’s stock performance would be (1200/2000)*100%= 60%. Such
a definition of the impact o crisis on firm’s performance is used in the literature (Johnson et al.,
2000, Mitton 2002). In figure 2 one can see the dynamic of stock prices for two US companies
(Alcoa and Exxon) and S&P500 index. The magnitude of falling is different for these firms, while
both prices on oil and aluminum (primary Alcoa and Exxon products) dramatically got down.
Figure 2 Stock price performance during financial crisis for two US companies and S&P500

To control on firm specific characteristics I employ measures of leverage, size, debt, capital
intensity (measured by PPE, Property, plant and equipment) along with country and industry
dummies in the regressions. I do that to control for factors that could affect stock prices. Using
leverage, however, may cause some problems in detecting the influence of disclosure quality on
stock prices. Low level of firm’s transparency may affect its leverage before the crisis. To deal with

3
Standard & Poor’s is a leading provider of financial market intelligence. The world’s foremost source of credit ratings,
indices, investment research, risk evaluation and data, Standard & Poor’s provides financial decision-makers with the
intelligence they need to feel confident about their decisions.
4
Standard Industry Classification
6
this possibility, I include both variables into regression. Also I control for market betas, which are
calculated on the pre-crisis period. Overall statistics of the sample is outlined in table 1.
Table 1 Summary statistics
Long term PPE,
Firm Stock price Sales, mln Disclosure
debt, mln Leverage mln
statistics performance USD Index
USD USD
mean 47.0 51.6 182.0 0.8 66.2 58.6
sd 20.9 448.0 1670.0 10.7 543.0 15.3
min 1.4 0.0 0.0011 0.0 0.0017 5.2
max 100.1 7860.0 29700.0 315.3 10400.0 88.7
To measure transparency quality of the firm I use pre-crisis Transparency & Disclosure
index which is scored by Standard & Poor’s agency. This score differs from 0 to 100 points and in
our final sample it ranges from 5 to 89 points.

3. Results
3.1. Main results
The basic specification in this paper is regression of firm’s stock performance on disclose quality
controlling on firm’s characteristics:

where index i denotes firm, Stock Performance is a minimum stock price over the crisis period,
from June 2007 through March 2009 with June 2007 equal 100, Disclosure is T&D Score and CVs
are control variables including country and industry dummies. Below I will consider more
sophisticated models using country-level interaction terms. The main hypothesis is that firms with
the higher level of transparency and disclosure have better stock performance during the financial
crisis, i.e. β>0.
Table 2 presents the results of regressions of crisis-period stock performance on measures of
transparency quality. Different columns represent the results of different specifications with debt
measures and with capital intensity measures using either country fixed effects and clustered by
industry errors or industry fixed effects and clustered by country errors. When country fixed effects
are included and errors clustered by industry I remove omitted variable bias on not observed
country characteristics and control on correlations of errors within industry, i.e. industrial shocks
can be general. Vice versa if industry fixed effects are included and errors clustered by country I
remove omitted variable bias on not observed industry characteristics and control on correlations of
errors within country, i.e. country-level shocks can be general. The magnitude and sign of the
7
coefficients indicates that firms with higher level of disclosure had, on average, a higher stock price
during the crisis. The decrease in level of a transparency of the company by 10 points gets down the
stock price by 2.2 % (columns 1-3) and 1.5% (columns 4-6) over the crisis period.
The coefficient on disclosure is significant at 0.9% – 2% level in all specifications. The
results are economically significant as well. If firm use poor financial transparency and information
disclosure, managers are more likely to use private information to derive benefits, which will
increase agency costs on shareholders. As agency costs become higher, managers can easily extract
benefits and rights of small shareholders. Moreover better disclosure quality works as a signal to
potential investors that their rights will be protected once they decided to pursue stake of the firm.
Thus investors withdraw their assets from non-transparent firms and reinvest them into less risky
assets. Negative coefficient on beta indicates that less correlated with the market firms suffer less
during the crisis. Also high leverage and debt before the crisis imply significant falling of stock
prices. Another key finding here that large firms, on average, suffer less during the crisis.
Table 2
Stock prices during the crisis and level of transparency
Dependent variable: Stock price at the lowest point during financial crisis with June, 2007 = 100.
Country fixed effects, errors clustered by Industry fixed effects, errors
industry clustered by country
(1) (2) (3) (4) (5) (6)
Disclosure 0.222** 0.219** 0.216** 0.135** 0.146** 0.155***
[0.0973] [0.0922] [0.0928] [0.0513] [0.0544] [0.0501]
beta -11.64 -11.74 -11.76 -6.255* -6.439* -6.461**
[8.524] [8.647] [8.523] [3.113] [3.200] [3.148]
Size 3.173* -0.277 2.358 3.135* 0.669 4.480**
[1.662] [0.849] [2.269] [1.635] [0.613] [1.836]
Leverage 0.109*** -2.413 -0.765 0.148*** -1.325 0.868
[0.0377] [1.533] [0.904] [0.0417] [1.126] [0.868]
Debt -4.303* -3.512 -2.086 -3.468
[2.507] [2.916] [2.556] [2.714]
Liability 2.507 0.704 -0.531 -0.859
[2.272] [2.366] [1.954] [2.175]
Capital intensity 1.807** 1.901** -0.0184 0.459
[0.854] [0.886] [0.594] [0.642]
Country fixed yes yes yes no no no

8
effects
Industry fixed
no no no yes yes yes
effects
R-squared 0.166 0.177 0.18 0.315 0.315 0.318
Notes:
Standard errors in brackets, *** p<0.01, ** p<0.05, * p<0.10.
The table reports coefficients of regressions of crisis-period stock performance on disclosure quality variable measured
by S&P for 885 firms from 43 countries and 60 industries. Stock performance is a minimum stock price over the crisis
period (June, 2007 - March, 2009) with June 2007 = 100. Firms with missing data on any control variable are
excluded. Firm size is the logarithm of total sales. Debt is the logarithm of LT debt, Capital intensity is the PPE and
Liability is the logarithm of liability. The leverage is LT debt over total sales. Beta is the market 1 year pre-crisis beta.
All firm variables correspond to pre-crisis period (as of end 2006).

