MATERIA¸Y I STUDIA
Paper No. 36
ECMOD
Model of the Polish Economy
Tatiana Fic, Marcin Kolasa, Adam Kot,
Karol Murawski, Micha∏ Rubaszek, Magdalena Tarnicka
Warsaw, October 2005
Authors work in the National Bank of Poland (Macroeconomic and Structural Analyses
Departament).
Tatiana Fic – project manager; correspondence e-mail address: tatiana.fic@mail.nbp.pl
Design:
Oliwka s.c.
Layout and print:
NBP Printshop
Published by:
National Bank of Poland
Department of Information and Public Relations
00-919 Warszawa, 11/21 Âwi´tokrzyska Street
phone: (22) 653 23 35, fax (22) 653 13 21
© Copyright by the National Bank of Poland, 2005
http://www.nbp.pl
Acknowledgements
The authors would like to thank the following persons for their
suggestions and remarks concerning both the current and previous versions of
the model:
Piotr Boguszewski, Wojciech Charemza, Adam Czy˝ewski, Ryszard
Kokoszczyƒski, Krzysztof Marczewski, Witold Or∏owski, Krzysztof Rybiƒski,
Krystyna Strza∏a, Piotr Szpunar, Aleksander Welfe, Leszek Zienkowski, and the
heads of the divisions involved in work on the model – Jakub Borowski,
Ma∏gorzata Golik, Gra˝yna Mierzejewska, Ewa Rzeszutek, Zbigniew ˚ó∏kiewski
and Roman Sawiƒski.
In particular, the authors would like to thank Micha∏ Gradzewicz who codeveloped the previous version of the model. The model would not have the
present form were it not for Micha∏’s contribution.
The authors would also like to express their thanks to Tomasz Kuciƒski for
his help in the preparation of this paper.
The authors bear full responsibility for any errors.
MATERIA¸Y I STUDIA – PAPER 36
3
Contents
Contents
Non-technical summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1 General description of the ECMOD model . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.1 ECMOD construction methodology, forecasting and simulation methodology . . . . 11
1.2 ECMOD - theoretical aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2 Modules of the ECMOD model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1 Real economy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.1.1 Supply side . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.1.2 Demand side . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.1.2.1 Individual consumption (by households) . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.1.2.2 Disposable income . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.1.2.3 Fixed capital formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.2 Prices and costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.2.1 CPI index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.2.2 GDP deflator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.2.3 Wages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.2.4 Exchange rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.2.5 Interest rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.3 External sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.3.1 Balance of payments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.3.2 Foreign trade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.3.2.1 Exports of goods and services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.3.2.2 Imports of goods and services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.3.2.3 Export deflator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.3.2.4 Import deflator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.4 General government . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.4.1 GG revenues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.4.1.1 Indirect tax revenues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.4.1.2 PIT revenues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.4.1.3 CIT revenues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.4.1.4 Revenues from health and social insurance contributions . . . . . . . . . . . . . . 42
2.4.1.5 Customs revenues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2.4.1.6 Real estate tax revenues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.4.1.7 NBP profit transfer and other revenues . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.4.2 GG expenditure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.4.2.1 Current expenditure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.4.2.2 Transfers to the population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.4.2.3 Costs of debt service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.4.2.4 Investment expenditure, subsidies to enterprises and other expenditure . . . . 47
4
N a t i o n a l
B a n k
o f
P o l a n d
Contents
2.4.3 GG deficit and public debt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3 Simulation analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.1 Monetary policy rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.2 Fiscal policy rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.3 Forecast properties of the model - simulations in- and out-of-sample . . . . . . . . . . 52
3.4 Interest rate impulse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3.5 Exchange rate impulse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.6 Fiscal impulse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
5 Selected references . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
6 Annex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Annex 2. “Model-experts” interation method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
Annex 3. Long-term equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Annex 4. Cointegration analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Annex 5. Definitions of variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
MATERIA¸Y I STUDIA PAPER 36
5
Non technical summary
NON-TECHNICAL SUMMARY
This paper describes the multi-equation macroeconomic model of the Polish
economy (ECMOD) developed for the purpose of conducting monetary policy in
Poland.
Under the direct inflation targeting strategy, the assessment of inflation
prospects, including projection of inflation, is an important factor in the decisionmaking process. The significance of forecasting in contemporary monetary policy is
reflected in the words of Alan Greenspan, Federal Reserve Chairman:
„Implicit in any monetary policy action or inaction is an expectation of how the future
will unfold, that is, a forecast” [Greenspan, 2002].
Macroeconomic models systematise the way of thinking about complex
economic reality and enable economists to identify factors relevant to the explanation
of the mechanisms governing the economy in the future.
The ECMOD model has been developed to generate forecasts for all key
macroeconomic categories, mainly inflation, as well as GDP and its components and
many other variables relevant to monetary policy. Another objective of the ECMOD
model is to conduct simulations that enable quantification of the effects of economic
policy and exogenous shocks to the main macroeconomic variables.
The ECMOD model reflect the macroeconomic processes observed in Poland
during the sample period. At the same time, in describing the functioning of the
economy, it refers to the contemporary macroeconomic achievements. Consequently,
the ECMOD model combines the information coming from the empirical data with
theoretical relationships, and thus it belongs to the class of hybrid models.
Like most models used in central banks to pursue monetary policy, the ECMOD
model’s theoretical fundamentals include both classical and keynesian elements. In the
short-term, economic growth is mainly determined by demand factors, while, in thelong term, it is shaped by factors related to the supply-side of the economy.
The model describes the economy in a way that resembles the manner of
reporting macro-processes in the national accounts. The ECMOD model differentiates
among six types of entities: households, enterprises, financial intermediaries, monetary
authorities, fiscal authorities, and foreign countries.
6
N a t i o n a l
B a n k
o f
P o l a n d
Non technical summary
7KHILJXUH)LJSUHVHQWVWKHVWUXFWXUHRIWKH(&02'PRGHOEURNHQGRZQLQWR
PRGXOHV
)LJ0RGXOHVRIWKH(&02'PRGHO
35,&(6 &2676
&RUHLQIODWLRQ
*'3GHIODWRU
5($/(&2120<
3RWHQWLDOSURGXFW
:DJHV
([FKDQJHUDWH
*'3 FRQVXPSWLRQ
*URVVFDSLWDOIRUPDWLRQ
IRUHLJQWUDGHEDODQFH
5HYHQXH
([SHQGLWXUH
'HILFLW
'HEW
(&02'
2XWSXWJDS
(PSOR\HGSHUVRQV
8QHPSOR\PHQWJDS
GHYLDWLRQRIXQHPSO
UDWHIURP1$:58
*(1(5$/
*29(510(17
(;7(51$/6(&725
([SRUWV
,PSRUWV
([SRUW GHIODWRU
,PSRUW GHIODWRU
58/(6
0RQHWDU\SROLF\
UXOH
)LVFDOSROLF\
UXOH
6RXUFH 1DWLRQDO%DQNRI3RODQG
7KH (&02' LV EDVHG RQ D TXDUWHUO\ GDWD DQG FRQVLVWV RI PRUH WKDQ
HTXDWLRQV ,QPRVW FDVHV EHKDYLRXUDO HTXDWLRQ SDUDPHWHUV DUH HVWLPDWHG VRPHWLPHV
KRZHYHUWKH\DUHFDOLEUDWHG7KH(&02'PRGHOHQFRPSDVVHVGDWD IURP4 WR
4
*LYHQ WKH VLPSOLILHG E\ GHILQLWLRQ PDSSLQJ RI WKH UHDOLW\ LQ PRGHOV WKH
DVVHVVPHQWVRIH[SHUWVDUHRIJUHDWLPSRUWDQFHGXULQJWKHSURMHFWLRQSURFHVV7KHUHIRUH
HVWDEOLVKLQJ DIUDPHZRUN WR HQVXUH FRKHVLRQ RI IRUHFDVWV REWDLQHG ERWK ZLWKLQ DQG
RXWVLGHRIWKHPRGHOZDVRQHRIWKHFULWHULDIRUWKHPRGHO¶VGHYHORSPHQW&RQVHTXHQWO\
D IXOO\ LQWHJUDWHG IRUHFDVW LV DFKLHYHG DV D UHVXOW RI FRPELQLQJ LQIRUPDWLRQ IURP WKH
PRGHODQGH[SHUWV
7KLV SDSHU GHVFULEHV WKH PHWKRGRORJ\ XVHG LQ FRQVWUXFWLQJ WKH PDFURPRGHO
JHQHUDWLQJIRUHFDVWVDQGFRQGXFWLQJVLPXODWLRQV,WUHYLHZVWKHPRGHO¶VNH\HTXDWLRQV
DQG WKHUHVXOWV RI VLPXODWLQJ WKH HFRQRP\¶V UHVSRQVH WR WKUHH VWDQGDUG LPSXOVHV WKH
LQWHUHVWUDWHLPSXOVHWKHH[FKDQJHUDWHLPSXOVHDQGWKHILVFDOLPSXOVH
MATERIA¸Y I STUDIA – PAPER 36
7
General Description of the ECMOD Model
1 GENERAL DESCRIPTION OF THE ECMOD MODEL
1
GENERAL DESCRIPTION OF THE ECMOD MODEL
1
The class, form and degree of detail used in the models developed for the
purposes of monetary policy are conditional upon their application. ECMOD, the multiequation macroeconomic model of the Polish economy has been developed to generate
forecasts and conduct simulation analyses. Forecasting is conditioned on the starting
point assessment of future developments in the economy. Given the fact that the
transmission mechanism is spread over time, the course of future economic growth is
both determined by and determines current decisions related to macroeconomic policy.
Conducting a simulation enables economists to quantify the effects of various economic
policies and exogenous shocks on macroeconomic variables.
The model has been shaped by three criteria. First, it should provide a
comprehensive image of the mechanisms functioning in the economy; i.e. it should not
focus on a single variable, a section of the economy or partial equilibrium but rather on
a set of variables that provide a comprehensive description of economic reality. Second,
the model should enable the preparation of an internally and externally cohesive
description of economic processes. The model’s internal cohesion is construed as a
mutual complementation of its interdependencies, whereas its external cohesion is
supposed to guarantee that the model-based view of the economy reflects the actual
course of macroprocesses in the Polish economy. Third, given the simplified, by
definition, representation of reality, the model should enable economists to take into
account information from outside of the model (experts´ adjustments) in a manner that
guarantees cohesion of future paths for all variables.
ECMOD is a quarterly model, which structure comprises all of the main
macrocategories of the economy —inflation, GDP and its components, wages,
employment, the exchange rate, fiscal categories, etc. A proactive monetary policy, i.e.
a policy reacting to developments in the economy, is an integral element of the model1.
The ECMOD model is of medium size — it comprises approximately 100
equations, including approximately 15 behavioural equations2. Therefore, the model is
1
Taylor’s Rule is a nominal mechanism that stabilises the model’s performance in the long-term. In the
short and medium terms, the interest rate path may be determined according to Taylor’s Rule (it refers,
inter alia, to simulations that analyse less probable or even unrealistic scenarios) or on the basis of market
expectations (it refers, inter alia, to forecasts depicting the most probable course of events). (Cf.: 3.1).
2
According to standard categories in literature, the ECMOD model could be classified as a big
macroeconometric model; however, compared with the models used by other institutions (for instance,
FRB US – 300 equations), it may be assumed to be a medium-sized model.
8
N a t i o n a l
B a n k
o f
P o l a n d
General Description of the ECMOD Model
detailed enough to generate forecasts and analyse essential macroeconomic categories
while remaining transparent, which ensures the simplicity of its operation and facilitates
reporting of results3.
The form and distinctive features of the model are conditional upon analytical
1
and research requirements in the pursuit of monetary policy under the direct inflation
4
targeting strategy. The quarterly macroeconomic projections are optional forecast for
the economy under different interest rate scenarios, assuming the realisation of the
expected exogenous variables. In principle, at least two projection scenarios are
developed — using interest rates expected by the market and using fixed interest rates.
The scenario of interest rates expected by the market is considered equivalent to the
most probable economic growth scenario, whereas the scenario of fixed interest rates is
of an informative nature only - “how would the economy develop, should interest rates
remain unchanged.” The simulation analyses, on the other hand, are conducted to
determine the significance of individual transmission channels in the economy, and the
response of the economy to selected impulses (a detailed analysis covers the exchange
rate channel, the interest rate channel or the response of the economy to various fiscal
policies). The model is also a tool used to carry out more complex analyses (concerning
the adequacy of monetary and fiscal policies, scenarios of threats of exceeding safe
debt-to-GDP thresholds, etc.).
The concept behind the construction of the ECMOD model is based on the
attempt to reflect the macroeconomic processes observed in Poland during the sample
period, so that their mutual relationships are properly explained and their future
developments predicted. Adjustment to data is an empirical test of an adopted economic
theory in the case of Poland, and may, in consequence, determine the quality of
forecasting properties of the model. Embedding the model in the theory of the economy,
on the other hand, is aimed at explaining the mechanisms governing the economy,
according to which forecast and simulation results are interpreted. Consequently, in line
with Pagan’s classification (2001)5, the ECMOD model belongs to the class of hybrid
models that combine theory with information in the sample. In principle, Pagan
differentiates between a few classes of models of different degrees of theoretical and
empirical cohesion — in extreme cases, the form of the model is wholly derived from
3
The importance of transparency of the models used for the purposes of monetary policy is also
mentioned by C. Walsh in Transparency in Monetary Policy, Federal Reserve Bank of San Francisco
Economic Letter 2001-66 (2001).
4
The forecasts are made in iteration with experts. Section 1.1 presents more details concerning the
forecasting methodology.
5
Cf.: A. Pagan, Report on modelling and forecasting at the Bank of England, BoE, mimeo, 2003. Cf.
also: the discussion in R. KokoszczyĔski, B. Kłos, Modelowanie i prognozowanie w nowej kadencji RPP
— problemy i propozycje ich rozstrzygniĊcia, the NBP, mimeo, 2003. Pagan distinguishes between two
types of hybrid models — type I and II hybrid models.
MATERIA¸Y I STUDIA PAPER 36
9
General Description of the ECMOD Model
the existing economic theory or the model is of a statistical nature. Hybrid models fall
in between the purely theoretical and the purely empirical models. Most of the models
used by central banks to pursue monetary policy are hybrid models6, where the steady
state is based on the theory of economy, whereas short-term adjustments depend on
1
specific features of particular economies.
The construction of the ECMOD model resembles that of the models used by the
banks comprising the European System of Central Banks (the ESCB)7. Tab. 1 describes
the basic characteristics of the models used for forecasting inflation by ESCB banks.
Tab. 1 Comparative analysis — basic characteristics of the ECMOD model as
compared to ESCB models
ESCB models (prevailing
approach)
ECMOD
Model structure
Supply/demand model
• Supply side
- production function
- number of goods
- production factors
• Demand side
• Inflation
Monetary policy —
transmission channels
• Interest rate
- cost of credit
- cost of capital
• Exchange rate
• Asset
• Cash-flow
Fiscal policy – categories
• Number of expenditure
categories
• Number of revenue
categories
Short term – demand factors
Long term – supply factors
Short term – demand factors
Long term – supply factors
Cobb-Douglas
Cobb-Douglas, CES
1
1
Capital, labour
Capital, labour
disclosed according to the national
accounts; i.e. the following are modelled: disclosed according to the
consumption, capital formation, foreign national accounts
trade balance
The
modelling
usually
A few price indices are modelled: CPI,
involves GDP deflators and
GDP deflator, export and import
deflators of GDP components.
deflators. Inflation is modelled on the
The cost-push theory of
basis of the cost-push theory.
inflation prevails.
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
9
The factors that are taken into account
include transfers to the population,
current expenditure, investment expenditure (Cf.: Section 2.4) Data from
Reports on the execution of the Budget
Act
14
The following can be singled out: direct
taxes, indirect taxes (Cf: Section 2.4)
Data from Reports on the execution of the
Budget Act
3 - 22
The degree of detail depends
on available data.
4 – 26
The degree of detail depends
on available data.
6
The type I and II hybrid models are used e.g. by: the EBC, the FRB US, the Bank of England,
Bundesbank.
7
For more details on the models used by the European System of Central Banks for inflation forecasting,
cf.: B. Kłos, R. KokoszczyĔski, T. Łyziak, J. Przystupa, E. Wróbel, Modele strukturalne w
prognozowaniu inflacji w Narodowym Banku Polskim, Materials and Studies 180, the NBP, 2004.
10
N a t i o n a l
B a n k
o f
P o l a n d
General Description of the ECMOD Model
Macroeconomic policy
rules
Monetary policy
Interest rate
Endogenous interest rate — Taylor’s
Rule is included to conduct a simulation
analysis.
Exogenous interest rate — in the initial
option, interest rate forecasts are
compliant with market expectations.
Fiscal policy
The rule is based on the deficit.
The rule is applied via effective PIT rate
and social contributions.
In most cases — exogenous
interest rate (in the euro area,
monetary policy is exogenous
in relation to national banks).
Taylor’s Rule conjugated with
euro area aggregates is
included to conduct an
analysis of the models’
properties.
The rule is based on the
deficit.
The rule is applied via direct
taxes.
1
Given the changes observed in the economy and the development of the theory
of the economy and the methods of macromodelling, work on the model should
be of continuous character; i.e. the changes observed in the economy and the development of the methods of macromodelling should be accompanied by constant updates
of the model, which means that no version of the model should be considered final.
This paper is composed of four sections. Section I explains the construction
methodology, the forecasting methods and simulations and the theoretical basis
of the ECMOD model. Section II presents a review of the model’s equations —
individual equations have been presented on a modular basis, the modules comprising
the real economy, costs and prices, the external sector and the general government.
Section III covers simulation analyses — it contains the results of simulation of the
economy’s response to three standard impulses: the interest rate impulse, the exchange
rate impulse, and the fiscal impulse. Section IV is a summary of the findings.
1.1 ECMOD CONSTRUCTION METHODOLOGY, FORECASTING AND
SIMULATION METHODOLOGY
The ECMOD model uses quarterly data seasonally adjusted with the
TRAMO/SEATS method. The sample covers the period from Q1 1995 to Q4 20048.