In table 3 I provide results of regression where both country and industry fixed effects are
included. Notice that the sign of the coefficient on disclosure index remain same while it becomes
significant at 10% level. The reason why significance has fallen is including wide number of
industry and country dummies (totally 103) into regression. The magnitude of the coefficient
indicates that if firm’s Disclosure score decrease by 10 points its stock price falls by 1.4%.
Table 3
Stock prices during the crisis and level of transparency
Dependent variable: Stock price at the lowest point during financial crisis with June, 2007 = 100.
(1) (2) (3)
Disclosure 0.144* 0.144* 0.144*
[0.0872] [0.0851] [0.0846]
beta -7.739*** -7.739* -7.739
[1.477] [3.933] [7.042]
Size 4.408*** 4.408** 4.408*
[1.676] [1.980] [2.464]
Leverage -0.0111 -0.0111 -0.0111
[1.357] [0.917] [1.037]
Debt -1.677 -1.677 -1.677
[2.303] [2.850] [2.414]
Liability -2.323 -2.323 -2.323
[1.857] [1.660] [1.914]
Capital intensity 0.568 0.568 0.568

9
 [0.739] [0.575] [0.917]
Country fixed effects yes yes yes
Industry fixed effects yes yes yes
R-squared 0.333 0.333 0.138
Notes:
Standard errors in brackets, *** p<0.01, ** p<0.05, * p<0.10.
The table reports coefficients of regressions of crisis-period stock performance on disclosure quality variable
measured by S&P for 885 firms from 43 countries and 60 industries. Totally 103 country and industry dummies.
Stock performance is a minimum stock price over the crisis period (June, 2007 - March, 2009) with June 2007 = 100.
Firms with missing data on any control variable are excluded. Firm size is the logarithm of total sales. Debt is the
logarithm of LT debt, Capital intensity is the PPE and Liability is the logarithm of liability. The leverage is LT debt
over total sales. Beta is the market 1 year pre-crisis beta. All firm variables correspond to pre-crisis period (as of end
2006). First column presents specification with robust standard errors, second with clustered by country errors and
third with clustered by industry errors.

As the robustness check I will consider different specifications of basic model, such as
specification with weighted by countries observations, use alternative measures of disclosure
(Mitton 2002) and see stock performance at the single point in time, analyze pre-crisis years to see
whether disclosure factor is specific to the crisis or not.

3.2. Firm-level interactions

In this chapter I consider specification of the basic model with firm-level interactions in order to
specify to what kind of firms the level of disclosure is more important over the crisis period. Thus
regression specification becomes following:

Coefficient on interaction term show the significance of disclosure across firms with different level
of given control variable.
Table 4 represents the results of regressions with interaction terms. First column correspond
to specification with interaction between the disclosure quality and the size of the company. The
results are coincides with our preliminary hypothesis: investors know more about large firms and
effect of transparency is lower for such firms and higher for small firms. This hypothesis supported
by the negative and significant coefficient on interaction term between size and disclosure. Also I
expect that firms with higher debt and debt ratio are less sensitive to the transparency because such
firms suffer more from their high level of debt. During the crisis it becomes more and more difficult
to refinance debt, so investors looking at such firms more focus on their debt ratios and firms’
possibility to restructuring debt and less focus on level of transparency. Negative coefficients on

10
interactions in second and third columns confirm these hypotheses. Notice that all coefficients on
interaction terms are significant at 1% and 5 % level.
Table 4
Stock prices during the crisis, level of transparency and firm-level interactions
Dependent variable: Stock price at the lowest point during financial crisis with June, 2007 = 100.
(1) (2) (3)
Direct effect:
Disclosure 0.903* 0.168* 0.835**
[0.467] [0.0881] [0.340]
Interaction Disclosure with:
Size -0.0479*
[0.0271]
Leverage -0.0626***
[0.0226]
Debt -0.0469**
[0.0190]
beta, Size, Leverage, LT debt, Liability, Capital
Control variables intensity
Country fixed effects yes yes yes
Industry fixed effects yes yes yes
Notes:
Standard errors in brackets, *** p<0.01, ** p<0.05, * p<0.10. The table reports coefficients of regressions of crisis-
period stock performance on disclosure quality variable measured by S&P for 885 firms from 43 countries and 60
industries. Stock performance is a minimum stock price over the crisis period (June, 2007 - March, 2009) with June
2007 = 100. Firms with missing data on any control variable are excluded. Firm size is the logarithm of total sales.
Debt is the logarithm of LT debt, Capital intensity is the PPE and Liability is the logarithm of liability. The leverage is
LT debt over total sales. Beta is the market 1 year pre-crisis beta. All firm variables correspond to pre-crisis period (as
of end 2006). Interactions are variables counted by multiplying level of transparency and corresponding control
variable. Errors clustered by industry.

3.3. Country-level interactions

In chapter 4 I analyze country-level effects of level of corporate governance and quality of
institutions on stock market performance over the crisis period. I use broad number of measures of
corporate governance and institutions such as ICRG indexes, LLSV measures and different
measures of corruption. In this chapter I do not describe all these variables because it has been done
below in chapter 4.1.2. Here I focus on interactions between quality of firm’s disclosure practice
11
and country-level measures of corporate governance and institutions. Therefore the specification
here is following:

where Stock Perf. is a minimum firm’s stock price over the crisis period, from June 2007 through
March 2009 with June 2007 equal 100, Discl. is T&D Score, CVs are firm’s control variables and
CountryVar is country-level measure of corporate governance and quality of institutions. Analyzing
interactions between country-level and firm-level measures of corporate governance during the
crisis is interesting because it was known that in normal times country-level and firm-level
measures of corporate governance are substitutes, i.e. the higher country-level measure yields the
lower firm-level and opposite (Durnev and Kim, 2005). Here I can clarify is there any dependence
between firm- and country-level measures over the crisis and whether it coincide with results in
normal times.
Table 5 summarizes all regressions including firms’ control variables and all interactions
between firm disclosure quality and country-level measures of corporate governance. First I should
notice that the majority of interaction terms are significant and have similar positive signs. Second
is that coefficients which signs do not coincide with majority are insignificant. Final finding is that
size of country and its development impact differently on effect of being transparent for the firm. I
find strong evidence that level of transparency is more important in countries with better level of
corporate governance and quality of state institutions over the crisis period. Previous conclusion
keeps for broad range of variables that measures corporate governance and quality of state
institutions. Effect of being more transparent is more significant for firms in more developed
countries (with higher GDP p.c.).
Table 5
Stock prices during the crisis, level of transparency and country level interactions
Dependent variable: Stock price at the lowest point during financial crisis with June, 2007 = 100.
World Governance Indicators ICRG LLSV measures
Voice and
0.200* Bureaucracy 0.172** 0.0931***
Accountability Rule of law
Quality
[0.101] [0.0855] [0.0344]
Political Stability 0.164** 0.0198** Efficiency of 0.123**
Composite
judicial
[0.0725] Risk Rating [0.00887] [0.0509]
system
Government Repudiation
0.244*** Corruption 0.172** 0.162***
Effectiveness of contracts

12
 by
[0.0887] [0.0715] [0.0526]
government
Regulatory Quality 0.302*** Democratic 0.0385 0.164***
Risk of
Accountabilit
[0.0998] [0.0585] expropriation [0.0560]
y
Rule of Law 0.223*** Economic 0.0408** Accounting 0.00352
[0.0796] Rating [0.0169] Standards [0.00558]
Control of Corruption 0.204*** 0.0311 Disclosure 0.0287
Financial
requirements
[0.0722] Rating [0.0216] [0.955]
index
Size and development Government 0.158
Population -0.148*** Cohesion [0.178] -0.0019
Shareholders
[0.0423] Government 0.0368 [0.238]
GDP p.c. 0.219** Stability [0.106] Inside 0.345
[0.0918] 0.228*** ownership [0.524]
Law & Order
Corruption measures [0.0683] Irregular 0.15
Kaufmann corruption 0.194** Legislative 0.145 contracts [0.269]
index (2003-2005) [0.0731] Strength [0.201] 0.653*
Transparency Transactions
0.0873*** 0.0138** [0.354]
International Political Risk
corruption index Rating
[0.0313] [0.00649] Investor -0.377
(2003-2007)
Protection
Heritage Foundation 0.00850** [0.284]
corruption index
[0.00321]
(2003-2007)
GCR corruption index 0.196***
(2003-2007) [0.0705]
Firms that pay bribes 0.0118
(%) [0.0142]
Country fixed effects: yes Industry fixed effects: yes Errors clustered by industry.
Notes:
Standard errors in brackets, *** p<0.01, ** p<0.05, *p<0.10. The table reports coefficients of regressions of crisis-
period stock performance on disclosure quality variable measured by S&P and its interactions with country-level
measures of corporate governance and quality of government institutions for 885 firms from 43 countries and 60
industries. Each coefficient on interaction represents the results of regression of firm’s stock price performance over
13
crisis period on level of disclosure, all control variables and one interaction term. Stock performance is a minimum
stock price over the crisis period (June, 2007 - March, 2009) with June 2007 = 100. Firms with missing data on any
control variable are excluded. Firm size is the logarithm of total sales. Debt is the logarithm of LT debt, Capital
intensity is the PPE and Liability is the logarithm of liability. The leverage is LT debt over total sales. Beta is the
market 1 year pre-crisis beta. All firm variables correspond to pre-crisis period (as of end 2006). Measures of a
corporate governance and quality of the state institutions as of June, 2007, are presented by agency ICRG, WGI
agency and LLSV papers. The higher value of an index, the higher level of a corporate governance and quality of the
state institutions. Measures of corruption counted for the historical period prior to crisis (2003-2007) by different
agencies. The higher value of an index, the less corrupted country, except variable which is responsible for number of
firms that pay bribes. Fewer firms pay bribes, less corrupted country. All firm controls are included. Description of
country-level corporate governance measures see in chapter 4.1.2 and in LLSV (1997).