The model is based on a co-integration analysis approach, commonly used in
macroeconometric modelling. This approach enables differentiation between short- and
long-term performance of the analysed macrocategories. Co-integrating relations
correspond to long-term relations, whereas short-term dynamic is explained by current
and lagged values of first differences of variables included in the co-integrating relation,
8
The model has been re-estimated/re-specified on two occasions since work on the model started (i.e.
since February 2003).
MATERIA¸Y I STUDIA PAPER 36
11
General Description of the ECMOD Model
and other exogenous variables. A brief description of the co-integration analysis is
included in Annex 4.
Individual equations are estimated separately, which results from the great
degree of detail in the model and the limited length of the time series in Poland. Long-
1
term relations have been estimated using the least-squares method, the modified leastsquares method and the Johansen method.
In most cases, equation parameters are estimated, but they can also be calibrated.
Three criteria have been used for the estimation and calibration of the equations: their
simulation properties, forecasting properties, and the quality of its fitting to historical
data.
Medium-term macroeconomic forecasts (covering eight to eleven quarters,
depending on the quarter in which the forecast is made) are generated using the model.
Experts’ judgment play a significant role when generating forecasts based on the model.
Experts set the paths of exogenous variables and their uncertainty, and produce
judgments concerning the selected endogenous variables (mainly the basic categories
of national accounts) in the forecasting period. The model is solved and, subsequently,
the obtained solution is compared with the experts’ judgments, using the model-experts
iteration method. It means that the forecast derived from the model undergoes
evaluation by experts. The experts’ adjustments are then included in the model, and the
procedure is repeated until convergence is reached. A more detailed description of the
model-experts iteration method is presentedin Annex 2. The CPI inflation forecast is
generated by compiling core inflation, obtained on the basis of the model and the
experts’ forecasts of food and fuel prices. In order to eliminate the effects of food price
shocks, the weight of this component is reduced to zero in the projection horizon (for
more details, see Section 2.2.1 on inflation).
For the purposes of monetary policy, the majority of simulations are run during
the forecasting period and are out-of-sample rather than in-sample simulations. The
starting point for a simulation is a forecast; i.e. the most probable future development of
selected macroeconomic categories, derived in the manner specified above9. Next, the
path of a variable (or variables) is determined, conditioned on which the simulation is
performed. The reaction of macroeconomic variables under analysis is determined by
solving the model, and experts’ interference with the endogenous variables is not
repeated. In particular, after the forecast has been made using the most probable path of
short-term interest rates, the experts’ judgment no longer have any impact on the
9
The experts’ judgments refer to the most probable scenario of future events rather than to a scenario
conditioned on fixed interest rates (projection).
12
N a t i o n a l
B a n k
o f
P o l a n d
General Description of the ECMOD Model
projection; i.e. on the model solving under the assumption of fixed interest rates. Thus,
the difference between a forecast and a projection is a reflection of the model’s
multipliers.
Potential interdependencies between the model’s exogenous variables give rise
1
to considerable restrictions on the simulations. In particular, given a strong disturbance
to some exogenous variables, it should not be assumed that the developments in other
exogenous variables remain unchanged (e.g. a strong adjustment of the expected
economic growth rate in the case of Poland’s major trading partners should be
accompanied by a change in external prices). The ability to make such simulations
depends on experts’ ability to jointly and coherently affect a few exogenous variables.
1.2 ECMOD — THEORETICAL ASPECTS
ECMOD is a supply-demand model. In the short-term, the economic growth
path is determined mainly by demand factors, taking into account inertia and rigidity.
The longer the analytical horizon, the greater the role of the factors associated with the
supply-side of the economy.
The model describes the economy in a way that resembles the manner of
disclosing macroprocesses in the national accounts. The ECMOD model consists of six
types of entities: households, enterprises, financial intermediaries, monetary authorities,
fiscal authorities and foreign countries.
Households receive income in the form of remuneration for work performed
for enterprises, social benefits from the government and transfers from foreign
countries. On the income they receive, households pay personal income tax as well as
health and pension insurance contributions, and using the remaining part of their income
(disposable income), they purchase goods manufactured at home or imported from
abroad. In their decisions concerning the level of consumption, households take into
account intertemporal substitution.
Enterprises produce goods that are sold on the domestic market and to foreign
countries. In the production process, they use two production factors — capital and
labour. When setting production levels (and investment expenditure), enterprises follow
the rule of profit maximisation, according to which the marginal product of capital is
compared with the real user cost of capital. On their income earned for the sale of
goods, enterprises pay corporate income taxes to the general government.
Fiscal authorities obtain revenues from taxes, which they subsequently
redistribute for retirement and disability pensions, social benefits for the unemployed,
MATERIA¸Y I STUDIA PAPER 36
13
General Description of the ECMOD Model
subsidies for enterprises, etc. The revenues are also spent by the government on
purchases of consumer goods and for financing investments. To cover possible deficits,
the state assumes debt on which interest is paid.
Monetary authorities control the level of nominal interest rates so that the
1
inflation target is met.
Financial intermediaries conduct currency sale and purchase transactions,
based on their evaluation of the attractiveness of the Polish economy when compared
with Poland’s major trading partners and expectations concerning its future growth
(including, in particular, the possibility of receiving EU funds). Financial intermediaries
also constitute an element of the transmission channel of the central bank’s policy
concerning interest rates on loans.
Foreign countries provide goods purchased by domestic entities and are, at the
same time, potential purchasers of goods manufactured at home.
In the model the conditions that optimise the decisions made by entities, given
the predefined restrictions, are not directly derived. The relationships between the above
mentioned entities shape the macroeconomic processes observed in the economy. Both
the supply and demand sides are modelled explicitly.
In the long-term, the potential product is determined by the production function;
i.e. it is obtained as a result of the combination of labour and capital outlays at an
exogenously predefined level of technology (total factor productivity). The form of the
production function determines the form of the demand for individual production
factors. The demand for labour is modelled via a reversal of the production function,
which is additionally influenced by demographic factors and labour costs. The desirable
capital stock is determined by the profit maximisation condition, in line with the
adopted production function. The actual capital stock gradually moves to the desirable
level through fixed capital formation, and both the adjustment and the process of
transforming capital formation into production assets occur over time. Consequently, it
becomes possible to take into account the cost of capital adjustment, on the one hand,
and delays between the outlay and building the effective capital stock (time-to-build),
on the other.
Due to the nominal rigidity, i.e. the delay in adjustment processes, the GDP level
in the short-term is determined by aggregate demand: consumption (individual
and collective), gross capital formation (fixed capital formation and changes in
inventories) and the foreign trade balance (exports adjusted for imports).
14
N a t i o n a l
B a n k
o f
P o l a n d
General Description of the ECMOD Model
Individual consumption is determined by the disposable income of households
and the amount of wealth. Disposable income is mainly composed of wages, income
earned by self-employed individuals, income from property and net flows from the
general government and foreign countries. Wealth comprises three main stock variables
1
of the model: capital, public debt and net foreign assets. In the short-term, households
follow intertemporal substitution, making the current consumption level conditional
upon the interest rate level.
Fixed capital formation is determined by the required capital level, set by the
current demand and the real user cost of capital. Moreover, the financial accelerator
effect, i.e. the interdependence between the ability of enterprises to form capital and
their financial standing, and cost factors (oil prices), have been taken into account.
Changes in inventories, on the other hand, depend on the level of economic activity.
Exports, in the long-term, depend on the potential product level and the ratio of
export prices to domestic production prices. In the short-term, external demand is an
important determinant of exports. The import function corresponds to the export
function: in the long-term, the level of imports depends on the potential product level
and the ratio of import prices to domestic production prices; in the short-term, imports
depend on the current level of economic activity.
Potential labour input is determined by demographic and structural factors,
i.e. the number
of
professionally
active
persons
adjusted
by
the
NAWRU
unemployment rate (Non-Accelerating Wage Rate of Unemployment), which is
considered equivalent to structural unemployment. In the short-term, a gap may occur
between the actual rate of unemployment and the NAWRU rate and a corresponding
output gap, defined as the difference between the actual product and the potential
product (consistent with the production function and potential labour input) may also
exist. Both gaps have an impact on the level of prices and wages.
In the long-term, nominal wages are adjusted for labour productivity and price
changes, taking into account restrictions in the form of production material costs. Apart
from the real and nominal rigidities, one of the factors which makes the level of wages
differ from the labour productivity level is the unemployment gap. In a situation where
the actual rate of unemployment is higher than the NAWRU indicator, employees’
power to negotiate wage increases is relatively weak and the pressure to increase wages
is consequently lower. A corresponding situation can be observed where the actual rate
of unemployment is lower than the NAWRU indicator — in such a case, the pressure to
increase wages is greater than expected as the result of changes in labour productivity.
MATERIA¸Y I STUDIA PAPER 36
15
General Description of the ECMOD Model
7KH OHYHO RI SULFHV LQ WKH ORQJWHUP LV GHWHUPLQHG E\ SURGXFWLRQ FRVWV
UHSUHVHQWHG E\XQLW ODERXU FRVWV DQG LPSRUW SULFHV ,Q WKH VKRUWWHUP LQIODWLRQ LV
LQIOXHQFHGE\WKHFXUUHQWOHYHORIHFRQRPLFDFWLYLW\PHDVXUHGE\WKHRXWSXWJDSDQGWKH
VXSSO\VKRFNVUHSUHVHQWHGE\H[RJHQRXVFKDQJHVLQIRRGDQGIXHOSULFHV
1
7KH ORQJWHUP SDWK RI WKH H[FKDQJH UDWH LV FRQVLVWHQW ZLWK WKH WKHRU\ RI
SXUFKDVLQJ SRZHU SDULW\ 7KH H[FKDQJH UDWH¶V GHYLDWLRQ IURP WKH ORQJWHUP SDWK
GHSHQGV RQ WKH UHDO LQWHUHVW UDWH GLVSDULW\ DW KRPH DQG DEURDG DQG WKH OHYHO RI QHW
IRUHLJQDVVHWVDVZHOODVWKHULVNSUHPLXP
7KH JHQHUDO JRYHUQPHQW KDV DQ LPSDFW YLD LWV VSHQGLQJ SROLF\ FROOHFWLYH
FRQVXPSWLRQ WUDQVIHUV DQG VXEVLGLHV DQG LQFRPH SROLF\ WD[ UDWHV RQ ERWK WKH UHDO
DQGQRPLQDOVHFWRUVRIWKHHFRQRP\
)LJSUHVHQWVDVLPSOLILHGGLDJUDPRIWKHLQWHUGHSHQGHQFLHVLQWKH(&02'PRGHO
)LJ6LPSOLILHGGLDJUDPRILQWHUGHSHQGHQFLHVEHWZHHQYDULDEOHVLQWKH(&02'
PRGHO
([WHUQDOHQYLURQPHQW
([WHUQDOGHPDQG
)RUHLJQSULFHV
3/1(85UDWH
*'3
,QGLYLGXDOFRQVXPSWLRQ
&ROOHFWLYHFRQVXPSWLRQ
&DSLWDOIRUPDWLRQ
([SRUW
,PSRUW
3RWHQWLDOSURGXFW
2XWSXWJDS
,QWHUHVWUDWHV
7D\ORU¶V5XOH
VLPXODWLRQ
3ULFHV
&3,
([SRUWGHIODWRU
,PSRUWGHIODWRU
*'3GHIODWRU
/DERXUPDUNHW
:DJHV
(PSOR\HGSHUVRQV
8QHPSOR\PHQWJDS
*HQHUDOJRYHUQPHQW
5HYHQXHV
([SHQGLWXUH
'HILFLW
'HEW
)LVFDOUXOH VLPXODWLRQ
6RXUFH 1%3
7R HQVXUH WKH GLDJUDP¶V WUDQVSDUHQF\ DUURZV DUH XVHG WR PDUN RQO\ WKH PRVW LPSRUWDQW OLQNDJHV
7KHHTXDWLRQV VHW IRUWK LQ VHFWLRQ ,, RI WKLV SDSHU FRQWDLQ D GHWDLOHG GHVFULSWLRQ RI LQWHUGHSHQGHQFLHV
ZLWKLQWKHPRGHO
16
N a t i o n a l
B a n k
o f
P o l a n d
Modules of the Economic Model
2 MODULES OF THE ECMOD MODEL
2
Modules of the Economic Model
The table below defines the main equations and identities, broken down into
2
blocks, found in the model. Small letters were used to mark the logarithms of the
variables. A description of the variables is included in Annex 5.
I. Supply side
(1)
Production
function:
GDPt = EMPt 0,68 Kt0,32TFP
(2)
Capital stock:
K t = (1 − DEPR) K t −1 + 0,38FINV t +0,22FINVt −1 + 0,12FINVt −2
+ 0,09FINVt −3 + 0,08FINVt − 4 + 0,06FINVt −5 + 0,05FINVt −6
(3)
Employed
persons:
∆ emp t = 0 ,16 ( gdp t −1 − 0 , 68 emp
2 ,8
+ 0 , 70 ∆ lf t + 0 ,30 ∆ emp
6,4
− 0 ,1( ∆ ln( WAGE
(4)
Potential
product:
(5)
Output gap:
t −1
t −1
t −1
− 0 , 32 k t −1 − tfp t −1 ) / 0 , 68
+ 0 , 23 ( ∆ gdp t − ∆ tfp t / 0 , 68 )
1, 8
/ PGDP
t −1
) − ∆ tfp t −1 / 0 , 68 )
GDP _ POTt = ( LFt (1 − NAWRU t ))0,68 K t0,32TFP
GAPt =
GDPt
−1
GDP _ POTt
II. Demand side
(6)
GDP
GDP = CONP + CONGOV + INV + (GDP_EXP-GDP_IMP)
(7)
Individual
consumption:
∆conp t = 0.01− 0.12(conp t −1 − 0,94 yd t −1 − (1 − 0,94) wealth t − 2 − 0.25)
3 .1
2 .2
+ 0,27 ( ∆conp t −1 + ∆conp t − 2 + ∆conp t − 3 + ∆conp t − 4 ) / 4 + 0,09 ∆yd t
1 .6
1 .4
− 0,0006 i _ w3mr _ cpi t + DUMMIES
0 , 78
(8)
Disposable
income:
YD _ N t = WAGEFUND _ N t + YD _ NO t + YD _ WL t + GTR _ N t
+ TR _ WPR _ CON t + YD _ POZOSTALE t − GPIT _ N t
− GLT _ EMP _ N t − GLT _ HC _ N t − GLT _ SB _ N t
(9)
Fixed capital
formation:
∆finvt = 0,27(0,32
1, 9
GDPt −1
− RUCCt −1 ) + 0,46 ∆finvt −1 + 0,27 ∆finvt − 2
5, 5
3, 6
K t −1
+ 0,17 ∆FINACCt
1,8
III. Prices and costs
(10) Core inflation:
∆netcpit = 1,33∆dlt − 0,15(netcpit −1 − (0,61 + 0,83ulct −1 + 0,17 p _ impt*−1
( 2, 78)
( −1, 7 )
+ 0,7dlt −1 )) + 0,33 ∆netcpit −1 + 0,22 ∆ulct + 0,1 GAPt
( 2,16)
( 2,34 )
(1, 25)
+ 0,02 ∆p _ imp
( 0, 65)
MATERIA¸Y I STUDIA PAPER 36
*
t
17
Modules of the Economic Model
(11)
CPI inflation:
(12)
GDP deflator:
∆cpi = waga _ net * ∆netcpi+ waga _ food * ∆foodcpi+ waga _ fuel * ∆fuelcpi
where: waga_net + waga_fuel + waga_food = 1
∆pgdpt = − 0,18( pgdpt −1 − 0,8 ulct −1 − 0,20 p _ impt −1 + 4,0) + 0,32 ∆pgdpt −1
3, 5
4 ,1
+ 0,16 ∆pgdpt −2
2, 2
(13)
Wages:
∆waget = 0,43∆waget −1 + 0,57 ∆cpit −1 + 0,57(∆gdpt − ∆empt )
4, 2
− 0,23(UNEMPt −1 − NAWRUt −1 ))
2
3, 4
− 0,18(waget −1 − gdpt −1 + empt −1 − 1,25 pgdpt −1 + 0,25 p _ impt −1 − 5,02)
3,9
(14)
Real effective
zloty exchange
rate:
∆s _ reert = −0,75∆ ( gdp _ pot − gdp _ ext _ pot ) t − 0,28 ∆ (i _ w3mr _ pgdp
( 0 , 39 )
− i 3mr _ ext − s _ risk _ pr ) t −1 − 0,26 ( s _ reer − 10,55
( 0 , 09 )
NFA
)+
4 * GDP _ N
+ 1,31(i _ w3mr _ pgdp − i 3mr _ ext − s _ risk _ pr )) t −1 + DUMMIES
+ 0,75( gdp _ pot − gdp _ ext _ pot ) + 0,25(
(15)
Interest rates on
loans to
enterprises:
I _ RENTP t = − 0, 27 − 0, 44 ( I _ RENTP t −1 − 2,16 − I _ W 3 M t −1 )
( − 2 , 76 )
( − 4 , 56 )
+ 0,64 ∆ I _ W 3 M t + I _ RENTP t −1
(11 , 02 )
IV. External sector
(16)
Balance of
payments:
CAB t = CAB _ NT t + CAB _ INC t + CAB _ TRANS t
(17)
Exports:
∆gdp _ expt = 0,02− 0,35( gdp _ exp+ 1,44 − 0,01trend − gdp _ pot
( 2 , 96 )
( 3, 92 )
(10 ,1)
− 0,74( p _ exp− pgdp))t −1 + 1,01 ∆gdp _ extt + DUMMIES
(1, 96 )
(18)
Imports:
( 0 ,81)
∆gdp _ impt = 0,01− 0,12 ( gdp _ imp + 1, 44 − 0,01 trend − gdp _ pot
( 0 , 91)
(1, 75 )
(10 ,1)
+ 0,70 ( p _ imp _ noil + gtar _ tr − pgdp ))t −1 + 1,59 ∆gdp t
(1, 3 )
( 2 , 26 )
+ DUMMIES
(19)
Export deflator: ∆ p _ exp t = ∆ pgdp t − 0 , 61 (( p _ exp − pgdp ) + 6 ,12 − 0 ,58 ( s _ reer +
( 4 , 92 )
( 6 , 31 )
+ 0 , 75 ( gdp _ pot − gdp _ pot _ ext ))) t −1 + 0 , 45 ∆ ( s _ reer ) t
( 4 , 72 )
+ DUMMIES
(20)
Import deflator ∆p _ imp _ noilt = ∆pgdpt − 0,21(( p _ imp _ noil − pgdp) + 6,88 − 0,61(s _ reer +
( 2, 54)
( 4,32)
adjusted by oil
price
+ 0,75( gdp _ pot − gdp _ ext _ pot)))t −1 + 0,63∆(s _ reer)t
( 4,32)
movements
according to the
+ DUMMIES
national
accounts:
(21)
Import deflator:
p _ imp t = 0,92 p _ imp _ noil t + 0,08(oil + s _ usd _ p ln) t
V. General government (GG)
(22)
18
GINC_N = GIDT_N + GPIT_N + GCIT_N + GLT_N + GTAR_N + GCT_N +
General
Government
CGNBP_N + GINC_UE_N + GRT_N
(GG) revenues:
N a t i o n a l
B a n k
o f
P o l a n d
Modules of the Economic Model
(23)
GG
expenditure:
GEXP_N = GCE_N + GTR_N + GINV_N + GTR_CORP_N + GDS_N
(24)
GG balance:
GDEF_N = GINC_N - GEXP_N
(25)
Public debt:
GDEBT_N = GDEBT_N(-1) * [(GDEBT_EUR * S_EUR_PLN +
+ GEXP_UE_N + GRE_N
+ (1-GDEBT_EUR) * S_USD_PLN) / (GDEBT_EUR *
* S_EUR_PLN(-1) + (1-GDEBT_EUR) * S_USD_PLN(-1)) *
* (1 - GDEBT_PL) + GDEBT_PL] - GDEF_N - GPRIV_N +
+ adjustment for the balance of pre-financing of EU projects +
+ adjustment for funds in government accounts +
2
+ adjustment for hospital debt
VI. Rules governing monetary and fiscal policies
(26)
(27)
Monetary
policy (Taylor’s
Rule):
i _ w3mt = (1 − α ) * [φ + inf t + βGAPt + γ (inf t − inf_ cel )]
Fiscal policy:
∆et = θ × (d t −1 − d ∗ )
2.1
+ α * i _ w3mt −1
REAL ECONOMY
2.1.1 Supply side
The two-factor Cobb-Douglas production function is the core element of the
supply side of the model. The parameters have been calibrated, on the basis of the share
of remuneration of both production factors in gross value added, less the balance of
taxes and subventions from producers11, according to the assumption of perfect
competition on the product market in the long-term.