3.4. Robustness check

3.4.1. Observations weighted by country
In this chapter I try to correct our basic econometric specification controlling on how many firms
represented from each country and industry. I use country and industry fixed effects in all our
regressions. Here I try to weight observations proportionally to country and proportionally to
industry, e.g. firms from country which represent majority firms get less weight. Table 6 presents
regressions results. First column correspond to the specification with weighted by country
observations and second is a specification with weighted by industry observations. Notice that
Disclosure index becomes larger in absolute value and more significant with respect to
specifications without weights (see table 3). Thus reduction in Disclosure index by 10 points will
decrease stock price by 1.5 - 2 percent over the crisis period (in previous specification I get 1.4%
falling in stock prices).
Table 6
Stock prices during the crisis and level of transparency (robustness check)
Dependent variable: Stock price at the lowest point during financial crisis with June, 2007 =
100.
(1) (2)
Disclosure 0.194** 0.153*
[0.0828] [0.0799]
beta -5.909** -5.013
[2.895] [5.591]
Size 3.834* 4.880*
[1.930] [2.830]
Leverage -0.192 0.301

14
 [1.033] [1.024]
LT Debt -0.711 -2.845
[2.667] [3.070]
Liabilities -3.274** -1.518
[1.336] [2.118]
Capital intensity 1.072* 0.662
[0.547] [0.945]
R-squared 0.354 0.128
Country fixed effects yes yes
Industry fixed effects yes yes
Errors clustered by country industry

Notes:
Standard errors in brackets, *** p<0.01, ** p<0.05, * p<0.10. The table reports coefficients of regressions of
crisis-period stock performance on disclosure quality variable measured by S&P for 885 firms from 43
countries and 60 industries. Stock performance is a minimum stock price over the crisis period (June, 2007 -
March, 2009) with June 2007 = 100. Firms with missing data on any control variable are excluded. Firm size
is the logarithm of total sales. Debt is the logarithm of LT debt, Capital intensity is the PPE and Liability is the
logarithm of liability. The leverage is LT debt over total sales. Beta is the market 1 year pre-crisis beta. All
firm variables correspond to pre-crisis period (as of end 2006). First column is a regression where observations
weighted by country and errors clustered by country; second column represent regression where observations
weighted by industry and errors clustered by industry

3.4.2. Alternative measures of disclosure and dependent variable

Mitton (2002) as a measure of disclose quality used only two indicators which are responsible for
whether firm have BIG6 auditor and issue American Depository Receipts (ADR). As robustness
check here I try to use quality of disclose measured by BIG4 indicator which is equal 1 if firm
audited by BIG4 company and 0 otherwise. Results are presented in first column of table 7.
Coefficient on BIG4 indicator is significant and has a positive sign which additionally indicate
about validity of our results. If firm pursue assurance services from BIG4 company firm’s value, on
average, increase by 4.7% over the crisis period.
Table 7  
Stock prices during the crisis and level of transparency (robustness
check)
Dependent variable:
Stock price at the lowest point during financial Stock price at the end of 2008 with
15
 crisis with June, 2007 = 100. June, 2007 = 100.
big4 auditor 4.669** Disclosure 0.215**
[2.227] [0.101]
beta -5.197* beta -5.258
[2.812] [3.128]
Size 2.566 Size 1.663
[1.633] [1.455]
Leverage -0.15 Leverage -0.498
[0.803] [0.905]
LT Debt -0.33 LT Debt 1.12
[2.470] [2.221]
Liabilities -2.613* Liabilities -3.104*
[1.368] [1.628]
Capital intensity 1.239** Capital intensity 1.349**
[0.594] [0.647]
Country fixed
yes Country fixed effects yes
effects
Industry fixed
yes Industry fixed effects yes
effects
Notes:
Standard errors in brackets, *** p<0.01, ** p<0.05, * p<0.10. The table reports coefficients of regressions
of crisis-period stock performance on disclosure quality variable measured by S&P for 885 firms from 43
countries. Stock performance is a minimum stock price over the crisis period (June, 2007 - March, 2009)
with June 2007 = 100. Firms with missing data on any control variable are excluded. Firm size is the
logarithm of total sales. Debt is the logarithm of LT debt, Capital intensity is the PPE and Liability is the
logarithm of liability. The leverage is LT debt over total sales. Beta is the market 1 year pre-crisis beta. All
firm variables correspond to pre-crisis period (as of end 2006). First column is a regression where the main
interest variable is an indicator which is equal 1 if firm is audited by BIG4 auditor (KPMG, Deloitte, Ernst
and Young, PWC) and 0 otherwise. Second column represent regression with alternative dependent
variable which is firms stock price performance at the single point. Here I use end of 2008 as one general
single point to all firms. Observations are weighted by country.

Taking minimum price of stock value during the crisis period is not exceptional, but here I
try to use alternative measure of impact of the crisis on a firm. Again I consider falling of stock
prices, but as the starting point I consider same point as previous (1st June, 2007) and look at price
decline at the single for all firms point in time. As this single point I take end of 2008, i.e. last close
16
price in December, 2008. Second column in table 7 presents results of regressions with this
dependent variable. Notice that the value and the sign of the coefficient on disclosure index remain
same and significance of the coefficient is at 5% level.

3.5. Pre-crisis analysis

In this chapter I focus on analyzing pre-crisis period in order to understand whether our results hold
during the crisis or in normal times too. To see how the dynamics of stock prices depend on
disclosure quality in the pre-crisis period I consider same sample but as the dependent variable I use
pre-crisis price dynamic. I take Jan, 2006 as a baseline and look at a maximum fall of the stock
price during the Jan, 2006 – Jun, 2007 period which is become a new dependent variable. Table 8
presents results of regression in two different specifications (different types of errors).
The most unexpected result is that disclosure quality does not have significant effect on
stock prices over pre-crisis period from what I can conclude that quality of transparency has effect
only over crisis-period. This might be strange because there were broad range of literature
indicating the significance of quality of corporate governance on firm value. As an explanation to
these findings might be the existence of speculative “bubbles” on many markets (oil, household,
metals) in this specific period and no doubt it has impact on stock markets, where investors act
more in speculative manner and likely ignore fundamental factors such as disclosure quality.
Table 8
Stock prices prior the crisis period and level of transparency (robustness check)
Dependent variable: Stock price at the lowest point during prior to the
financial crisis (Jan, 2006 – June, 2007) with January, 2006 = 100.
(1) (2)
Disclosure -0.00683 -0.00683
[0.0687] [0.0667]
beta 1.522 1.522
[0.944] [1.184]
Size 2.697 2.697
[2.434] [1.656]
Leverage 1.743 1.743***
[1.044] [0.612]
LT Debt -3.859 -3.859*
[3.078] [2.170]
Liabilities 0.447 0.447

17
 [1.518] [1.545]
Capital intensity 1.285 1.285
[0.775] [0.812]
Country fixed effects yes yes
Industry fixed effects yes yes
Errors clustered by country industry
Notes:
Standard errors in brackets, *** p<0.01, ** p<0.05, * p<0.10. The table reports coefficients of
regressions of pre-crisis-period stock performance on disclosure quality variable measured by S&P
for 885 firms from 43 countries and 60 industries. Stock performance is a minimum stock price
over the pre-crisis period (Jan, 2006 - Jun, 2007) with Jan 2006 = 100. Firms with missing data on
any control variable are excluded. Firm size is the logarithm of total sales. Debt is the logarithm of
LT debt, Capital intensity is the PPE and Liability is the logarithm of liability. The leverage is LT
debt over total sales. Beta is the market 1 year pre-crisis beta. All firm variables correspond to pre-
crisis period (as of end 2006).

4. Country-level analysis
4.1. Variable descriptions and econometric issues
4.1.1. Measuring the crisis
Financial crisis 2007-2009 has impacted on many countries. Following the previous literature as a
measure of influence of financial crisis on the country I use the value of MSCI index at the
minimum point during June, 2007 – December, 2008 (at the beginning of the period the index is
supposed equal 100). The data consists of the day market indexes counted by Morgan Stanley
Capital International. Start date is June, 2007 coincides with the crisis beginning in the market of
sub-prime mortgages in the USA. The historical dynamic of indexes for four countries is shown at
the figure 3. Values of MSCI indexes at the minimum point are resulted in table 9.
Figure 3 MSCI Index for China, USA, Germany and Russia during the financial crisis

18
 4.1.2. Measuring institutions
International Country Risk Guide (ICRG) makes a monthly rating of the countries in three
categories of risk: policy, finance, economy. Table 10 shows the basic components of a category of
political risk. Final country risk index consist of three basic type of risk where the political risk has
the largest weight (50 %), and financial and economic parts add 25 % each. Notice that financial
and economic ratings do not represent a particular interest as they are made exclusively of the
analysis of macroeconomic variables which are supposed to be used as control variables.
I am interested in the first group - political risks which define various aspects of corporate
governance and quality of the state institutes in the country. Measures of risk ICRG are considered
prior the crisis.
The largest composite rating of ICRG belongs to Norway and Switzerland whereas the least
risky countries from the point of view of political risk are Finland and Sweden. The countries with
the highest political risk are Nigeria and Pakistan.
Table 9 MSCI Index value at lowest point during June, 2007 - December, 2008 with June,
2007 = 100.
MSCI MSCI MSCI
Country Country Country
Index Index Index
SLOVENIA 41.9 NIGERIA 39.3 AUSTRIA 21.9
UNITED
KOREA, SOUTH 34.8 PAKISTAN 24.2 41.8
KINGDOM
BAHRAIN 50.9 INDIA 43.7 FINLAND 41.4
NEW ZEALAND 35.8 INDONESIA 41.1 SINGAPORE 44.4
SPAIN 47.6 UKRAINE 12.5 IRELAND 18.2
GREECE 33.2 PHILIPPINES 43.7 SWEDEN 34.0
ITALY 39.5 CHINA 48.0 DENMARK 46.3
KUWAIT 46.6 EGYPT 49.8 HONG KONG 50.7
AUSTRALIA 40.3 KAZAKHSTAN 31.9 SWITZERLAND 53.4
UNITED
FRANCE 43.6 BULGARIA 17.1 48.8
STATES
BELGIUM 23.7 ROMANIA 23.9 JAPAN 54.1
GERMANY 45.5 THAILAND 49.6 MALAYSIA 55.5
UNITED ARAB
30.1 PERU 43.1 TURKEY 34.6
EMIRATES
CANADA 46.4 SOUTH 41.3 CROATIA 35.0

19
 AFRICA
NETHERLANDS 42.5 BRAZIL 45.2 POLAND 38.0
CZECH
PORTUGAL 39.8 53.8 CHILE 55.1
REPUBLIC
NORWAY 30.6 ARGENTINA 32.3 HUNGARY 29.4
MEXICO 39.5

Table 10 Political risk components and their points

Political Risk Components 
Points
Component (max.)
Government Stability 12
Socioeconomic Conditions 12
Investment Profile 12
Internal Conflict 12
External Conflict 12
Corruption 6
Military in Politics 6
Religious Tensions 6
Law and Order 6
Ethnic Tensions 6
Democratic Accountability 6
Bureaucracy Quality 4
Total 100

Worldwide Governance Indicators (WGI) annually counts up the aggregated and private indicators
of corporate governance in 212 countries and territories during the period since 1996 on 2007. Basic
components of WGI measures are:
- Voice and Accountability
- Political Stability and Absence of Violence
- Government Effectiveness
- Regulatory Quality
- Rule of Law
- Control of Corruption