The so-called Solow residuals, which, having been HP filtered, constitute
an approximation of the trend in the total factor productivity, have been calculated
on the basis of the estimated product elasticity against production factor inputs. As a
result, the production function is expressed in the following equation:
GDP t = EMP t 0 , 68 K t0 , 32 TFP
E. 1
where:
GDP – gross domestic product,
EMP – employed persons according to BAEL (equation E. 4),
11
The problem concerning calculation of remuneration of production factors on the basis of data from
the national accounts is the household-related category of mixed income, which combines remuneration
of both the labour and capital factors (cf.: A. CzyĪewski (2002)). Upon distribution of the mixed income
MATERIA¸Y I STUDIA PAPER 36
19
Modules of the Economic Model
K – capital stock at the end of the period (equation E. 3),
TFP – total factor productivity.
The factors of production — capital and labour inputs – are determined
in the following manner. The capital stock in the base year corresponds to the value of
fixed assets according to the Central Statistical Office’s national accounts. Gross capital
formation, on the other hand, is calculated using the perpetual inventories method, with
2
regard to the lags´distribution describing the relationship between the completed fixed
assets and the outlays incurred on them (cf.: M. Gradzewicz, M. Kolasa, 2004)12:
KEPt = (1 − depr ) KEPt −1 + 0,38FINVt + 0,22 FINVt −1 + 0,12 FINVt − 2 + 0,09 FINVt −3
+ 0,08FINVt − 4 + 0,06 FINVt −5 + 0,05FINVt −6
E. 2
where:
KEP – capital stock at the end of the period
FINV – investment input (equation E. 11)
K t = ( KEPt + KEPt −1 ) / 2
E. 3
where:
K- average stock of capital KEP in the period.
Labour input in the long-term are determined via a reversal of the production
function. The demand for labour in the short-term is also influenced by GDP growth
and by the real cost of labour incurred by the employer, with both categories adjusted
for the long-term growth rate in labour productivity13. Moreover, short-term growth has
been supplemented by the autoregressive part, taking account of the dynamic
homogeneity.
∆ emp
t
= 0 ,16 ( gdp
2 ,8
+ 0 , 30 ∆ emp
t −1
t −1
− 0 , 68 emp
+ 0 , 23 ( ∆ gdp
− 0 ,1( ∆ ln( WAGE
1,8
t −1
/ PGDP
t
t −1
− 0 , 32 k t − 1 − tfp t −1 ) / 0 , 68 + 0 , 70 ∆ lf t
6,4
− ∆ tfp t / 0 , 68 )
E. 4
t − 1 ) − ∆ tfp t − 1 / 0 , 68 )
where:
EMP –employed persons according to BAEL,
between both factors, it has been assumed that the share of the remuneration thereof in the household
sector is the same as in the whole economy.
12
The annual data on fixed assets obtained from investment activity and their structure by type, published
by the Central Statistical Office, has been used. The split into quarters has been based on the data on gross
fixed capital formation from the national accounts. In the case of discrepancy between the sum of
quarterly gross fixed capital formation from the national accounts and the annual data published by the
Central Statistical Office (which follows from taking into account in the quarterly data in the national
accounts the outlays on renovation, intangible assets and outlays incurred in the grey economy) the prorata adjustment thereof has been made.
13
Upon the growth of real wages, the parameter has been set at -0.1.
20
N a t i o n a l
B a n k
o f
P o l a n d
Modules of the Economic Model
GDP – gross domestic product (equation E. 1),
TFP– total factor productivity,
LF – labour force, in thousands, according to BAEL,
WAGE – gross wages (equation E. 16),
PGDP – GDP deflator (equation E. 15).
Adjusted R2: 61.0%
Autocorrelation of the random component Ȥ2 (1)=1.97 (p=0.16)
2
Autocorrelation of the random component: Ȥ2 (4)=4.05 (p=0.40)
Normality of distribution of the random component: JB=0.93 (p=0.63)
Sample: 1996 Q1 – 2004 Q4
The estimated production function (E. 1) is also used to determine the potential
product in which the potential labour input is determined by the labour force, adjusted
by the NAWRU indicator (non-accelerating wage rate of unemployment). Both the
number of professionally active persons and the NAWRU14 indicator are set
exogenously.
GDP _ POTt = ( LFt (1 − NAWRU t ))0, 68 K t0,32TFP
E. 5
The output gap is defined as the relative difference between actual and potential
GDP:
GAPt =
GDPt
−1
GDP _ POTt
E. 6
2.1.2 Demand side
The demand-side of the economy (the market for goods and services) is
modelled explicitly. This means that equations are estimated for individual GDP
components. GDP is calculated as the sum of all its components.
2.1.2.1 Individual consumption (by households)
Households provide entrepreneurs with labour and receive remuneration in
return. They also receive transfers from both the government (social benefits) and
14
In the sample period, the NAWRU indicator is estimated using the modified Elmeskov method (1993),
according to which the change in wage growth rate is proportional to the difference between the actual
2
rate of unemployment and the NAWRU indicator; i.e.: ∆ log W = −α (U − NAWRU ) , α >0, where
W — wage level, U — rate of unemployment or its lags´distribution . Under the assumption of a constant
NAWRU indicator for two subsequent quarters, the α parameter is estimated for the following periods,
and on its basis a NAWRU series is calculated; a smoothed series is used in the model.
MATERIA¸Y I STUDIA PAPER 36
21
Modules of the Economic Model
foreign countries (at the moment, these are mainly benefits received as part of the
Common Agricultural Policy). In order to smoothen consumption, households, as
holders of all assets; i.e. government securities (the amount of which is reflected in the
amount of public debt), net foreign assets and the value of capital, control the amount of
their savings/debt.
Consequently, the long-term level of consumption is determined by the amount
of disposable income (cf: E. 8) and the value of wealth. The value of wealth is construed
2
as a sum of public debt, net foreign assets and the amount of capital. In the short-term,
consumption growth is also influenced by the interest rate (according to the hypothesis
of intertemporal consumption substitution, higher interest rates encourage postponement
of current consumption). Consequently, the individual consumption function is as
follows:
∆conpt = 0.01− 0.12(conpt −1 − 0,94 yd t −1 − (1 − 0,94) wealtht − 2 − 0.25) + 0,27( ∆conpt −1
3.1
2.2
1.6
+ ∆conpt − 2 + ∆conpt −3 + ∆conpt − 4 ) / 4 + 0,09 ∆yd t − 0,0006 i _ w3mr _ cpit − 0.009 * i 00q3 E. 7
1.4
0 , 78
− 0.008 * i 02q 2
where:
CONP – individual consumption,
YD – disposable income (equation E. 8),
WEALTH – wealth,
I_W3MR_CPI – real interest rate (WIBOR 3M deflated by CPI inflation).
Adjusted R2: 42%
Autocorrelation of the random component: Ȥ2(4)=2.21 (p=0.09)
Normality of distribution of the random component: JB=1.65 (p=0.44)
Sample: 1996Q2 – 2004Q4
2.1.2.2 Disposable income
Gross disposable income is calculated as the sum of income earned from
employment (wages), the operating surplus of the household sector, income from
property, social benefits from the public sector, other disposable income and, since
2004, transfers from the European Union granted as part of the Common Agricultural
Policy. Personal income taxes and health and pension insurance contributions are paid
on the above-mentioned income:
22
N a t i o n a l
B a n k
o f
P o l a n d
Modules of the Economic Model
<' B 1 W = :$*()81' B 1 W + <' B 12W + <' B :/W + *75 B 1 W
+ 75 B :35 B &21 W + <' B 32=267$/( W − *3,7 B 1 W − */7 B (03 B 1 W
(
− */7 B +& B 1 W − */7 B 6% B 1 W
ZKHUH
<'B1±JURVVQRPLQDOGLVSRVDEOHLQFRPH
<'B12±QRPLQDORSHUDWLQJVXUSOXVRIWKHKRXVHKROGVHFWRU HTXDWLRQ(
<'B:/±QRPLQDOSURSHUW\LQFRPHLQKRXVHKROGVVHFWRU HTXDWLRQ(
2
:$*()81'B1±TXDUWHUO\ZDJHIXQG
*75B1±WUDQVIHUVWRWKHSRSXODWLRQ VRFLDOEHQHILWVXQHPSOR\PHQWEHQHILWVSHQVLRQV
HTXDWLRQ(
75B:35B&21 ± UHYHQXH REWDLQHG DV SDUW RI WKH &RPPRQ $JULFXOWXUDO 3ROLF\
DOORFDWHGWRFRQVXPSWLRQ
<'B32=267$/(±RWKHUGLVSRVDEOHLQFRPH
*3,7B1±SHUVRQDOLQFRPHWD[ HTXDWLRQ(
*/7B(03B1±VRFLDOFRQWULEXWLRQVSDLGE\HPSOR\HHV HTXDWLRQ(
*/7B+&B1±KHDOWKLQVXUDQFHFRQWULEXWLRQV HTXDWLRQ(
*/7B6%B1±VRFLDOFRQWULEXWLRQVSDLGE\EXVLQHVVRZQHUVDQGIUHHODQFHSURIHVVLRQDOV
HTXDWLRQ(
7KH JURVV RSHUDWLQJ VXUSOXV RI KRXVHKROGV FRQVWLWXWHV PRUH WKDQ RI JURVV
GLVSRVDEOHLQFRPH7KLVLVWKHJURVVYDOXHDGGHGFUHDWHGE\GRPHVWLFXQLWVFRPSULVLQJ
WKHKRXVHKROG VHFWRU PDLQO\ HPSOR\HUV DQG VHOIHPSOR\HG LQGLYLGXDOV H[FOXGLQJ
WKHDJULFXOWXUDO VHFWRU OHVV WKH FRVWV RI HPSOR\PHQW DQG SURGXFWLRQUHODWHG WD[HV
DQGLQFUHDVHGE\VXEVLGLHVWRSURGXFHUV
7KHJURVVRSHUDWLQJVXUSOXVRIKRXVHKROGVLVFRQGLWLRQDOXSRQDJJUHJDWHGHPDQG
LQWKHHFRQRP\ LQWKHORQJWHUPWKHHODVWLFLW\ZLWKXQLWDU\UHVWULFWLRQ
∆\G B QRW = − \G B QRW − − JGS B QW − + + ∆\G B QRW − + ∆JGS B QW
−
(
ZKHUH
<'B12±RSHUDWLQJVXUSOXVRIWKHKRXVHKROGVHFWRU
*'3B1±QRPLQDOJURVVGRPHVWLFSURGXFW
$GMXVWHG5
$XWRFRUUHODWLRQRIWKHUDQGRPFRPSRQHQW Ȥ S
1RUPDOLW\RIGLVWULEXWLRQRIWKHUDQGRPFRPSRQHQW Ȥ S
6DPSOH 4±4
7KHKRXVHKROGVVHFWRULQFOXGHVVHOIHPSOR\HGLQGLYLGXDOVHPSOR\LQJXSWR ZRUNHUV
MATERIA¸Y I STUDIA PAPER 36
23
Modules of the Economic Model
Real property income from households´sector (CPI-deflated) is a function
of the product of the real interest rate and wealth. Part of the property income is directly
dependent on the interest rate level. It refers to investment in certain financial
instruments, including bank deposits, loans and bonds:
YD _ WL = − 46,1 + 5,88 * 10 −7 I _ W 3M * WEALTH
( −1, 51)
E. 10
( 2 , 69 )
where:
2
YD_WL – property income in the households´ sector,
I_W3M – average quarterly value of WIBOR3M (equation E. 66),
WEALTH – wealth.
Adjusted R2: 28.5%
Autocorrelation of the random component: Ȥ2(2)=8.84 (p=0.01)
Normality of distribution of the random component: Ȥ2 (2)=0.8 (p=0.66)
Sample: 1995Q1 – 2004Q4
2.1.2.3 Fixed capital formation
In the case of the investment equation (fixed capital formation), a decision has
been made to substitute estimation of the co-integrating relation with a theoretical
relation describing fixed assets (capital) in a steady-state of the neoclassical model16.
The long-term relationship is linked to the rule of profit maximisation by
business entities, according to which the marginal capital product is equal to a real user
cost of capital (RUCC). Assuming the production function is expressed as a CobbDouglas production function (equation E. 1), the required (optimum) capital level is
expressed in the following equation:
Kt = α
Yt
RUCCt
where:
Į- product elasticity against capital,
K – capital,
GDP – gross domestic product,
RUCC – real user cost of capital.
16
It is extremely difficult to estimate the investment equation. In the investment equation, long-term
relationships have been accepted on the basis of the theory of economy, consistent with the disclosure in
the supply-side of the model (the condition of maximisation of profits earned by enterprises, under the
Cobb-Douglas production function).
24
N a t i o n a l
B a n k
o f
P o l a n d
Modules of the Economic Model
Deviation of the actual capital stock from its optimum level causes adjustment in
fixed capital formation.
Short-term growth in fixed capital formation is explained using the error
correction model. Given the inertia of investment processes, the autoregressive part was
added to the equation. Moreover, the short-term equation takes into account the
financial accelerator , according to which in the condition of information asymmetry the
improvement of the financial standing of enterprises increase their capital formation
2
ability17.
∆finvt = 0,27(0,32
1,9
GDPt −1
− RUCCt −1 ) + 0,46 ∆finvt −1 + 0,27 ∆finvt − 2 + 0,17 ∆FINACCt
5, 5
3,6
1,8
Kt −1
E. 11
where:
FINV – investment input,
K – capital (equation E. 3),
GDP – gross domestic product (equation E. 1),
RUCC – real user cost of capital (interest rates on enterprises´loans deflated with the
GDP deflator, adjusted by capital depreciation and the corporate income tax),
FINACC – share of disposable income of enterprises18 in national disposable income.
Adjusted R2: 88.5%
Autocorrelation of the random component: F(1)=0.44 (p=0.51)
Autocorrelation of the random component: F(4)=0.49 (p=0.74)
Normality of distribution of the random component: JB=1.34 (p=0.51)
Sample: 1996 Q1 – 2004 Q4
2.2
PRICES AND COSTS
2.2.1 CPI index
In the ECMOD model, inflation is represented by the core inflation index.
The inflation equation is based on the concept of cost-push inflation, according to which
prices depend on the costs of production factors, i.e. unit labour costs and import prices.
The price level is also influenced by demand factors, quantified using the output gap.
Where demand is higher than supply, sellers are able to demand higher margins, which
results in price growth. Otherwise, the low demand barrier necessitates a reduction in
17
For the purpose of cost-push shock, the negative logarithm of oil prices and the domestic goods price
ratio is added to the equation – to reflect changes in relative price levels. The cost-push shock results in
price increases, but (contrary to the demand shock) the demand for a product decreases.
MATERIA¸Y I STUDIA PAPER 36
25
Modules of the Economic Model
margins. Inflation is additionally affected by inflation expectations. Research on the
mechanism of formation of inflation expectations in the Polish household sector has
confirmed its adaptation properties (see: Łyziak (2003)). For this reason, they are
represented in the model by the lagged inflation.
The following is the core inflation equation:
∆netcpit = 1.33∆dl t − 0.15(netcpit −1 − (0.61 + 0.83ulc t −1 + 0.17 p _ imp t*−1 + 0.7 dl t −1 ))
( 2.78 )
2
( −1.7 )
E. 12
+ 0.33 ∆netcpit −1 + 0.22 ∆ulc t + 0.1 GAPt + 0.02 ∆p _ imp t*
( 2.16 )
( 2.34 )
(1.25 )
( 0.65 )
where:
NETCPI – net core inflation price index – excluding food and fuel prices,
ULC – unit labour costs, ulc = emp + wage − gdp
P_IMP* – import prices adjusted by customs duties and import taxes (applicable in the
mid-90s),
GAP – output gap (equation E. 6),
dl – an artificial variable used to quantify the growth of net core inflation faster than the
cost determinant of inflation in the sample.