20
On an official web site of the agency it is specified that the given indicators are made of surveys of
firms, citizens of the countries, experts.
In addition to the measures set above, I consider the measures made in following articles:
- «Law and Finance» (with R. La Porta, F. Lopez-de-Silanes, and R. Vishny). Journal of
Political Economy, December 1998
- «Investor Protection and Corporate Governance» (with R. La Porta, F. Lopez-de-Silanes,
and R. Vishny), Journal of Financial Economics, October, 2000
- «What Works in Securities Laws?» (with R. La Porta, and F. Lopez-de-Silanes), Journal of
Finance, February, 2006
- «Disclosure by Politicians» (with with S. Djankov, R. La Porta, and F. Lopez-de-Silanes),
January, 2009
Authors of specified articles consider measures of quality of legislative system, corruption level in
the country, norms of the right and corporate governance.
As measures of corruption and norms of right LLSV used measures presented by agency
ICRG (therefore the given indicators are excluded from consideration), and also the measures
constructed by various economists and agencies (Kauffman, Corruption Index, Transparency
International Corruption Index, Heritage Foundation, Corruption Index, GCR Corruption Index).
The level of corporate governance is a measure calculated by authors and estimating "the
real" rights of shareholders. Also, LLSV consider a measure of protection of the rights of creditors
and a measure of development of standards of the financial reporting in the country. To estimate
influence of an openness of politicians in the country, LLSV used the measures showing how much
requirements on disclosing of the information to a society politician are fulfilled.
In tables 16-17 (in the appendix) variables and their descriptions are summarized.

4.1.3. Control variables

As the basic control variables I consider main macroeconomic variables: Cash surplus/deficit (% of
GDP), Central government debt, total (% of GDP), GDP (constant 2000 US $), GDP per capita,
PPP (constant 2005 international $), Market capitalization of listed companies (% of GDP),
Population, total, Total reserves (% of external debt), Current account balance (% of GDP),
External debt, total (DOD, current US $), Expense (% of GDP). Data is given by the international
statistical agency WorldBank (http://www.worldbank.com). In table 3 the basic descriptive statistics
of macroeconomic variables are resulted.

21
 Table 11 Summary statistics of macroeconomic variables
Government GDP Market
Current Surplus/Deficit GDP Population,
Statistics Debt of (2000=constant) Cap to
Account of GDP p.c. ths
GDP bln USD GDP

mean 1.7 0.8 51.2 700 15,5 113 93,200

sd 11.4 5.3 25.2 1,760 12,8 89 237,000
min (21.7) (8.6) 7.1 11 0.47 27 752
max 49.9 21.3 109.8 11,600 42,4 562 1,320,000
Note: Data from Worldbank as of 2007

4.1.4. Final sample and econometric models

At the first stage "clarification" of the data has been done. I excluded countries for the following
reasons:
- MSCI availability: at the moment MSCI index counts for 69 countries in the world. I
excluded indexes which have started to count after the considered period
- Development of financial market: market capitalization to GDP on the end of 2007 less than
25 % (an average of sample is 113 %, a standard deviation of 89 %)
- Size of country: Gross national product per capita on the end of 2007 less than 450 US
dollars
Finally there are 53 countries in our sample.
The indicator of debt is available only for 32 countries. Then it has been done robustness
check whether this variable influence on signs and significance of main interest variables, i.e.
measures of corporate governance, quality of institutions. Notice that inclusion/exclusion of a
variable of debt does not change our results.
In figures 4 - 7 it is shown the dependence of index MSCI on various measures of a
corporate governance and quality of the state institutes. Basically, all factors are significant and
have that sign which corresponds to the basic hypothesis - in the countries with lower level of the
state institutes and corporate governance impact of the financial crisis was reflected more
significant, than in the countries with better level. Using various measures yields similar results that
convinces of fidelity of the given hypothesis even more.

22
Figure 4 Country MSCI index at the minimum point Figure 5 Country MSCI index at the minimum point
during the crisis period and ICRG index of government during the crisis period and LLSV measure of accounting
stability standards

Figure 6 Country MSCI index at the minimum point Figure 7 Country MSCI index at the minimum point
during the crisis period and LLSV measure of Disclosure during the crisis period and LLSV measure of investor
requirements protection

The basic specification is linear regression of a dependent variable on measures of a

corporate governance and quality of institutes, and also on control variables:

where MSCI is a MSCI index at the lowest point during financial crisis, Z is main variable of
interest (the level of corporate governance, quality of institutions) and CV’s are control variables.
The main hypothesis is γ > 0.

4.2. County-level results

The analysis of dependence of index MSCI from the main macroeconomic variables shows that the
market index significantly correlates with such variables, as gross national product, the current
account and market capitalization to GDP. These variables explain a country position in the world
development. With such variables as a public debt, reserves and deficit significant correlation
dependence is absent. The second group of variables explains current condition of the country and
does not show fundamental situation and possibility of country to struggle with crisis. From table 6
23
where results of regression are presented, it can be drawn that any of variables from group two is
not significant. Notice that R-squared in the given regressions is approximately equal 10%.
Table 12
MSCI Index and macroeconomic variables
Dependent variable: MSCI index at the lowest point during financial crisis with June, 2007
= 100.
GDP (log) 2.150**
[1.002]
Government Debt (log) 0.757
[0.685]
Current Account (log) 0.353***
[0.121]
Market Capitalization-to-
0.0436***
GDP
[0.0155]
Surplus/Deficit to GDP 0.122
[0.302]
Total Reserves 0.0203
[0.021]
Constant -16.23 19.75 39.50*** 35.01*** 39.03*** 37.06***
[26.19] [18.77] [1.375] [2.199] [1.586] [3.005]
Observations 53 32 52 53 46 24
R-squared 0.083 0.039 0.145 0.134 0.004 0.043
Notes:
Standard errors are in
brackets
*** p<0.01, ** p<0.05,
* p<0.1
Macroeconomic variables are given from Worldbank as of end 2006.