Adjusted R2: 85.0%
Autocorrelation of the random component: Ȥ2(4)=0.96 (p=0.62)
Normality of distribution of the random component: Ȥ2 (4)=3.62 (p=0.16)
Sample: 1995Q1 – 2004Q4
The homogeneity of the impact of unit labour costs and import prices on net
inflation has been achieved through the imposition of relevant restrictions. The dl
variable has been added to the equation as a result of a number of factors occurring in
the sample in favour of a higher level of net inflation rather than a higher rate of growth
in the cost factors found in E. 12. These factors include price deregulation, a higher
share of services and a smaller share of industrial goods in the NETCPI index structure
than in the structure of GDP.
In the model, core inflation is quantified using the net inflation index, which is
derived by removing food and fuel prices from the complete CPI index. In order to
compute the CPI index in simulations and forecasts, it is necessary to add food and fuel
prices to net inflation derived from the E. 12 equation. Thus, CPI inflation is calculated
according to the following equation:
18
Enterprises consist of two institutional sectors specified in the national accounts: the business sector
and the sector of financial institutions.
26
N a t i o n a l
B a n k
o f
P o l a n d
Modules of the Economic Model
∆cpi = waga _ net * ∆netcpi + waga _ food * ∆foodcpi + waga _ fuel * ∆fuelcpi
E. 13
where:
waga_net+waga_fuel+waga_food=1
E. 14
In the forecasting process, inflation series for food and fuel prices are forecasted
outside of the model by experts. Experience shows that independent forecasting of food
price inflation is justifiable only on a time horizon of approximately one year; in the
case of long-horizon forecasting, forecasts of food price inflation are consistent with
2
forecasts concerning inflation in general. Thus, it has been assumed in the model that
over the forecast horizon, the weight of food prices (waga_food) in CPI inflation
decreases, and after one year it equals zero. At the same time, waga_net is increased, so
that the condition contained in E. 14 is maintained. In the case of fuel prices, the most
important determinant is the change in crude oil prices on international markets. For this
reason, no analogous mechanism of reducing to zero has been applied in the the
waga_fuel series. This corresponds to the assumption concerning the external
environment and the fact that crude oil prices have been taken into account in other
segments of the model.
2.2.2 GDP deflator
As in the case of the core inflation equation (cf. E. 12) the GDP deflator, in the longterm, is determined by unit labour costs and import prices, with the requirement of
homogeneity being met. The co-integrating relationship corresponds to the cointegrating relationship in the wage equation (cf. E. 16).
∆pgdp t = − 0,18( pgdpt −1 − 0,8 ulct −1 − 0,20 p _ impt −1 + 4,0) + 0,32 ∆pgdpt −1
3, 5
+ 0,16 ∆pgdp t − 2
4 ,1
E. 15
2, 2
where:
PGDP – GDP deflator,
ULC – unit labour costs, ulc = emp + wage − gdp ,
P_IMP – import prices adjusted by customs duties and import taxes (applicable in the
mid-90s.) (equation E. 35).
Adjusted R2: 83.9%
Autocorrelation of the random component: Ȥ2 (1)=3.52 (p=0.07)
Autocorrelation of the random component: Ȥ2(4)=5.81 (p=0.21)
Normality of distribution of the random component: Ȥ2 (2)=5.43 (p=0.07)
Sample: 1995Q4 – 2004Q3
MATERIA¸Y I STUDIA PAPER 36
27
Modules of the Economic Model
2.2.3 Wages
The concept of the function of real wages has been based on the productivity
mechanism of determining average wages in the economy consistent with the adopted
Cobb-Douglas production function. In the short-term, growth in nominal wages is
determined by the autoregressive factor on the one hand, and movements in consumer
goods prices and short-term changes in labour productivity (the requirement of dynamic
homogeneity has been imposed in both cases) on the other. The impact of the cost of
2
living, measured using the CPI index, on nominal wages is spread over time and reflects
the existence of nominal rigidities. Moreover, it has been assumed that, in the longterm, changes in prices of goods fully translate into changes in wages, which is one of
the conditions necessary to ensure equilibrium on the goods and labour markets.
Wages react to deviations from two types of equilibria, which may occur in the
short-term. The first disequilibrium is the difference between actual and natural
unemployment (measured using the NAWRU indicator). It reflects the situation on the
labour market and, in the case of a high unemployment rate, leads to a reduction in the
supply pressure to increase labour prices. The other disequilibrium, related to the
specification of the production function, is the difference between productivity and real
wages. Here, both the price index of goods manufactured in the domestic economy and
the import price index19 have been used as deflators of nominal wages in the long-term,
with the imposed requirement of homogeneity. Excessive growth in the productivity of
employed persons above the real marginal labour cost defined in this way results in
upward pressure on wages, and offsets the initial disequilibrium.
∆ wage t = 0 , 43 ∆ wage t −1 + 0 ,57 ∆ cpi t −1 + 0 ,57 ( ∆ gdp t − ∆ emp t )
4,2
− 0 , 23 (UNEMP t −1 − NAWRU
3, 4
t −1
))
E. 16
− 0 ,18 ( wage t −1 − gdp t −1 + emp t −1 − 1, 25 pgdp t −1 + 0 , 25 p _ imp t −1 − 5,02 )
3, 9
where:
WAGE – nominal wages,
CPI – consumer price index,
PGDP – GDP deflator (equation E. 15),
GDP – gross domestic product (equation E. 1),
EMP – number of employed persons according to BAEL (equation E. 4),
UNEMP – rate of unemployment,
NAWRU – Non-Accelerating Wage Rate of Unemployment,
28
N a t i o n a l
B a n k
o f
P o l a n d
Modules of the Economic Model
P_IMP – import prices adjusted by customs duties and import taxes (applicable in the
mid-90s) (equation E. 35).
Adjusted R2: 81.1%
Autocorrelation of the random component χ2 (1)=0.52 (p=0.47)
Autocorrelation of the random component χ2 (4)=2.79 (p=0.59)
Normality of distribution of the random component χ2(2)=4.37 (p=0.11)
Sample: 1995Q1-2004Q4
2
2.2.4 Exchange rate
In 2002, GDP per capita in Poland, adjusted for the difference in prices, was
40.5% of the average GDP per capita of Poland’s major trading partners (the weights
are shown in Tab. 2). In the same period, the price level in Poland represented 51.0% of
the price level of Poland’s major trading partners20. The basis for this long-term
exchange rate equation lies in the hypothesis that real convergence is accompanied by
nominal convergence; namely21:
s _ reert LR = 10,55− 0,75( gdp _ pot t − gdp _ ext _ pot t )
E. 17
( 0 , 022 )
where:
S_REERLR – long-term level of real effective zloty exchange rate (weights Tab. 2),
GDP_POT – potential GDP (equation E. 5),
GDP_EXT_POT – weighted (weights in Tab. 2) potential GDP abroad.
Elasticity of exchange rate with respect to the difference in potential GDPs has been
calibrated at -0.75, so that equalisation of GDP levels at home and abroad is
accompanied by the equalisation in the level of prices. If the growth rate of GDP in
Poland equals growth rates abroad, the long-term level of the real exchange rate remains
steady. In this context, it may be assumed that the long-term exchange rate path is
consistent with the relative version of the purchasing power parity theory.
Tab. 2 Share in Poland’s foreign trade and the weights adopted (%)
Weights
1995
2002
1995-2002
Germany
32.5
28.3
30.0
46.9
Italy
6.7
6.9
7.4
11.5
6.2
5.6
5.9
-
Russia
b)
adopteda)
France
4.2
6.5
5.5
8.6
Great Britain
4.6
4.5
4.5
7.1
19
The import price index is an approximate measure of material production costs, whose growth above
producer prices limits the ability to finance wage increases.
20
Calculated on the basis of OECD data, International Comparison Program.
21
A growth in variables used to determine the exchange rate denotes depreciation of the zloty.
MATERIA¸Y I STUDIA PAPER 36
29
Modules of the Economic Model
The Netherlands
5.1
4.0
4.4
6.8
Czech Republic
3.1
3.6
3.4
5.4
The United States
3.3
3.0
3.3
5.1
Sweden
2.8
2.9
2.7
4.3
Belgium
2.5
3.0
2.7
4.2
a)
The adopted weights are directly proportional to a given country’s share in Poland’s exports in 1995-
2002.
b)
2
Given the difficulties in accessing comparable historical data and a small number of published forecasts
of economic values, a decision was made not to take Russia into account.
Source: OECD.
Deviations of the real exchange rate from its long-term path depend on two
factors: the level of net foreign assets and the real interest rate disparity. An increase in
net foreign assets results in exchange rate appreciation. Such a mechanism prevents
excessive accumulation of receivables and liabilities with respect to foreign entities. The
other factor, which leads to deviation of the exchange rate from its long-term path, is the
real interest rate disparity — its increase results in appreciation of the zloty exchange
rate. Consequently, the exchange rate equation is as follows:
∆s _ reert = −0,75∆( gdp _ pot − gdp _ ext _ pot ) t −
− 0,28 ∆ (i _ w3mr _ pgdp − i3mr _ ext − s _ risk _ pr ) t −1 −
( 0 , 39 )
− 0,26( s _ reer − 10,55 + 0,75( gdp _ pot − gdp _ ext _ pot ) + 0,25(
( 0 , 09 )
NFA
)+
4 * GDP _ N
E. 18
+ 1,31(i _ w3mr _ pgdp − i3mr _ ext − s _ risk _ pr )) t −1 + dummies
where:
S_REER – real effective zloty exchange rate,
i_w3mr_pgdp – real WIBOR 3M rate deflated using the GDP deflator,
i3mr_ext – real foreign WIBOR 3M rate deflated using the GDP deflator
(equation E. 19),
s_risk_pr – risk premium valuation22,
GDP_POT – potential GDP (equation E. 5),
GDP_EXT_POT – weighted (weights Tab. 2) potential GDP abroad,
NFA – net foreign assets (equation E. 29),
GDP_N – GDP nominal level.
Adjusted R2: 50.8%
Autocorrelation of the random component: Ȥ2(4)=2.6 (p=0.63)
Normality of distribution of the random component: Ȥ2 (2)=1.6 (p=0.45)
Sample: 1997Q3-2004Q4
22
In the case of simulations, it is assumed that the risk premium (s_risk_pr) is fixed. For forecasting
purposes, on the other hand, it is assumed that the risk premium is conditional upon public debt and
increases a quarter before parliamentary elections. Consequently, the variable is determined based on the
following equation:
30
N a t i o n a l
B a n k
o f
P o l a n d
Modules of the Economic Model
The real foreign interest rate is defined by the following formula:
§ PGDP _ EXTt
·
i3mr _ ext t = i3m _ extt − ¨¨
− 1¸¸
PGDP
EXT
_
t −4
©
¹
E. 19
where:
i3m_ext – weighted (weights Tab. 2) nominal 3M rate abroad,
PGDP_EXT – weighted (weights Tab. 2) GDP deflator abroad.
2
With a defined path of the real effective zloty exchange rate, the nominal
effective zloty exchange rate is calculated using the following identity:
S _ NEERt =
S _ REERt * PGDPt
PGDP _ EXTt
E. 20
where:
S_NEER – nominal effective zloty exchange rate – an increase corresponds to
depreciation of the zloty,
S_REER – real effective zloty exchange rate (equation E. 18),
PGDP – GDP deflator (equation E. 15),
PGDP_EXT – weighted (weights Tab. 2) GDP deflator abroad.
The zloty exchange rate against the US dollar is subsequently calculated using
the following equation:
s_usd_plnt = s_neert – 0,782s_usd_eurt – 0,071s_usd_gbpt –
- 0,053s_usd_czkt – 0,04s_usd_sekt
E. 21
where:
S_USD_PLN – nominal zloty exchange rate against the US dollar,
S_NEER – nominal effective zloty exchange rate,
S_USD_EUR – nominal euro exchange rate against the US dollar,
S_USD_GBP – nominal exchange rate of the pound sterling against the US dollar,
S_USD_CZK – nominal exchange rate of the Czech crown against the US dollar,
S_USD_SEK – nominal exchange rate of the Swedish crown against the US dollar.
The exchange rate against the euro (as in the case of other currencies) is
calculated using the following equation:
s_risk_prt=-0,05+0,49s_risk_prt-1 + 0,09(0,25GDEBT_N/GDP_N) + 0,05(i97q2+i01q2+i05q2).
MATERIA¸Y I STUDIA PAPER 36
31
Modules of the Economic Model
S _ EUR _ PLN t =
S _ USD _ PLN t
S _ USD _ EURt
E. 22
where:
S_EUR_PLN – nominal zloty exchange rate against the euro – an increase corresponds
to depreciation of the zloty,
S_USD_PLN – nominal zloty exchange rate against the US dollar (equation E. 21),
2
S_USD_EUR – nominal euro exchange rate against the dollar.
2.2.5 Interest rates
The interest rates that affect consumption in the ECMOD model are interest rates
on commercial bank loans. The central bank monitors the level of short-term interest
rates on the interbank market, and those, in turn, have an impact on interest rates on
loans. In the model, WIBOR3M is the variable directly controlled by the central bank.
WIBOR3M values are transmitted via the banking system for the purpose of
determining interest rates on loans to the non-banking sector.
Transmission mechanism for interest rates on loans granted by commercial banks
Interest rates on loans in commercial banks differ, depending on the sector to
which the loans are granted. In the ECMOD model, the capital formation equation uses
the average weighted interest on loans to enterprises23 (see equation E. 11).
The equation of transmission from WIBOR3M is as follows:
I _ RENTPt = − 0,27 − 0,44 ( I _ RENTPt −1 − 2,16 − I _ W 3M t −1 ) + 0,64 ∆I _ W 3M t +
( −2 , 76 )
( −4, 56 )
(11, 02 )
E. 23
+ I _ RENTPt −1
where:
I_RENTP –average quarterly value of average weighted interest rates on loans to
enterprises by commercial banks,
I_W3M – average quarterly value of WIBOR3M.
Adjusted R2: 79.4%
Autocorrelation of the random component: Ȥ2(2)=0.91 (p=0.64)
Normality of distribution of the random component: Ȥ2 (2)=1.94 (p=0.38)
Sample: 1995Q1 – 2004Q4
23
Average weighted interest rate on loans used in the model: quarterly average from the data published
by the Statistical Department of the NBP, adjusted for the effects of changes in methodology in January
2004 and March 2002.
32
N a t i o n a l
B a n k
o f
P o l a n d
Modules of the Economic Model
(TXDWLRQ ( LV FRQVLVWHQW ZLWK WKH SUHYDLOLQJ OLWHUDWXUH HJ :HWK
:LQNHU &KPLHOHZVNL 7KH SDUDPHWHUV RI HTXDWLRQ ( KDYH EHHQ
HVWLPDWHGZLWKRXWUHVWULFWLRQV7KHXQLWIOH[LELOLW\RIWKHDYHUDJHLQWHUHVWUDWHRQORDQV
DJDLQVWWKHH[FKDQJHUDWHRQWKHLQWHUEDQNPDUNHWKDVEHHQWHVWHGXVLQJDQLQGHSHQGHQW
ORQJWHUPHTXDWLRQDQGKDVQRWEHHQUHMHFWHG7KHUHVXOWVKRZVWKDWFKDQJHVLQLQWHUHVW
UDWHVRQWKHLQWHUEDQNPDUNHWIXOO\IORZLQWRFKDQJHVLQLQWHUHVWUDWHVRQFRUSRUDWHORDQV
VHH)LJ
2
)LJ
7UDQVIHU RI FKDQJHV LQ LQWHUHVW UDWHV RQ WKH LQWHUEDQN PDUNHW WR LQWHUHVW UDWHV RQ
FRUSRUDWH ORDQV± WKH GLDJUDP VKRZV WKH UHDFWLRQ RI WKH ,B5(173 LQWHUHVW UDWH WR
WKHLQFUHDVHLQ,B:0E\RQHSHUFHQWDJHSRLQW
,B5(173
3HUFHQWDJHSRLQWV
4XDUWHUV
6RXUFH 1%3HVWLPDWHV
(;7(51$/6(&725
7KLV SDUW RI WKH SDSHU SURYLGHV D GHVFULSWLRQ RI WKH LPSDFW RI WKH H[WHUQDO
HQYLURQPHQW RQ 3RODQG¶V HFRQRP\ ,Q WKH (&02' PRGHO WKH H[WHUQDO HQYLURQPHQW
LQIOXHQFHV WKHGRPHVWLF HFRQRP\ WKURXJK WKUHH FKDQQHOV )LUVW VXFK YDULDEOHV DV WKH
GLVSDULW\ LQ UHDO LQWHUHVW UDWHV WKH UHODWLRQ EHWZHHQ GRPHVWLF DQG IRUHLJQ SULFHV WKH
UHODWLRQ EHWZHHQ SRWHQWLDO *'3 LQ 3RODQG DQG DEURDG DV ZHOO DV QHW IRUHLJQ DVVHWVDOO
GHWHUPLQH WKH HYROXWLRQDU\ SDWK RIWKH UHDO HIIHFWLYH H[FKDQJH UDWH HTXDWLRQ (
6HFRQG WKH QRPLQDO H[FKDQJH UDWH DQGH[WHUQDO SULFHV KDYH DQ LPSDFW RQ H[SRUW DQG
LPSRUW GHIODWRUV ZKLFK LQ WXUQ LQIOXHQFH GRPHVWLF SULFHV 7KLUG WKH OHYHO RI WKH UHDO
H[FKDQJH UDWH H[WHUQDO GHPDQG DQG GRPHVWLF GHPDQG LQIOXHQFH QHW H[SRUWV DQG WKH
FXUUHQWDFFRXQWEDODQFH
MATERIA¸Y I STUDIA PAPER 36
33
Modules of the Economic Model
2.3.1 Balance of payments24
The current account balance (CAB) is described with the following identities:
2
CABt = CAB _ NTt + CAB _ INCt + CAB _ TRANS t
E. 24
CAB _ NTt = GDP _ EXP _ N t − GDP _ IMP _ N t
E. 25
GDP _ EXP _ N t = GDP _ EXPt * P _ EXPt
E. 26
GDP _ IMP _ N t = GDP _ IMPt * P _ IMPt
E. 27
where:
CAB – current account balance in zloty (equation E. 24),
CAB_NT – balance of goods and services (equation E. 25),
CAB_INC – income balance,
CAB_TRANS – current transfers balance,
GDP_EXP_N – nominal value of exports of goods and services according to the
national accounts (equation E. 26),
GDP_IMP_N – nominal value of imports of goods and services according to the
national accounts (equation E. 27),
GDP_EXP – exports of goods and services in constant prices according to the national
accounts (equation E. 31),
GDP_IMP – imports of goods and services in constant prices according to the national
accounts (equation E. 32),
P_EXP – export deflator according to the national accounts (equation E. 33),
P_IMP – import deflator according to the national accounts (equation E. 35).
Section 2.3.2 presents estimates of the imperfect substitute model on the basis
of which the balance of goods and services is determined. When making forecasts,
balances of income and transfers are exogenous variables. For the sake of the
simulation, it is assumed that the income balance is proportional to net foreign assets,
i.e.
CAB _ INCt ~ 0,013 NFAt −1
E. 28
8, 3
where:
NFAt – value of net foreign assets at the end of period t (equation E. 29).
Adjusted R2: 71.5%
Autocorrelation of the random component: Ȥ2(4)=11.7 (p=0.02)
Normality of distribution of the random component: Ȥ2 (2)=0.32 (p=0.85)
Sample: 1995Q1-2004Q3
24
34
Transaction-based balance of payments is discussed.
N a t i o n a l
B a n k
o f
P o l a n d
Modules of the Economic Model
The above estimates imply that revenues on net foreign assets amount to about
1.3% quarterly, i.e. slightly above 5.0% annually. The value of net foreign assets is
measured as cumulated balances in current accounts (cf.: Lane and Milesi-Ferretti,
1999):
NFAt = NFAt −1 (1 + δ tNFA ) + CABt
E. 29
The value of the net investment position as at the end of 2003 has been assumed
to be the initial value of net foreign assets. Due to the impact of exchange rate
2
movements on the value of net foreign assets, the following adjustment has been
introduced:
0,4
δ tNFA =
0,6
S _ USD _ PLN t * S _ EUR _ PLN t
−1
0,4
0,6
S _ USD _ PLN t −1 * S _ EUR _ PLN t −1
E. 30
where:
S_USD_PLNt – zloty exchange rate per US dollar (equation E. 21),
S_EUR_PLNt – zloty exchange rate per euro (equation E. 22).
2.3.2 Foreign trade
The set of equations describing factors impacting foreign trade growth is based
on the theory of imperfect substitutes (cf. Goldstein M., Khan M., 1985), according to
which there is no perfect substitution of imported (exported) goods for goods present in
the domestic market. This is an assumption confirmed by reality, in particular in
countries with a diverse structure of imports and exports.
2.3.2.1 Exports of goods and services
It has been assumed that, in the long-term, real volumes of Polish exports are
influenced by supply factors. An equation has been adopted, based on a transformation
function, according to which the share of exports in potential GDP is a function of the
relation between export prices and domestic producer prices. For the short-term, it has
been assumed that exports are largely influenced by domestic demand. A combination
of trend and long-term dynamics is aimed at allowing for phenomena described by
Mroczek and Rubaszek (2003 and 2004), such as integration with the European Union,
trade liberalisation or inflow of foreign direct investments. The following results were
obtained:
∆gdp _ exp t = 0,02 − 0,35( gdp _ exp + 1,44
( 2 , 96 )
( 3, 92 )
− 0,01trend − gdp _ pot − 0,74( p _ exp − pgdp )) t −1 + 1,01 ∆gdp _ ext t + dummies
(10 ,1)
MATERIA¸Y I STUDIA PAPER 36
(1, 96 )
E. 31
( 0 ,81)
35
Modules of the Economic Model
where:
GDP_EXP – exports of goods and services in constant prices according to the national
accounts,
GDP_POT – level of potential GDP (equation E. 5),
GDP_EXT – weighted GDP for Poland’s major trading partners (weights Tab. 2),
P_EXP – export deflator according to the national accounts (equation E. 33),
PGDP – GDP deflator (equation E. 15).
2
Adjusted R2: 56.0%
Autocorrelation of the random component: Ȥ2(4)=6.0 (p=0.20)
Normality of distribution of the random component: Ȥ2 (2)=0.2 (p=0.90)
Sample: 1995Q2-2004Q3
2.3.2.2 Imports of goods and services
In the case of import volume, an assumption has been made that supply is
unlimited. For the long-term, a function specification with constant elasticity of
substitution has been adopted, according to which the share of imports in potential GDP
depends on the domestic-import price ratio. The goal of adding the trend to the longterm growth rate is to take into account the same phenomena as in the export volume
equation. Therefore, an equality requirement has been imposed on the trend parameters
in the export and import equations. The following results were obtained:
∆gdp _ impt = 0,01− 0,12( gdp _ imp + 1,44
( 0 , 91)
(1, 75 )
− 0,01trend − gdp _ pot + 0,70( p _ imp _ noil + gtar _ tr − pgdp))t −1 + 1,59 ∆gdpt + E. 32
(10 ,1)
(1, 3 )
( 2 , 26 )
+ dummies
where:
GDP_IMP – import of goods and services in constant prices according to the national
accounts,
GDP_POT – level of potential GDP (equation E. 5),
GDP – Gross Domestic Product (equation E. 1),
P_IMP_NOIL – import deflator according to the national accounts, adjusted for oil price
fluctuations (equation E. 34),
GTAR_TR – rate of customs duties,
PGDP – GDP deflator (equation E. 15).
Adjusted R2: 40.1%
Autocorrelation of the random component: Ȥ2(4)=5.5 (p=0.24)
Normality of distribution of the random component: Ȥ2 (2)=1.72 (p=0.42)
Sample: 1995Q2-2004Q3
36
N a t i o n a l
B a n k
o f
P o l a n d
Modules of the Economic Model
2.3.2.3 Export deflator
The export deflator is modelled on the concept “price maker — price taker”,
adjusted for the convergence effect. It has been assumed that export profitability,
influencing export volumes according to equation E. 31 , depends on the deviation of
the real exchange rate from its long-term path (cf. E. 17). In other words, exchange rate
appreciation resulting from long-term fundamentals does not change export
profitability, whereas an appreciation of the zloty due to factors other than those
2
mentioned in E. 17 does influence exporters’ decisions. The following results were
obtained:
∆p _ expt = ∆pgdpt − 0,61(( p _ exp− pgdp ) + 6,12 − 0,58( s _ reer +
( 4 , 92 )
( 6 , 31)
+ 0,75( gdp _ pot − gdp _ pot _ ext )))t −1 + 0,45 ∆ ( s _ reer )t + dummies
E. 33
( 4 , 72 )
where:
P_EXP – export deflator according to the national accounts,
PGDP – GDP deflator (cf. equation E. 15),
S_REER – real effective zloty exchange rate (equation E. 18),
GDP_POT – potential GDP (equation E. 5),
GDP_EXT_POT – weighted potential GDPs abroad (weights Tab. 2).
Adjusted R2: 70.4%
Autocorrelation of the random component: Ȥ2(4)=4.8 (p=0.30)
Normality of distribution of the random factor: Ȥ2 (2)=2.1 (p=0.35)
Sample: 1995Q2-2004Q3
2.3.2.4 Import deflator
The import deflator adjusted for oil prices movements is also modelled on the
concept “price maker — price taker”, adjusted for the convergence effect. As in the
export deflator equation, it has been assumed that the measure of price competitiveness
influencing import volumes, as in E. 32, depends on the deviation of the real exchange
rate from its long-term path (cf. E. 17). The following results were obtained:
∆p _ imp _ noil t = ∆pgdp t − 0,21(( p _ imp _ noil − pgdp) + 6,88 − 0,61( s _ reer +
( 2 , 54 )
( 4 , 32 )
+ 0,75( gdp _ pot − gdp _ ext _ pot ))) t −1 + 0,63 ∆( s _ reer ) t + dummies
E. 34
( 4 , 32 )
where:
P_IMP_NOIL – import deflator adjusted for oil price movements according to the
national accounts,
PGDP – GDP deflator (cf. equation E. 15),
MATERIA¸Y I STUDIA PAPER 36
37
Modules of the Economic Model
S_REER – real effective zloty exchange rate (equation E. 18),
GDP_POT – potential GDP (equation E. 5),
GDP_EXT_POT – weighted potential GDPs abroad (weights Tab. 2).
Adjusted R2: 57.0%
Autocorrelation of the random component: Ȥ2(4)=3.0 (p=0.55)
Normality of distribution of the random component: Ȥ2 (2)=0.2 (p=0.90)
Sample: 1995Q2-2004Q3
2
The import deflator is derived from the weighted sum of the import price
deflator, adjusted for oil price movements and oil prices on global markets denominated
in zloty:
p _ impt = 0,92 p _ imp _ noil t + 0,08(oil + s _ usd _ p ln) t
E. 35
where:
P_IMP – import deflator according to the national accounts,
P_IMP_NOIL – import deflator adjusted for oil price movements according to the
national accounts (cf. equation E. 34),
OIL – US dollar-denominated world oil prices,
S_USD_PLN – zloty exchange rate per US dollar (cf. equation E. 21).
2.4 GENERAL GOVERNMENT
The fiscal module of the model is based on data covering general government
(GG), which includes the central budget, local governments, special state funds, the
National Health Fund, government agencies and extra-budgetary entities. Thus,
problems related to modelling the effect of decentralisation of the public administration
or transfers between entities of the general government, have been avoided. In this
section, we model the volumes of GG revenues and expenditures and the resulting
deficit and public debt. All equations presented below are identities. There are
numerous interactions between specific categories of expenditures and revenues and the
real economy, frequently resulting from legal regulations (e.g. determining a tax base
for tax revenues).
The fiscal module has been constructed on the basis of nominal data, published
in Reports on the execution of the Budget Act for the years 1995-2003. Annual data
from that source have been converted into quarterly data, on the basis of other available
information. For example, annual central budget data has been broken down into
quarters on the basis of monthly information from Operational reports on the execution
38
N a t i o n a l
B a n k
o f
P o l a n d
Modules of the Economic Model
of the central budget; local government data — on the basis of quarterly Reports on the
execution of budgets of local governments; and data concerning some agencies and
funds has been broken down on the basis of quarterly information on their current
financial standing as submitted to the NBP under the existing reporting system. A small
percentage of data, which could not be broken down into quarters in this manner, has
been divided into quarters on the basis of nominal GDP growth, nominal wage growth
or the consumer price index, depending on the nature of the data.
2
2.4.1 GG revenues
GG revenues represent the sum of individual revenue categories, as presented
in the following equation:
GINC_N = GIDT_N + GPIT_N + GCIT_N + GLT_N + GTAR_N +
GCT_N + CGNBP_N + GINC_UE_N + GRT_N
E. 36
where:
GINC_N – revenues of the general government (equation E. 36),
GIDT_N – indirect tax revenues (equation E. 37),
GPIT_N – personal income tax revenues (equation E. 41),
GCIT_N – corporate income tax revenues (equation E. 42),
GLT_N – health and social contributions (equation E. 43),
GTAR_N – customs revenues (equation E. 48),
GCT_N – real estate tax revenues (equation E. 49),
CGNBP_N – the NBP profit transfer (exogenous variable),
GINC_UE_N – special cash-flow facility from the EU budget (exogenous variable),
GRT_N – other revenues (equation E. 50).
In the accession treaty, Special EU funds were allocated to Poland to
compensate for the negative effects of accession (known as a special cash-flow facility).
These funds will continue to flow into the central budget in the years 2004-2006, in the
amount specified in the treaty. For this reason, the GINC_UE_N is treated as an
exogenous variable in the model. The other revenue categories (except for the NBP
profit transfer) are endogenous variables.
In all tax categories, revenues from social contributions and customs duties
are determined on the basis of effective tax rates and the tax base. The effective tax rate
makes it possible to include a number of factors, such as the nominal rate, the system of
tax deductions and exemptions or the share of grey economy, in a single variable. It is
MATERIA¸Y I STUDIA PAPER 36
39
Modules of the Economic Model
which data are available. For the forecast period, rates are adjusted for the effects of
known systemic changes, e.g. freezing of PIT tax thresholds in 2005, growth of the
health insurance contribution rate from 8.25% in 2004 to 8.50% in 2005, 8.75% in 2006
and 9.00% in 2007, additional personal income tax revenues in 2007 (the 2006 annual
settlement) due to elimination of the home improvement tax deduction in 2006.
2.4.1.1 Indirect tax revenues
2
Indirect tax revenues can be divided into value-added tax revenues, excise tax
revenues and gaming tax revenues, i.e.:
GIDT_N = GVAT_N + GEXT_N + GGAM_N
E. 37
where:
GVAT_N –value added tax revenues (equation E. 38),
GEXT_N – excise tax revenues (equation E. 39),
GGAM_N – gaming tax revenues (equation E. 40).
The tax base for VAT and the excise tax is calculated as the sum of private
consumption, public purchases (which account for about 30% of current government
expenditure) and public investment25, which is illustrated with the following identities:
GVAT_N = GVAT_TR * (CONP_N + 0,3 * GCE_N + GINV_N)
E. 38
GEXT_N = GEXT_TR * (CONP_N + 0,3 * GCE_N + GINV_N)
E. 39
where:
GVAT_TR – effective VAT rate (exogenous variable),
GEXT_TR – effective excise tax rate (exogenous variable),
CONP_N – nominal individual consumption (equation E. 7),
GCE_N – current government expenditure, consisting of expenditure on purchases of
non-investment goods (about 30% of total spending) and wages (about 70% of
spending) (equation E. 52),
GINV_N – public investment (equation E. 61).
Gaming tax revenues are a product of the effective tax rate and private
consumption.
GGAM_N = GGAM_TR * CONP_N
25
40
E. 40
Public entities may not deduct VAT paid on their own investment since they are not VAT payers.
N a t i o n a l
B a n k
o f
P o l a n d
Modules of the Economic Model
where:
GGAM_TR – gaming tax effective rate (exogenous variable),
CONP_N – nominal individual consumption (equation E. 7).
2.4.1.2 PIT revenues
The tax base for PIT revenues consists of the wage fund increased by retirement
and disability pension transfers and unemployment benefits, and is illustrated by the
2
following relation:
GPIT_N = GPIT_TR * (GTR_RETIRED_N + GTR_UNEMP_N +
WAGEFUND_N)
E. 41
where:
GPIT_TR – effective PIT rate (exogenous variable),
GTR_RETIRED_N – retirement and disability pension transfers (equation E. 58),
GTR_UNEMP_N – unemployment benefits (equation E. 54),
WAGEFUND_N – quarterly wage fund in the economy.
2.4.1.3 CIT revenues
Corporate entities pay CIT on profit generated, i.e. on revenues less expenses.
CIT revenues are forecast on the basis of the assumed constant ratio of CIT revenues to
gross added value26 across the whole country, net of labour costs and depreciation of
fixed assets. Thus, the so-called tax base in the case of CIT revenues includes gross
domestic product less indirect taxes27, wage fund, non-wage labour costs, depreciation
of fixed assets28 and is increased by subsidies to enterprises29.
GCIT_N = GCIT_TR * (GDP_N – GIDT_N + GTR_CORP_N –
WAGEFUND_N – GLT_CORP_N – KEP * PGDP * DEPR)
E. 42
where:
GCIT_TR – effective CIT rate (exogenous variable),
GDP_N – nominal GDP value,
GIDT_N – indirect tax revenues (equation E. 37),
GTR_CORP_N – subsidies to enterprises (equation E. 62),
WAGEFUND_N – quarterly wage fund in the economy,
26
Gross added value is constructed as GDP value decreased by taxes on products and increased by
subsidies to products.
27
Indirect taxes are a variable replacing taxes on products.
28
Depreciation of fixed assets is a variable replacing amortisation of fixed assets.
29
Subsidies to enterprises are a variable replacing subsidies to products.
MATERIA¸Y I STUDIA PAPER 36
41
Modules of the Economic Model
GLT_CORP_N – quarterly non-wage labour costs (equation E. 45),
KEP – capital value equal to cumulated investment (equation E. 2),
PGDP – GDP deflator (equation E. 15),
DEPR – quarterly depreciation rate (exogenous variable).
2.4.1.4 Revenues from health and social insurance contributions
2
Revenues from health and social contributions consist of the sum of social
contributions (paid by employees, employers, business owners and freelance
employees) and health insurance contributions (paid by employees, business owners,
freelance employees and pensioners).
GLT_N = GLT_EMP_N + GLT_CORP_N + GLT_SB_N + GLT_HC_N
E. 43
GLT_EMP_N = GLT_EMP_TR * WAGEFUND_N
GLT_CORP_N = GLT_CORP_TR * WAGEFUND_N
E. 44
E. 45
GLT_SB_N = GLT_SB_TR * WAGEFUND_N
E. 46
GLT_HC_N = GLT_HC_TR * (WAGEFUND_N + GTR_RETIRED_N)
E. 47
where:
GLT_EMP_N – social contributions paid by employees (equation E. 44),
GLT_CORP_N – social contributions paid by employers (equation E. 45),
GLT_SB_N – social contributions paid by business owners and freelance employees
(equation E. 46),
GLT_HC_N – health insurance contributions (equation E. 47),
GLT_EMP_TR – effective rate of revenues from social contributions paid by
employees (exogenous variable),
GLT_CORP_TR – effective rate of revenues from social contributions paid by
employers (exogenous variable),
GLT_SB_TR – effective rate of revenues from social contributions paid by business
owners and freelance employees (exogenous variable),
GLT_HC_TR – effective rate of revenues from health insurance contributions
(exogenous variable).
2.4.1.5 Customs revenues
Customs revenues depend on the value of imported goods30. The following
identity is used:
30
Since Poland joined the European Union, i.e. the customs union, it has been receiving customs
revenues from goods crossing the EU border in Poland, but not necessarily intended for Polish end users.
Similarly, some products that are Polish imports cross the EU external border outside of Poland and thus
42
N a t i o n a l
B a n k
o f
P o l a n d
Modules of the Economic Model
GTAR_N = GTAR_TR * GDP_IMP * P_IMP
E. 48
where:
GTAR_TR – effective rate of customs duties (exogenous variable),
GDP_IMP – volume of imports of goods and services at constant prices, according
to the national accounts (equation E. 32),
P_IMP – import deflator according the national accounts (equation E. 35).
2
2.4.1.6 Real estate tax revenues
In the case of real estate tax revenues, a simplified assumption is used that the
value of real estate is proportional to GDP. The form of the assumed relation is as
follows:
GCT_N = GCT_TR * GDP_N
E. 49
where:
GCT_TR – real estate tax effective rate (exogenous variable).