Next I analyze influence of corporate governance and quality of institutions given by ICRG agency
on stock market value (see table 13). Composite risk rating is significant at 5 % level and if the
country rating rises on 2 units the market index will increase approximately on 1 %. Most
substantial growth in dynamics of a market index in crisis is caused by the variables which are

24
responsible for the government, namely its unity and stability. Reduction of the risk connected with
Government Cohesion by 1 unit will add to stock market index about 6 %, and with Government
Stability, about 1.5 %. Financial and economic ratings are significant, but they are not interesting to
consideration. Notice that R-squared has considerably increased after inclusion the measure of a
corporate governance and quality of the government institutions (from 0.1 to 0.4). Many variables
have a sign which supports the basic hypothesis. The corruption index variable is significant on 15
% level that also indicates about importance and influence of the given indicator. Below I will
consider alternative measures of corruption and will get more significant results.
Table 13
MSCI Index and ICRG
Dependent variable: MSCI index at the lowest point during financial crisis with June,
2007 = 100.
Government
Bureaucracy Quality (L) 1.706 Cohesion 5.806***
[2.275] [2.022]
Composite Risk Rating 0.440** Government Stability 1.476*
[0.205] [0.851]
Corruption 2.319 Law & Order 1.515
[1.697] [1.546]
Democratic Accountability -0.349 Legislative Strength -0.25
[1.078] [1.948]
Economic Rating 0.628* Financial Rating 0.700**
[0.319] [0.277]
Macroeconomic control variables
GDP (log)
Market Capitalization-to-
GDP
Notes:
Standard errors are in brackets, *** p<0.01, ** p<0.05, * p<0.1
In the table presented results of separate OLS regression of a dependent variable on one ICRG index
including control variables and constant. Macroeconomic variables are given from Worldbank as of end
2006. Measures of a corporate governance and quality of the state institutions as of June, 2007, are
presented by agency ICRG. The higher value of an index, the higher level of a corporate governance and
quality of the state institutions. Dummy-variables on market type are included (developed, developing,
frontier according to MSCI classification). Number of observations is 52. R-squared range from 0.3 to 0.4.

25
The analysis of LLSV measures (1998) is presented in table 14. Signs on all factors coincide with
the basic hypothesis. However, the majority of factors turn out insignificant. It is possible to explain
it to that LLSV indexes are accessible only to 35 countries of sample and, thereby, factors and
standard errors can are estimated incorrectly because of selection bias. The factor which is
responsible for development of the financial reporting in the country appears very significant. The
increase at 10 units of the given indicator (it varies from 0 to 83 across countries) will allow
increasing a market index approximately by 2 %. It partially confirms a hypothesis that the more
company disclose the information, the less investors "withdraw" assets from it, hence, the more
slightly market index falling. The factor which is responsible for the law on disclosure of names
shareholders appears significant. Acceptance of the given law, will allow increasing an index by
8.5% during the crisis.
Table 14
MSCI Index and LLSV measures
Dependent variable: MSCI index at the lowest point during financial crisis with June,
2007 = 100.
Anti-director Rights 0.832 Shareholders 8.487*
[1.218] [4.225]
Creditor Rights 0.546 Inside ownership -2.931
[1.271] [4.318]
Efficiency of judicial system 1.136 Irregular contracts 1.22
[1.005] [3.449]
Repudiation of contracts by
3.179* Transactions 2.005
government
[1.888] [4.302]
Disclosure
Risk of expropriation 3.124 requirements 6.682
index
[2.219] [7.436]
Investor
Accounting Standards 0.199*** 4.873
Protection
[0.0721] [5.930]
Macroeconomic control variables
GDP (log), Market

26
 Capitalization-to-GDP
Notes:
Standard errors are in brackets, *** p<0.01, ** p<0.05, * p<0.1
In the table presented results of OLS regression of a dependent variable on one measure of corporate
governance including control variables and constant. Macroeconomic variables are given from
Worldbank as of end 2006. Measures of a corporate governance and quality of the state institutions as of
June, 2007, are presented by LLSV papers. The higher value of measure, the higher level of a corporate
governance and quality of the state institutions in the country. The explanation of each variable is
presented at the appendix in tables 16-17. Dummy-variables on market type are included (developed,
developing, frontier according to MSCI classification). Number of observations is 35.

The analysis of alternative measures of corruption is presented in the table 15. Signs on all factors
coincide with the basic hypothesis and the majority of them turn out significant at 10 % level.
Reduction of corruption by 1 unit (according to Transparency Corruption Index) will allow
receiving 2% more on a market index. It is interesting that the factor at a variable of number of
firms which give bribes, turns out significant. If the number of firms in the country which practice
bribes falls at least by 10 % on the average market index will earn on 1.5 % more. R-squared in the
given specifications also increases considerably (from 0.1 to 0.36).
Table 15
MSCI Index and Corruption measures
Dependent variable: MSCI index at the lowest point during
financial crisis with June, 2007 = 100.
Kaufmann corruption 3.747
index (2003-2005) [2.328]
Transparency
1.884*
International corruption
index (2003-2007) [1.087]
Heritage Foundation
0.189*
corruption index (2003-
2007) [0.104]
GCR corruption index 3.844
(2003-2007) [2.296]
Firms that pay bribes -0.137*
(%) [0.0706]
Macroeconomic control variables