2.4.1.7 NBP profit transfer and other revenues
The NBP profit transfer (comprising mostly profit on foreign reserves and open
market operations), is forecast by the NBP’s Accounting & Operations Department and
used in the ECMOD.
Other revenues - comprising dividend revenues and contributions paid by local
governments - depend on numerous factors, which are difficult to quantify. Due
to the difficulties related to modelling, it has been assumed that other GG revenues
grow at the same rate as nominal GDP.
GRT_N = GRT_N(-4) * GDP_N / GDP_N(-4)
E. 50
where:
GRT_N – other GG revenues (equation E. 50),
GDP_N – nominal GDP.
2.4.2 GG expenditure
GG expenditure consists of individual expenditure categories: current
expenditure, transfers to the population, investment spending, subsidies for economic
the respective customs duties are not paid to the Polish budget. However, data on the flows mentioned
above are not available, which prevents the construction of a better equation on customs revenues.
MATERIA¸Y I STUDIA PAPER 36
43
Modules of the Economic Model
projects, debt service costs, contribution to the EU budget, and other expenditures. It is
represented by the following equation:
GEXP_N = GCE_N + GTR_N + GINV_N + GTR_CORP_N +
E. 51
GDS_N + GEXP_UE_N + GRE_N
where:
GEXP_N – general government expenditure (equation E. 51),
2
GCE_N – current expenditure (equation E. 52),
GTR_N – transfers to the population (equation E. 53),
GINV_N – investment expenditure (equation E. 61),
GTR_CORP_N – subsidies to enterprises (equation E. 62),
GDS_N – costs of debt service (equation E. 60),
GEXP_UE_N – contribution to the European Union budget (exogenous variable),
GRE_N – other expenditure (equation E. 63).
The contribution made by member states to the EU budget is composed of four
parts: customs revenues, revenues from agricultural fees, VAT revenues and the part
of the contribution relative to the GNP volume of the member state in relation to the
GNP of the whole EU. In the case of Poland, the last part accounts for about 70% of the
total amount. In order to estimate its value for the coming years, a comparable GNP
growth forecast for the 25 EU countries would have to be applied. However, the NBP
does not have at its disposal such estimates and therefore the contribution has been
assumed exogenously on the basis of separate expert research31. The other variables are
endogenous, and their corresponding equations are presented below.
2.4.2.1 Current expenditure
Remuneration and related spending account for about 70% of current
expenditure; the remaining 30% comprises rents and purchases of non-investment
goods. This results in the following equation:
GCE_N = GCE_N(-4) * (70% * WAGE / WAGE(-4) + 30% * PGDP / PGDP(-4)) E. 52
where:
WAGE – average wage in the economy (equation E. 16),
PGDP – GDP deflator (equation E. 15).
31
Report on ”Benefits and costs of Poland’s membership in the European Union” European Centre
Natolin, Warsaw, 2003.
44
N a t i o n a l
B a n k
o f
P o l a n d
Modules of the Economic Model
2.4.2.2 Transfers to the population
Transfers to the population have been further disaggregated into unemployment
benefits, retirement and disability pensions, and other social welfare transfers, i.e.:
GTR_N = GTR_UNEMP_N + GTR_RETIRED_N + GTR_RELIEF_N
E. 53
where:
GTR_UNEMP_N – unemployment benefits (equation E. 54),
GTR_RETIRED_N – retirement and disability pensions and pre-pension benefits
2
(equation E. 58),
GTR_RELIEF_N – transfers from social assistance (equation E. 56).
The amount of unemployment benefits and social welfare transfers is calculated
as a product of the average benefit and the number of unemployed (i.e. the difference
between the labour force and the number of employed persons), under the assumption
that the number of persons requiring social welfare assistance is proportional to the
number of unemployed. An assumption has been made that indexation of transfers (i.e.
the growth of the average benefit) is based on inflation. This is illustrated by the
following equations:
GTR_UNEMP_N = AVGBENEFIT_N * (LF – EMP)
E. 54
AVGBENEFIT_N = AVGBENEFIT_N(-1) * CPI(-1) / CPI(-2)
E. 55
GTR_RELIEF_N = AVGRELIEF_N * (LF – EMP)
E. 56
AVGRELIEF_N = AVGRELIEF_N(-1) * CPI(-1) / CPI(-2)
E. 57
where:
AVGBENEFIT_N – the average quarterly benefit from the Labour Fund per
unemployed person (equation E. 55),
LF – labour force, according to BAEL (exogenous variable),
EMP – number of employees, according to BAEL (equation E. 4),
CPI – consumer price index (E. 13),
AVGRELIEF_N – the average quarterly social welfare benefit (equation E. 57).
Retirement and disability pension spending depends on the number of
beneficiaries and the amount of the average pension, as illustrated by the equation
below:
GTR_RETIRED_N = RETIRED * AVGPENSION_N
E. 58
where:
RETIRED – number of pensioners (exogenous variable),
AVGPENSION_N – the average pension (equation E. 59).
MATERIA¸Y I STUDIA PAPER 36
45
Modules of the Economic Model
The number of retirement and disability pensioners depends on the demographic
structure of the population, as well as on administrative and legal regulations —
therefore it is an exogenous variable in the model. The rate of pension indexation had
been, until recently, determined partly on the basis of regulations, and partly negotiated
by the Trilateral Commission. On March 1, 2004 indexation was performed using a
1.8% rate. According to the Act on retirement and disability pension benefits, starting in
2005, indexation will occur when accumulated inflation (since the previous indexation)
2
exceeds 5%, but not less frequently than every three years. Inflation forecasts indicate
that the next indexation will occur in 2006.
The equation for indexation of pensions is composed of two sections. The first
section describes price indexation in 2006, to be introduced on March 1st, i.e. 33% of
the indexation takes place in the first quarter of a given year and the remaining 66% in
the second quarter. The other section of the equation (average_pension_correction) is an
adjustment for flows to beneficiaries, which is determined by experts. Newly granted
retirement and disability pensions are based on average salaries in the years preceding
retirement32; thus, new beneficiaries of the system will receive a higher average benefit
than those whose benefits are only inflation-indexed33. An assumption has been made
that the average period of pension benefit pay-out is 15 years34.
AVG_PENSION_N / AVG_PENSION_N(-1) = 1 + 59/60 * (I06Q1 * (CPI(-1) /
CPI(-9) - 1) + 66% * I06Q2 * (CPI(-2) / CPI (-10) - 1)) + 1/60 *
E. 59
avg_pension_correction
where:
I06Q1 – dummy variable , equal to 1 in the first quarter of 2006, and equal to 0 in other
periods,
I04Q2 – dummy variable , equal to 1 in the second quarter of 2004, and equal to 0 in
other periods.
2.4.2.3 Costs of debt service
Costs of debt service depend on the debt volume and the effective debt interest
rate. In the model, the effective rate has been assumed as exogenous due to problems
32
33
34
46
In reality the system of pension calculation is more complex; however, it is fully based on wages.
Where nominal wages grow at a higher rate than prices.
Hence the quarterly index of 1/60.
N a t i o n a l
B a n k
o f
P o l a n d
Modules of the Economic Model
with modelling this variable. Apart from a very short time series record35, the basic
problem is the fact that costs of debt service are accounted on a cash basis and not an
accrual basis. Thus, the costs of debt service depend largely on the debt structure (both
term-structure and structure by type of debt instrument). For example, in the case of the
issuance of zero-coupon bonds, all costs of debt service are recorded (using the cash
method) on the maturity date, which distorts the picture of the real effective interest rate
of the debt.
2
In ECMOD, when preparing a fiscal forecast, the forecast for the effective
interest rate applicable to public debt is made on the basis of data from the Ministry of
Finance on the structure and maturity dates of Treasury bills and bonds.
In the model, public debt is divided into domestic and foreign debt, the value of
which denominated in zloty depends on the current exchange rate. Therefore, changes
in the exchange rate are reflected in the costs of debt service.
GDS_N = GDEBT_N * GDEBT_IR
E. 60
where:
GDEBT_N – nominal value of public debt (equation E. 65),
GDEBT_PL – share of domestic debt in public debt (exogenous variable),
GDEBT_IR – effective interest rate applicable to public debt (exogenous variable).
2.4.2.4 Investment expenditure, subsidies to enterprises and other
expenditure
The following expenditure categories are determined on the basis of discretional
decisions made by the government and the Parliament:
1) public investments,
2) subsidies to enterprises,
3) other spending of the general government, including pre-determined
contributions to international organizations (apart from the EU).
Expenditure in the above categories in 2003 accounted for about 7.3% of total
GG spending. The model assumes that the above categories change in line with growth
in nominal GDP, i.e.:
GINV_N = GINV_N(-4) * GDP_N / GDP_N(-4)
E. 61
GTR_CORP_N = GTR_CORP_N(-4) * GDP_N / GDP_N(-4)
E. 62
35
A variable indispensable to calculate the effective debt interest rate is the debt volume; estimates of
which are available from the fourth quarter of 1999 on.
MATERIA¸Y I STUDIA PAPER 36
47
Modules of the Economic Model
GRE_N = GRE_N(-4) * GDP_N / GDP_N(-4)
E. 63
where:
GINV_N – investment expenditures (equation E. 61),
GDP_N – nominal GDP,
GTR_CORP_N – subsidies to enterprises (equation E. 62),
GRE_N – other expenditures (equation E. 63).
2
2.4.3 GG deficit and public debt
The balance of the general government is calculated in a standard manner,
according to the following equation:
GDEF_N = GINC_N - GEXP_N
E. 64
As for public debt, State Treasury debt is the dominant component and, to a
lesser extent, the debt of local governments. Other general government entities,
although generating debt, have a relatively low level of indebtedness, since their debt is
usually assumed by the State Treasury.
Aggregate data on general government debt is available since 1999. Therefore,
the volume of GG debt in the period from the fist quarter of 1995 – third quarter of
1999 was estimated using information on the State Treasury debt.
The growth of debt depends on a number of factors. First, the debt is increased
by the GG deficit. Second, changes in exchange rates may have an impact on the value
of external debt calculated in domestic currency. Third, the debt may be decreased
by revenues from privatisation and increased by outflows from budget compensation.
Fourth, the shortcomings of the Polish financial system mean that not all operations
executed by the government are accounted for in the sector deficit. Fifth, the
government may fund the deficit either by borrowing, or by using cash accumulated in
accounts (so-called liquidity cushion)36. The fourth and the fifth reasons for changes in
the level of debt have been accounted for in the equation in the form of adjustments to
debt volume.
Having allowed for all the above comments, the equation looks as follows:
GDEBT_N = GDEBT_N(-1) * [(GDEBT_EUR * S_EUR_PLN + (1GDEBT_EUR) * S_USD_PLN) / (GDEBT_EUR * S_EUR_PLN(-1) + (1GDEBT_EUR) * S_USD_PLN(-1)) * (1 - GDEBT_PL) + GDEBT_PL] -
E. 65
GDEF_N - GPRIV_N + correction on pre-financing UE projects + cash
36
In this way, the deficit may be financed only temporarily, since funds in government accounts perform
the role of the so-called liquidity cushion and their level must not be too low.
48
N a t i o n a l
B a n k
o f
P o l a n d
Modules of the Economic Model
accumulated correction + hospital debt correction
where:
GDEBT_EUR – share of euro-denominated debt in total debt37 (exogenous variable),
GDEBT_PL – share of domestic debt in public debt (exogenous variable),
S_EUR_PLN – PLN/EUR exchange rate (equation E. 22),
S_USD_PLN – PLN/USD exchange rate (equation E. 21),
GDEF_N – GG balance (equation E. 64),
2
GPRIV_N – revenues from privatisation (adjusted for budget compensation)
(exogenous variable).
37
Debt denominated in currencies other than the euro and the US dollar has been attributed to the euro
and the US dollar in proportion to their share in the basket. This translates into an assumption that the
zloty exchange rate against other currencies changes in proportion to the euro and US dollar basket
exchange rate.
MATERIA¸Y I STUDIA PAPER 36
49
Simulation Analyses
3 SIMULATION ANALYSES
3
Simulation Analyses
This section of the paper deals with the simulated results of reactions of the
Polish economy to three impulses:
3
impulse of monetary policy;
exchange rate impulse;
impulse of fiscal policy.
All simulations have been conducted with a built-in interest rate rule38 and fiscal
policy rule39. During the simulations, it is necessary to allow in the model for the
function of reactions of the central bank, endogenising the short-term interest rate.
Otherwise, the results of the simulations would describe an improbable path of
development of the economy, one in which the central bank does not react to changes in
the economy. An assumption of no fiscal adjustments in the case of a shock which leads
to a growing general government deficit (and thus the debt) would be equally
unrealistic. Simulations are primarily aimed at analysing the function of reaction to an
impulse; should the interest rate rule be not taken into account, the results would be
distorted. In the process of preparation of ECMOD projections/forecasts, the interest
rate rule is substituted with an exogenously pre-defined path of short-term interest rates.
This allows for an analysis of “what if” scenarios, which are the basic reason behind
projection preparation.
In all cases, the impulse is sustained for eight quarters. Transmission channels
of individual impulses in the economy as well as the reactions of selected macroeconomic categories are presented below. The results are presented in Tab. 5-Tab. 7 in
the annexes, which show the differences between a scenario with an impulse and a
baseline scenario (scenario of economic growth without disturbances).
38
Taylor’s rule, backward-looking, cf. item 3.1. The monetary policy rule is included at the moment the
impulse is ended in the case of simulation of the interest rate. In the other two simulations, the rule is
included throughout the simulation period.
39
Cf. item 3.2. The fiscal policy rule is included at the moment the impulse is ended in case of fiscal
simulation. In the other two simulations, the rule is included with a delay, which — on the one hand —
does not distort short- and medium-term reactions of key macrocategories, and on the other — assures
that the deficit and the public debt return to their original values.
50
N a t i o n a l
B a n k
o f
P o l a n d
Simulation Analyses
3.1 MONETARY POLICY RULE
In simulations, a monetary policy rule is applied, which specification is
compliant with Taylor’s Rule (cf. Taylor (1997 and 2000)). Weights used in the output
gap, the inflationary gap and the parameter defining the smoothing of the interest rate
path have been determined on the basis of values commonly used in literature. The
value of the natural interest rate, determined at 4%, is compliant with NBP studies on its
level and changes through time (cf. Brzoza-Brzezina 2003). The value of 4% is in the
lower range of values received from independent estimates; a decision was made to use
the lower-end value, due to the downward trend of the natural interest rate in Poland,
3
based on the results of research. The form of Taylor’s Rule used in the simulations is as
follows:
i _ w3mt = (1 − α ) * [φ + inf t + β GAPt + γ (inf t − inf_ cel )] + α * i _ w3mt −1
E. 66
where,
i_w3mt – nominal WIBOR3M, interest rate, quarterly average,
inf –CPI inflation (equation E. 13),
GAP – output gap (equation E. 6),
inf_cel – inflation target (i.e. 2.5%, exogenous variable),
φ – natural interest rate,
α – smoothing coefficient,
β , γ - weights for output gap and distance of inflation from inflation target,
respectively.
3.2 FISCAL POLICY RULE
When conducting a simulation, a fiscal policy rule, which balances general
government revenues and expenditures, is applied. PIT revenues and social contribution
revenues change by adjusting their effective rates, so that the balance of GG remains at
zero. The following equation is assumed:
∆et = θ × ( d t −1 − d ∗ )
E. 67
where:
∆et – percentage growth of the effective rate,
d – balance level in relation to GDP,
d ∗ – assumed balance level in relation to GDP (0% of GDP),
θ – coefficient describing the speed of the adjustment.
MATERIA¸Y I STUDIA PAPER 36
51
Simulation Analyses
This guarantees that the balance returns to its assumed value, and through the costs of
public debt service, gradually stabilises the debt level.
The general government balance is improving gradually, so that the relation
of the balance to gross domestic product fully stabilises only in the long-term. The
parameters of the rule have been set at such a level that they stabilise the long-term
properties of the model, having little influence on the current economy. This is
necessary due to the fact that balancing of public sector revenues and expenditures
results in the weakening of automatic stabilisers in the model40. Assuming such a
solution, stabilisation of the deficit and public debt is obtained in the long-run, without
short-term shocks.
3
This rule is not an operational rule; it is only aimed at stabilising the balance
of the sector (and thus the public debt) in simulations and as an exercise for examining
long-term properties. The choice of categories of effective rates to be changed as well as
delays has been of technical nature, driven by the intention to smooth the movement of
fiscal variables (delay -1) and by the limited impact of the rule on inflation (omission of
effective rates of indirect taxes41).
3.3 FORECAST PROPERTIES OF THE MODEL – SIMULATIONS INAND OUT-OF-SAMPLE
In order to evaluate the forecasting properties of the model, out of sample
and in-sample simulations have been performed. An out-of-sample simulation begins
with model re-estimation on a shortened sample (until moment t), maintaining the
present specification of the equations42, and then the forecast is prepared for t+k
quarters forward, where k=1,2,…,8. An in-sample simulation consists in model reestimation on the sample ending at the moment t, and then on “moving back in time” to
the moment t 8 and preparing forecasts for t-p quarters, where p=7,6,5,…,1,0. In both
these exercises, actual values of exogenous variables are adopted, and forecast errors are
calculated. Such simulations enable the examination of forecasting properties of the
40
In the case of a supply shock in the economy, automatic stabilisers are not able to assure permanent
economic equilibrium. In such a situation, a fiscal adjustment is necessary, which is assured in the
ECMOD model by the application of the fiscal policy rule. In case of a demand shock, the automatic
stabilisers are able to assure a return to the equilibrium, and the fiscal policy rule weakens their impact on
the economy. The coefficient describing the speed of adjustment has been selected in such a way as to
provide a sufficiently strong impact of the rule in case a supply shock occurs and a sufficiently weak
impact in case of a demand shock.
41
It is precisely this tax category that the government can change easiest and fastest.
42
In the case of each re-estimation, the same uniform set of variables and their delays is maintained.
52
N a t i o n a l
B a n k
o f
P o l a n d
Simulation Analyses
entire model, which would be impossible only on the basis of analysing the quality of
individual equations’ fit to the data.
Out-of-sample simulations have been conducted under the assumption of gradual
extension of the sample, i.e. by the rolling-window out-of-sample forecast method.
No re-estimation of cointegrating relationships has been performed. In the first step,
the moment to start the out-of-sample forecast was chosen — it was assumed that the
shortest sample selected for model re-estimation ended in the fourth quarter of 2001.
Then, the solution of the re-estimated model from the first quarter of 2002 to the third
quarter of 2004 was examined, analysing the basic macroeconomic categories. In the
next step, the sample was extended by one quarter, and the model was re-estimated
3
again43 and solved from the second quarter of 2002 to the third quarter of 2004,
calculating forecast errors afterwards. This procedure was repeated until the model was
re-estimated on a sample ending in the second quarter of 2004, thus obtaining a forecast
for the third quarter of 2004.
It should be stressed that errors obtained in out-of-sample simulations are not
identical to forecast errors, since the simulation does not account for experts’
evaluations, which also have some impact on forecast results. Therefore, errors obtained
in such simulations are usually higher than the actual forecast errors. It is additionally
assumed that the relations incorporated in the present model version were also true for
previous years — i.e. there is an automatic re-estimation of equation parameters without
changing their specification, which may also influence the quality of forecasts obtained
in the exercise. Fig. 4 presents errors of out-of-sample forecast for inflation in 8-quarter
forecast horizon. The number of observations, on the basis of which the analysed errors
were calculated, are shown in parentheses. A small number of observations indicate a
necessity of a cautious interpretation of errors obtained in this simulation.
Tab. 3 Out-of-sample inflation simulation errors
Number of quarters for which the forecast is made
Quarters
Errors in
1
2
3
4
5
6
percentage
points
0.23
0.51
0.69
0.89
0.99
1.30
MAE
0.001 0.005 0.010 0.015 0.021 0.033
MSE
0.29
0.71
1.02
1.23
1.46
1.81
RMSE
43
7
1.56
0.038
1.96
8
2.25
0.068
2.61
At this stage, model re-estimation is made on the sample of 1995 Q1–2002 Q1.
MATERIA¸Y I STUDIA PAPER 36
53
Simulation Analyses
)LJ2XWRIVDPSOHLQIODWLRQVLPXODWLRQHUURUV
,QIODWLRQRXWRIVDPSOHVLPXODWLRQHUURUV
SHUFHQWDJHSRLQWV
TXDUWHUV
3
0$(
506(
2QHRIWKHPDLQIDFWRUVLQIOXHQFLQJWKHHUURUVLQDQRXWRIVDPSOHVLPXODWLRQ²
DSDUW IURP WKH ODFN RI H[SHUWV¶ HYDOXDWLRQV ² LV WKH DVVXPSWLRQ RI XQFKDQJHG
VSHFLILFDWLRQRIWKHHTXDWLRQV0DNLQJVXFKDQDVVXPSWLRQZHDUHQRWDEOHWRDFFRXQW
IRU WKH FKDQJHV WKDW KDYH RFFXUUHG GXH WR 3RODQG¶V DFFHVVLRQ WR WKH (XURSHDQ 8QLRQ
7KHVRFDOOHG³(8HIIHFW´FDXVHGDJURZWKLQLQIODWLRQZKLFKZDVGLIILFXOWWRIRUHVHHLQ
SUHYLRXV \HDUV 6LQFH WKH UHVXOWV SUHVHQWHG FRQFHUQ D UROOLQJZLQGRZ RXWRIVDPSOH
IRUHFDVWVLPXODWLRQWKHXQH[SHFWHGO\KLJKLQIODWLRQLQWKHVHFRQGTXDUWHURIKDG
DQ LPSDFW RQ WKH LQFUHDVHG IRUHFDVW HUURUV IURP WKH ILUVW TXDUWHU XS WR TXDUWHUV
IRUZDUG )LJ SUHVHQWV RXWRIVDPSOH IRUHFDVW HUURUV ZLWKRXW WDNLQJ LQWR DFFRXQW
IRUHFDVWVIRUWKHVHFRQGDQGWKHWKLUGTXDUWHURI7KHQXPEHURIREVHUYDWLRQVXVHG
LQFDOFXODWLQJWKHDQDO\VHGHUURUVDUHLQSDUHQWKHVHV
)LJ 2XWRIVDPSOH LQIODWLRQ VLPXODWLRQ HUURUV H[FOXGLQJ IRUHFDVWV IRU WKH VHFRQG
DQGWKHWKLUGTXDUWHURI
,QIODWLRQRXWRIVDPSOHVLPXODWLRQHUURUV
ZLWKRXW(8HIIHFW
3HUFHQWDJHSRLQWV
4XDUWHUV
0$(
506(
54
,QD³UDZ´PRGHOVROXWLRQZLWKRXWDFFRXQWLQJIRUH[SHUWV¶HYDOXDWLRQV
N a t i o n a l
B a n k
o f
P o l a n d
Simulation Analyses
Tab. 4 Out-of-sample Simulation errors for inflation, excluding forecasts for the second
and the third quarter of 2004
Number of quarters for which the forecast is made
Quarters
Errors in
1
2
3
4
5
6
7
8
percentage
points
0.16
0.28
0.31
0.47
0.31
0.52
0.77
0.93
MAE
0.000 0.001 0.002 0.003 0.001 0.004 0.012 0.009
MSE
0.20
0.34
0.40
0.57
0.36
0.62
1.11
0.94
RMSE
In the exercise excluding the EU effect, errors are much lower than in the
exercise which include the second and the third quarter of 2004. The errors for the 7th
and 8th quarters, however, should be interpreted with caution, due to a small number of
3
observations.
An in-sample simulation consists of examining forecast errors obtained in the
sample on which equations were re-estimated. Cointegrating relationships (similar to
the out-of-sample simulation) remain unchanged. In the case of the in-sample
simulation, the model is re-estimated 11 times, without changing the equations and
preserving the actual paths of exogenous variables. The first re-estimation is performed
on the sample ending in the fourth quarter of 2001. The model is solved from the fouth
quarter of 1999 to the fourth quarter of 2001 (i.e. until the end of the sample on which
the equations were re-estimated) and forecast errors are calculated. Then the model is
solved from the first quarter of 2000 to the fourth quarter of 2001 (the end of the sample
period on which the equations were re-estimated). The following forecasts are made
from one quarter later until the fourth quarter of 2001, and the procedure is repeated
until a forecast is made only for the fourth quarter of 2001. Then the entire model is reestimated using the sample ending with the first quarter of 2002, i.e. extended by one
quarter in relation to the previous re-estimation. The re-estimated model is solved,
starting from the fourth quarter of 1999 until the end of the sample (i.e. until the first
quarter of 2002). The subsequent solutions always begin one quarter later and end in the
first quarter of 2002, i.e. at the end of the sample period on which the re-estimation was
performed. Then the model is re-estimated again on the sample extended by one quarter
in relation to the previous re-estimation, and the procedure of solving the model is
repeated, always changing the starting point of the solution and leaving the last quarter
included in a solution at the end of the sample, on which the re-estimation was made.
Thus a much larger number of observations are obtained. In-sample simulation
errors are presented in figure Fig. 6.
MATERIA¸Y I STUDIA PAPER 36
55
Simulation Analyses
)LJ,QVDPSOHIRUHFDVWHUURUVIRULQIODWLRQ
,QIODWLRQLQVDPSOHVLPXODWLRQHUURUV
3HUFHQWDJHSRLQWV
4XDUWHUV
3
0$(
506(
7KH IRUHFDVW HUURUV REWDLQHG FRQFHUQ D ´UDZ´ PRGHO VROXWLRQ ZLWKRXW H[SHUWV¶
SUHGLFWLRQV 7KH ODFN RI FKDQJHV LQ WKH HTXDWLRQV LQ WKH FDVH RI WKH IRUHFDVW IURP WKH
IRXUWKTXDUWHURI GRHVQRWSURYLGHIRUWKHVWUXFWXUDOFKDQJHVWKDWKDYHRFFXUUHGLQ
3RODQG VLQFH WKDW WLPH ,Q WKH FDVH RI D ODUJH QXPEHU RI REVHUYDWLRQV DYDLODEOH LQ LQ
VDPSOHVLPXODWLRQVWKHLPSDFWRIWKH³(8HIIHFW´LVQHJOLJLEOHVLQFHLWFRQFHUQVRQO\
REVHUYDWLRQVRXWRI RIREVHUYDWLRQV +RZHYHUWKHFKDQJHVWKDWWRRNSODFH
LQWKHILUVWWZR\HDUVRIWKHDQDO\VHGSHULRGDFFRXQWIRURIREVHUYDWLRQVDQGWKXV
KDYHDVWURQJLQIOXHQFHRQWKHUHVXOWV&RQVLGHULQJWKHDERYHWKHHUURUVREWDLQHGFDQEH
WUHDWHGDVTXLWHORZ
,17(5(675$7(,038/6(
,QWHUHVW UDWHV LQIOXHQFH WKH HFRQRP\ LQ D QXPEHU RI ZD\V ,Q WKH (&02'
PRGHOWKHIROORZLQJWUDQVPLVVLRQFKDQQHOVDUHGLVWLQJXLVKHG WKHLQWHUHVWUDWHFKDQQHO
YLDWKHFRVWRIFUHGLWDQGWKHFRVWRIFDSLWDO WKHH[FKDQJH UDWHFKDQQHODQGWKHDVVHW
FKDQQHO$VLPSOLILHGWUDQVPLVVLRQPHFKDQLVPLVLOOXVWUDWHGLQ)LJ
56
N a t i o n a l
B a n k
o f
P o l a n d
Simulation Analyses
Fig. 7 Monetary transmission in the ECMOD model45
INV ↓
EMP ↓
i_rentpr ↑
i_credor↑
WAGE ↓
GAP ↓
CONP ↓
i_rentp ↑
i_credo ↑
GDP_EXP-GDP_IMP ↓
INF↓
ULC ↓
i_w3m ↓
i w3m ↑
3
(Taylor’s
rule)
i_w3mr ↑
S_REER ↓
P_IMP ↓
The monetary impulse has been defined as a rise in the short-term interest rate
controlled by the central bank by 1 percentage point over its level in the baseline
scenario and maintaining it at the increased level for eight quarters. A simulation has
been conducted with an endogenous interest rate (with Taylor’s Rule included). The
results are presented in Tab. 5 and in Fig. 10.
In the short term, the growth of the interest rate results in decreased individual
consumption due to higher interest rates on consumer loans and a decrease in
investments due to the increased cost of capital. This entails a decrease in GDP, which
induces a drop in employment. Additionally, wages decrease, which results in lower
unit labour costs. Lower employment and lower growth lead to a higher general
government deficit. Higher interest rates, through appreciation of the exchange rate, are
also reflected in a decrease in exports, influencing the decline of GDP and import
prices. Eventually, inflation declines in accordance with the above factors46.
After the monetary impulse decays, due to lower inflation and lower demand,
following the interest rate rule, the level of nominal and real interest rates drops below
the baseline scenario. Eventually, this causes an upward movement in the growth rates
of GDP components as well as wages and employment, to the level above the baseline
45
Due to a large number of relations in the ECMOD model, the presented diagram of monetary
transmission is simplified.
46
Food and fuel price paths are the same in both scenarios, thus they have no impact on the results of
their comparison.
MATERIA¸Y I STUDIA PAPER 36
57
Simulation Analyses
scenario, as well as a temporary growth in y/y inflation and its subsequent return
to the baseline level.
3.5 EXCHANGE RATE IMPULSE
In the ECMOD model, the direct influence of the exchange rate on the economy
is exerted in two ways. First, currency deprecation makes import prices grow and thus
leads to higher CPI inflation. Second, the weakening of the zloty, through changes in
relative prices, causes an improvement in the trade balance. The indirect impact on the
economy has been presented in figure Fig. 8.
3
Fig. 8 Exchange rate impulse in the ECMOD model47
The exchange rate impulse has been introduced into the ECMOD model
in the following way - the model has been solved with the interest rate rule included.
For the first eight quarters, the valuation of the FX risk premium in the second scenario
has been increased by 1 percentage point in relation to the baseline scenario. For the
remaining quarters, the risk premium has been determined at a similar level. The results,
presented in Tab. 6 and in Fig. 11, are as follows.
In the short-term, after introducing the impulse, depreciation of the currency
through an increase in exports and slightly lower imports both results in improved net
exports and an improved current account balance. Moreover, due to higher import
prices, growth of CPI inflation is in place. Higher demand and intensified inflation
processes lead to increased interest rates, according to the interest rate rule.
47
Due to a high number of interdependencies in the ECMOD model, the presented diagram of exchange
rate transmission is simplified.
58
N a t i o n a l
B a n k
o f
P o l a n d
Simulation Analyses
In subsequent quarters, higher interest rates cause a reduction in investment
demand. After eight quarters, the return of the risk premium to its baseline level results
in the return of the real exchange rate to its previous level. In the medium-term, the
main macroeconomic variables, i.e. GDP and the CPI inflation, return to their baseline
levels. The economy returns to its previous long-term development path, from which it
was temporarily pushed off.
3.6 FISCAL IMPULSE
The fiscal impulse is defined as a decrease in specific categories of public
3
expenditure — current expenditure, pension expenditure, unemployment benefits and
other social expenditure, so that their total decrease equals 1% of GDP.
A decrease in current expenditure, which is reflected in a drop in collective
consumption, has a direct impact on GDP, whereas a decrease in transfers influences the
real economy through the channel of disposable income. The mechanisms of
transmission are presented in the following figure.
Fig. 9 Fiscal impulse in the ECMOD model48
DEFICIT ↓
DEFICIT ↑
GTR ↑
INF↓
EMP ↓
GINC ↓
GCE ↓
CONGOV ↓
GTR
↓
GDP ↓
YD ↓
CONP ↓
GDP_IMP ↓
GAP ↓
i_w3m ↓
(Taylor’s
rule)
It should be noted that a decrease in public expenditure by 1% of GDP does not
necessarily mean an equivalent decrease in the general government deficit. This is due
to the feedback mechanism between the fiscal sector and the real economy (Fig. 9).
Results are presented in Tab. 7 and in Fig. 12.
Initially, a drop in government demand results in a GDP decrease. As a result
of decreased transfers, individual consumption shrinks, which implies a further GDP
48
Due to a high number of relations in the ECMOD model, the presented diagram of fiscal transmission
has been simplified.
MATERIA¸Y I STUDIA PAPER 36
59
Simulation Analyses
decrease and thus a reduction in the output gap. This results in lower inflation. Lower
demand leads to decreased imports and, thus, an improved current account balance.
Declining inflation, along with a negative output gap, according to the interest rate rule,
leads to lower interest rates. As a result, there is a rebound in demand — both for
investments and from consumers.
3
60
N a t i o n a l
B a n k
o f
P o l a n d
Concluding Remarks
4 CONCLUDING REMARKS
4
Concluding Remarks
This paper presents a quarterly macroeconometric ECMOD model, constructed
in order to forecast the main macroeconomic indicators and to perform simulation
analyses for the Polish economy. ECMOD is a structural model in which short-term
dynamics are conditioned upon demand; in the long-run, however, an important role is
played by elements of the supply side of the economy.
This paper reviews the key equations of the macromodel by modules,
and so it discusses the real sector, prices and costs, the external sector and the fiscal
4
sector. Simulated reactions of the economy to standard impulses (the interest rate
impulse, the exchange rate impulse and the fiscal impulse) are presented.
MATERIA¸Y I STUDIA PAPER 36
61
Selected References
5 SELECTED BIBLIOGRAPHY
5
Selected References
Models:
[1]
Bank of England, 1999, Economic models at the Bank of England, BoE.
[2]
Bank of England, 2005, The Bank of England Quarterly Model, BoE.
[3]
Brayton F., Tinsley P.(ed.), 1996, A Guide to FRB/US, a macroeconomic model
of the United States, Federal Reserve Bank, mimeo.
[4]
Czech National Bank, 2003, The Czech National Bank’s Forecasting and Policy
Analysis System, CNB.
[5]
Deutsche Bundesbank, 2000, Macro-econometric Multi-Country Model,
MEMMOD, DB.
[6]
Dieppe A., Henry J., 2004, The euro area viewed as a single economy: how does
it respond to shocks?, Economic Modelling, volume 21(5).
5
[7]
Fagan G., Henry J., Mestre R., 2001, An area-wide model (AWM) for the euro
area, ECB WP 42/2001.
[8]
[9]
Kłos B., 2003, Mały strukturalny model inflacji, version 3.5.6, MiS 154, NBP.
Kłos B., KokoszczyĔski R., Łyziak T., Przystupa J., Wróbel E., 2004, Modele
strukturalne w prognozowaniu inflacji w Narodowym Banku Polskim, Materials
and Studies 180, NBP.
[10]
KokoszczyĔski R., Łyziak T., Pawłowska M., Przystupa J., Wróbel E., 2002,
Mechanizm transmisji polityki pieniĊĪnej – współczesne ramy teoretyczne, nowe
wyniki empiryczne, mimeo, NBP.
[11]
McGuire M., Ryan M., Macroeconomic modelling developments in the central
bank, Bank of Ireland, mimeo.
[12]
Welfe A., Karp P., Kelm R., 2003, Makroekonometryczny kwartalny model
[13]
Willman A., Estrada A., 2002, The Spanish block of the ESCB-Multi-Country
gospodarki Polski, Wyd. UŁ, ŁódĨ.
model, ECB WP 149.
[14]
Zonzilos N., Sideris D., 2003, The Greek Model Block of the ESCB MultiCountry Model, Bank of Greece, mimeo.
62
N a t i o n a l
B a n k
o f
P o l a n d
Selected References
Specific issues:
[15]
Aschauer D., 1989, Does Public Capital Crowd Out Private Capital?, Journal
of Monetary Economics, vol. 24, s. 171-188.
[16]
Banerjee A., Russell B., A Markup Model for the Forecasting Inflation in the
Euro Area, EUI Working Paper ECO No. 2002/16.
[17]
Banerjee. A., Russem B., 2001, The relationship between the Markup and
Inflation in the G7 Economies and Australia, Review of Economics and
Statistics, vol. 83, No 2., May.
[18]
Bindseil U., 2003, Monetary Policy Implementation. Theory-Past-Present,
mimeo, European Central Bank.
[19]
Chmielewski T., 2004, Interest Rate Pass-Through in the Polish Banking Sector
and Bank-Specific Financial Disturbances, mimeo, National Bank of Poland.