27
 GDP (log) 2.248* 2.342* 2.337* 2.282* 2.126*
[1.248] [1.252] [1.239] [1.247] [1.194]
Market Capitalization-
to-GDP 0.0474** 0.0454* 0.0439* 0.0435* 0.0592***
[0.0229] [0.0231] [0.0231] [0.0238] [0.0205]
R-squared 0.344 0.35 0.354 0.347 0.36
Notes:
Standard errors are in brackets
*** p<0.01, ** p<0.05, * p<0.1
In the table presented results of OLS regression of a dependent variable on one measure of corruption
including control variables and constant.
Macroeconomic variables are given from Worldbank as of end 2006.
Measures of corruption counted for the historical period prior to crisis (2003-2007) by different agencies. The
higher value of an index, the less corrupted country, except variable which is responsible for number of firms
that pay bribes. Fewer firms pay bribes, less corrupted country.
Dummy-variables on market type are included (developed, developing, frontier according to MSCI
classification). Number of observations is 50.

5. Conclusion
In this paper I analyze how the decline in stock prices for firms depend on their disclosure
practice as measured by Standard & Poor firm transparency score. I use transparency score that was
constructed prior to the beginning of the crisis, which allows us to deal with some potential
simultaneity problems. Then I measure the magnitude of stock market decline across 53 counties
and examine how it depends on various measures of the quality of political and financial institutions
in a country.

First, I show that the results in Johnson et al. (2000) and Mitton (2002) hold for the current
crisis. As in the Asian financial crisis countries with better corporate governance and firms with
better disclosure practices experience smaller decline in firm value during the crisis. An important
improvement in our work over the previous studies is that this crisis allowed us to look more
broadly at country effects, as this crisis, unlike Asian financial crisis, now affected countries with
different levels of economic development and different political, financial, and legal institutions.
Second, I find that at the firm-level this relationship is more important for smaller and less indebted
firms. I also find that the effect of firm-level disclosure practices is more important for firms located
in countries with better quality of institutions. The last results indicate that during the current crisis
country- and firm-level corporate governance become complements, unlike normal times when they
work as substitutes.
28
 The financial crisis was an unprecedented event that negatively affected many firms and
countries. Therefore studying crisis becomes an important challenge to many economists. The
results of our paper contribute to previous works on analyzing an importance of corporate
governance and disclosure quality to countries and firms.

29
6. References
[1] Bekaert, Geert, Harvey, Campbell R., Lundblad, Christian T. and Siegel, Stephan, 2008.
What Segments Equity Markets? AFA 2009 San Francisco Meetings Paper.

[2] Caballero, R., Krishnamurthy, A., 1998. Emerging markets crises: an asset markets
perspective. Unpublished working paper. MIT, Cambridge, MA.

[3] Campbell, J., 1996. Understanding risk and return. Journal of Political Economy 104, 298-
345.

[4] Chen, Wei-Peng, Chung, Huimin, Lee, Chengfew and Liao, Wei-Li, 2007. Corporate
Governance and Equity Liquidity: Analysis of S&P Transparency and Disclosure Rankings.
Corporate Governance: An International Review, Vol. 15, No. 4, 644-660

[5] Durnev and Kim, 2005. To Steal or Not to Steal: Firm Attributes, Legal Environment, and
Valuation. Journal of Finance 60, 1461-1493

[6] Glaeser, E., Johnson, S., Shleifer, A. 2001. Coase versus The Coasians. Quarterly Journal of
Economics, forthcoming.

[7] Johnson, Simon H., Boone, Peter D., Breach, Alasdair and Friedman, Eric, 2000. Corporate
Governance in the Asian Financial Crisis. Journal of Financial Economics, Vol. 58, No. 1-2

[8] Johnson, S., La Porta, R., Lopez-de-Silanes, F., Shleifer, A. 2000. Tunnelling. American
Economic Review, May 90(2) 22-27.

[9] La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 1999. Corporate ownership around the
world. Journal of Finance 54, 471-517.

[10] La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 1997. Legal determinants of
external Finance. Journal of Finance 52, 1131-1150.

[11] La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 1998. Law and Finance. Journal
of Political Economy December 106, 1113-1155.

[12] La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 1999. The quality of
government. Journal of Law, Economics and Organization 15, 222-279.

[13] La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R. 1999. Investor protection and
corporate valuation. Unpublished working paper.

[14] La Porta, Rafael, Florencio Lopez-de-Silanes, Andrei Shleifer, and Robert Vishny, 2002.
Investor protection and corporate valuation, Journal of Finance 57, 1147–1170.

[15] Mitton, Todd, 2001. A Cross-Firm Analysis of the Impact of Corporate Governance on the
East Asian Financial Crisis. AFA 2001 New Orleans.

[16] Nicola, Gianni De, Luc Laeven, and Kenichi Ueda, 2007. Corporate Governance Quality:
Trends and Real Effects. IMF Working Papers.

30
[17] Shleifer, A., Vishny, R. 1997. A Survey of Corporate Governance. The Journal of Finance,
Vol. LII, No.2, 737-783.

[18] Shleifer, A., R. La Porta, and F. Lopez-de-Silanes, 2006. What Works in Securities Laws?
Journal of Finance, February

[19] Shleifer, A., S. Djankov, R. La Porta, and F. Lopez-de-Silanes, 2009. Disclosure by

Politicians.

[20] Shleifer, Andrei, and Daniel Wolfenzon, 2002. Investor protection and equity markets.
Journal of Financial Economics 66, 3–27.

[21] World Bank, http://www.worldbank.org/

[22] International Country Risk Guide (ICRG), http://www.prsgroup.com/

31