[20]
Chortareas G., Driver R., 2001, PPP and the real exchange rate – real interest
differential puzzle revised: evidence from non-stationary panel data, Bank of
England, London.
[21]
CzyĪewski, A.B., 2002, SprawnoĞü i konkurencyjnoĞü polskiego rolnictwa,
5
mimeo, IERiGĩ, Warszawa.
[22]
De Bondt G., 2002, Retail bank interest rate pass-through: New evidence at the
euro area level, ECB Working Paper No. 136.
[23]
Eisner R., 1989, Budget Deficits: Rhetoric and Reality, Journal of Economic
Perspectives, vol. 3, s. 73-93.
[24]
Elmeskov J., 1993, High and persistent unemployment: assessment of the
problem and its causes, OECD Economics Department Working Paper No. 132.
[25]
Els P., Locarno A., Morgan J., Villetelle J., 2001, Monetary Policy Transmission
in the Euro Area: What Do Aggregate And National Structural Models Tell Us?,
ECB WP No. 94, Frankfurt.
[26]
Engle
R.,
Granger
C.,
1987,
Cointegration
and
Error
Correction:
Representation, Estimation, and Testing, Econometrica, vol. 55, s. 251-276.
[27]
Faruqee H., 1995, Long-Run Determinants of the Real Exchange Rate: A Stock-
Flow Perspective, IMF Staff Papers, vol. 42, no. 1, International Monetary
Fund, Washington.
[28]
Fic T., Gradzewicz M., Kolasa M., Kot A., Murawski K., Rubaszek M., 2004,
Kwartalny model makroekonometryczny gospodarki polskiej ECMOD, mimeo,
NBP.
[29]
Gagnon J., 2004, Productive Capacity, Product Varieties and the Elasticities
Approach to the Trade Balance, article presented at ECB conference “The
MATERIA¸Y I STUDIA PAPER 36
63
Selected References
external dimension of the euro area: trade, capital flows and international
macroeconomic linkages”.
[30]
Gerlach S., Smets F., 1995, The Monetary Transmission Mechanism: Evidence
From the G-7 Countries, BIS WP No. 26, Bazylea.
[31]
Gradzewicz M., Kolasa M., 2004, Szacowanie luki popytowej dla gospodarki
polskiej przy wykorzystaniu metody VECM, “Bank i Kredyt” Nr 2.
[32]
Hendry D. F., 1995, Dynamic Econometrics, Oxford University Press, New
York.
[33]
Krugman P., 1989, Differences in Income Elasticities and Trends in Real
Exchange Rates, European Economic Review No. 33, s. 1055-1085.
[34]
Lane P., Milesi-Ferretti G.M., 1999, The External Wealth of Nations: Measures
of Foreign Assets and Liabilities for Industrial and Developing Countries, IMF
WP/99/115, International Monetary Fund.
[35]
Łyziak T., Consumer Inflation Expectations in Poland, ECB Working Paper no
287, November 2003.
[36]
5
MacDonald R., 1997, What Determines Real Exchange Rates? The Long and the
Short of it, IMF Working Paper WP/97/21, International Monetary Fund,
Washington.
[37]
Marczewski K., 2002, Zmiany Kursu Walutowego a Ceny i Reakcje
PrzedsiĊbiorstw w Handlu Zagranicznym, Instytut Koniunktur i Cen Handlu
Zagranicznego, Warszawa.
[38]
McAdam P., Morgan J., 2001, The Monetary Transmission Mechanism at the
Euro-Area Level: Issues and Results Using Structural Macroeconomic Models,
ECB WP No. 93, Frankfurt.
[39]
Mishkin F., 1996, The Channels of Monetary Transmission: Lessons for
Monetary Policy, NBER WP 5464, Cambridge.
[40]
Mroczek W., Rubaszek M., 2003, Determinanty Polskiego Eksportu i Importu,
Materiala and Studies No. 161, National Bank of Poland.
[41]
Mroczek W., Rubaszek M., 2004, Development of the trade links between
Poland and the European Union in the years 1992-2002, NBP Paper No. 30,
National Bank of Poland.
[42]
Obstfeld M., Rogoff K., 1996, Foundations of International Economics, MIT
press, Cambridg, Massachusetts.
[43]
Peersman G., Smets F., 2001, The Monetary Transmission Mechanism in the
Euro Area: More Evidence From VAR Analysis, ECB WP No. 91, Frankfurt.
[44]
Taylor J., 1997, The Policy Rule Mix: A Macroeconomic Policy Evaluation,
Stanford University.
64
N a t i o n a l
B a n k
o f
P o l a n d
Selected References
[45]
Taylor J., 2000, Using Monetary Policy Rules in Emerging Market Economies,
Stanford University.
[46]
Weth M.A., 2002, The pass-through from market interest rates to bank lending
rates in Germany, Discussion Paper 11/02, Deutsche Bundesbank.
[47]
Winker P., 1999, Sluggish adjustment of interest rates and credit rationing:
an application of unit root testing and error correction modelling, Applied
Economics, vol. 31, s. 267-277.
[48]
Wróbel E., Pawłowska M., 2002, Monetary transmission in Poland” some
evidence on interest rate and credit channels, Materials and Studies, No. 24,
National Bank of Poland.
5
MATERIA¸Y I STUDIA PAPER 36
65
6
66
Tab. 5 Reaction of the economy to the interest rate impulse (deviation in % from the baseline scenario; in the case of y/y inflation and y/y
GDP, deviation in percentage points)
Quarters
CPI
Inflation y/y (CPI)
GDP
GDP growth rate y/y
Individual consumption
Investments
Exports
1
0.00
0.00
0.03
0.03
0.05
0.02
2
0.01
0.01
0.06
0.06
0.10
0.07
3
0.04
0.04
0.13
0.14
0.14
0.16
0.07
0.16
0.83
0.43
0.55
0.06
0.05
4
0.07
0.07
0.20
0.21
0.19
0.28
0.16
0.21
1.00
0.55
0.67
0.08
0.10
5
0.12
0.12
0.28
0.26
0.25
0.43
0.24
0.27
1.14
0.65
0.77
0.12
0.16
0.02 0.02 0.00
6
0.18
0.16
0.36
0.32
0.31
0.60
0.31
0.31
1.27
0.75
0.86
0.18
0.24
0.05
7
0.26
0.22
0.45
0.34
0.36
0.79
0.37
0.35
1.40
0.84
0.96
0.25
0.33
0.12
Years
8
0.35
0.28
0.55
0.37
0.43
1.00
0.43
0.40
1.51
0.93
1.05
0.34
0.43
0.22
9
0.47
0.35
0.61
0.34
0.44
1.20
0.48
0.39
1.63
1.03
1.15
0.45
0.52
0.36
10
0.59
0.41
0.65
0.30
0.44
1.37
0.53
0.34
1.09
0.84
0.81
0.56
0.62
0.52
11
0.72
0.46
0.65
0.21
0.42
1.48
0.49
0.27
0.64
0.64
0.55
0.69
0.69
0.73
Unit labour costs
P o l a n d
Unemployment
General government deficit (% of
GDP)
NEER
Export deflator
Import deflator
Public debt (% of GDP)
Source: NBP calculations.
1
0.03
0.03
0.11
0.11
0.12
0.13
0.06
0.13
0.62
0.32
0.41
0.05
0.05
2
0.22
0.19
0.41
0.32
0.34
0.71
0.34
0.33
1.33
0.79
0.91
0.22
0.29
0.01 0.10
3
0.65
0.43
0.64
0.24
0.43
1.40
0.48
0.30
0.91
0.75
0.71
0.63
0.64
0.64
4
5
11
21
1.16 1.60 1.59 1.80
0.51 0.44 0.17 0.03
0.48 0.25 0.01 0.05
0.16 0.24 0.07 0.01
0.30 0.16 0.01 0.04
1.44 0.95 0.14 0.28
0.08 0.25 0.20 0.02
0.01 0.09 0.09 0.04
0.11 0.22 2.06 2.01
0.36 0.77 1.95 1.97
0.21 0.66 1.98 1.98
1.25 1.87 1.67 1.88
0.72 0.47 0.16 0.07
1.48 2.09 1.50 1.86
0.59 0.04 0.23 0.51 0.58 0.38 0.13 0.06
0.37 0.04 0.17 0.31 0.39 0.28 0.02 0.01
4.85 0.04 0.95 3.54 6.40 8.27 5.97 3.38
Annex
Nominal wages
o f
0.01 0.02 0.04 0.08 0.13 0.19 0.26 0.34 0.42 0.49 0.55
0.01 0.02 0.04 0.07 0.10 0.14 0.18 0.23 0.26 0.29 0.33
0.08 0.22 0.10 0.08 0.37 0.71 1.08 1.64 2.14 3.18 3.97
Imports
B a n k
0.01 0.01
N a t i o n a l
Employed persons (BAEL)
0.00 0.00
0.05 0.09
0.00 0.64
0.00 0.29
0.00 0.42
0.01 0.03
0.01 0.02
12
0.85
0.50
0.63
0.09
0.40
1.54
0.40
0.19
0.29
0.48
0.35
0.84
0.74
0.95
Annex
MATERIA¸Y I STUDIA PAPER 36
Tab. 6 Reaction of the economy to the exchange rate impulse (deviation in % from the baseline scenario; in case of y/y inflation and y/y
GDP, deviation in percentage points)
Quarters
Years
1
2
3
4
5
6
7
8
9
10
11
12
1
2
3
4
5
11
21
CPI
0.06
0.09
0.12
0.15
0.19
0.24
0.29
0.34
0.39
0.44
0.49
0.53
0.10
0.26
0.46
0.61
0.67
-0.02
0.15
Inflation y/y (CPI)
0.06
0.08
0.09
0.11
0.13
0.15
0.17
0.19
0.20
0.20
0.20
0.19
0.09
0.16
0.20
0.14
0.06
-0.11
-0.00
GDP
0.10
0.13
0.15
0.17
0.20
0.20
0.21
0.21
0.22
0.19
0.15
0.10
0.14
0.21
0.17
-0.01
-0.15
-0.16
0.02
GDP growth rate y/y
0.10
0.12
0.11
0.11
0.10
0.08
0.07
0.04
0.02 -0.01 -0.07 -0.11
0.11
0.07
-0.04
-0.18
-0.15
0.05
0.00
0.00
-0.00
-0.00
0.00
-0.02
-0.06
-0.16
0.03
Individual consumption
-0.00 -0.00 -0.00 -0.00 -0.00 -0.00 -0.00 -0.00 -0.00
0.00
0.00
Investments
0.01
0.01 -0.00 -0.02 -0.04 -0.08 -0.12 -0.18 -0.24 -0.32 -0.40 -0.49
-0.00
-0.11
-0.37
-0.71
-0.96
-0.10
0.16
Exports
0.23
0.30
0.22
0.32
0.43
0.34
0.01
-0.15
-0.12
0.00
Imports
0.03
0.01 -0.00 -0.03 -0.04 -0.08 -0.11 -0.15 -0.17 -0.23 -0.28 -0.32
0.00
-0.09
-0.25
-0.37
-0.37
-0.10
0.10
NEER
1.09
1.17
1.21
1.23
1.24
1.22
1.21
1.20
1.19
0.55
0.37
0.25
1.18
1.22
0.59
0.19
0.42
0.16
0.10
Export deflator
0.63
0.69
0.73
0.76
0.77
0.78
0.79
0.80
0.82
0.55
0.43
0.36
0.70
0.79
0.54
0.34
0.52
0.02
0.15
Import deflator
0.73
0.79
0.82
0.84
0.85
0.85
0.85
0.85
0.87
0.46
0.36
0.30
0.80
0.85
0.49
0.30
0.51
0.03
0.15
Nominal wages
0.01
0.01
0.03
0.05
0.09
0.13
0.18
0.23
0.29
0.34
0.40
0.46
0.03
0.16
0.37
0.59
0.68
-0.22
0.22
Employed persons (BAEL)
0.05
0.08
0.11
0.15
0.18
0.21
0.23
0.25
0.27
0.28
0.28
0.26
0.10
0.22
0.27
0.16
-0.01
-0.06
-0.03
-0.04 -0.03 -0.01
0.03
0.07
0.13
0.20
0.27
0.34
0.42
0.52
0.61
-0.01
0.17
0.47
0.76
0.82
-0.11
0.18
Unemployment
-0.04 -0.07 -0.09 -0.11 -0.14 -0.16 -0.18 -0.20 -0.21 -0.22 -0.22 -0.20
General government deficit (%
of GDP)
0.01 0.02 0.04 0.05 0.06 0.07 0.08 0.10 0.09 0.08 0.09 0.10
-0.08
-0.17
-0.22
-0.13
0.01
0.05
0.02
Unit labour costs
Public debt (%of GDP)
Source: NBP calculations.
0.23
0.14
0.35
0.39
0.42
0.43
0.44
0.44
0.43
0.41
0.32
0.04 -0.10 -0.30 -0.46 -0.63 -0.84 -1.07 -1.68 -1.84 -2.01
0.03
0.08
0.09
0.08
0.01
0.02
-0.00
0.08
-0.56
-1.65
-2.20
-2.13
1.27
-0.13
67
6
6
68
Tab. 7 Reaction of the economy to the fiscal impulse (deviation in % from the baseline scenario; in the case of y/y inflation and y/y GDP,
deviation in percentage points)
Quarters
1
12
1
2
3
4
5
11
21
CPI
-0.01 -0.05 -0.09 -0.15 -0.21 -0.31 -0.42 -0.53 -0.64 -0.76 -0.88 -0.98
-0.07
-0.37
-0.82
-1.22
-1.47
-0.56
-0.91
Inflation y/y (CPI)
-0.01 -0.05 -0.09 -0.15 -0.20 -0.26 -0.33 -0.38 -0.44 -0.46 -0.47 -0.46
-0.07
-0.29
-0.45
-0.40
-0.26
0.22
0.03
GDP
-0.30 -0.24 -0.30 -0.29 -0.60 -0.51 -0.55 -0.50 -0.51 -0.43 -0.35 -0.24
-0.28
-0.54
-0.38
-0.06
0.22
0.42
0.35
GDP growth rate y/y
-0.32 -0.25 -0.31 -0.31 -0.31 -0.29 -0.27 -0.22
0.28
-0.30
-0.27
0.17
0.34
0.29
-0.10
-0.02
Individual consumption
-0.05 -0.11 -0.17 -0.23 -0.30 -0.36 -0.41 -0.46 -0.44 -0.38 -0.29 -0.19
-0.14
-0.38
-0.33
0.09
0.53
1.50
1.39
Investments
-0.01
0.00
0.01
0.04
0.07
0.14
0.22
0.32
0.45
0.61
0.78
0.95
0.01
0.19
0.70
1.38
1.93
1.25
0.95
0.00
0.00
0.01
0.02
0.04
0.07
0.11
0.16
0.22
0.27
0.33
0.37
0.01
0.10
0.30
0.43
0.39
0.11
0.32
-0.48 -0.32 -0.38 -0.34 -0.80 -0.58 -0.60 -0.49 -0.48 -0.33 -0.23 -0.07
-0.38
-0.62
-0.28
0.11
0.35
0.54
0.27
0.11
0.48
0.57
-0.14
-1.16
-1.60
-1.28
Exports
Imports
N a t i o n a l
NEER
0.00
2
0.07
3
0.14
4
0.22
5
0.30
6
Years
0.45
7
0.54
8
0.61
9
0.10
0.64
10
0.09
0.64
11
0.21
0.56
0.42
Export deflator
0.00
0.04
0.07
0.11
0.14
0.21
0.24
0.25
0.23
0.19
0.10 -0.03
0.06
0.21
0.12
-0.48
-1.25
-1.01
-1.02
Import deflator
0.00
0.05
0.10
0.15
0.19
0.28
0.32
0.34
0.33
0.29
0.19
0.05
0.07
0.28
0.21
-0.44
-1.27
-1.08
-1.05
Nominal wages
-0.13 -0.15 -0.20 -0.23 -0.40 -0.46 -0.56 -0.64 -0.75 -0.84 -0.94 -1.03
-0.18
-0.52
-0.89
-1.27
-1.52
-0.00
-0.65
Employed persons (BAEL)
B a n k
o f
-0.16
-0.49
-0.63
-0.43
-0.10
0.11
0.07
0.10 -0.05 -0.09 -0.19 -0.17 -0.41 -0.54 -0.73 -0.88 -1.06 -1.22 -1.38
-0.06
-0.46
-1.14
-1.64
-1.84
-0.31
-0.93
Unemployment
General government deficit (% of
GDP)
0.06
0.10
0.16
0.20
0.29
0.36
0.43
0.48
0.51
0.52
0.51
0.47
0.13
0.39
0.50
0.34
0.08
-0.09
-0.06
0.63
0.62
0.59
0.57
0.87
0.85
0.81
0.79
0.30
0.17
0.07
0.02
0.60
0.83
0.14
0.04
0.09
-0.03
0.00
0.50 -0.38 -0.52 -1.16 -0.60 -1.57 -1.72 -2.58 -2.35 -2.41 -2.29 -2.39
-0.39
-1.62
-2.36
-2.22
-2.55
-5.60
-2.35
Public debt (% of GDP)
Source: NBP calculations.
Annex
P o l a n d
-0.07 -0.13 -0.19 -0.25 -0.36 -0.46 -0.54 -0.60 -0.64 -0.65 -0.64 -0.59
Unit labour costs
Annex
Fig. 10. Reaction of the economy to the interest rate impulse
Perc. pts
0.30
Perc. pts
0.30
CPI y/y
0.20
Perc. pts
0.20
GDP y/y
Individual consumption y/y
0.15
0.20
0.10
0.10
0.10
0.00
0.05
-0.10
0.00
0.00
-0.20
-0.10
-0.05
-0.30
-0.10
-0.20
-0.40
-0.15
-0.30
-0.50
-0.60
-0.20
-0.40
1
4
7
Perc. pts
0.80
10
13
16
19
22
-0.25
1
Investments y/y
0.40
0.40
0.30
7
Perc. pts
0.50
0.60
4
10
13
16
19
1
22
Exports y/y
4
7
Perc. pts
0.40
10
13
16
19
22
16
19
22
16
19
22
16
19
22
Imports y/y
0.30
0.20
0.20
0.20
0.00
0.10
-0.20
0.00
-0.40
-0.10
-0.60
-0.20
0.10
0.00
-0.10
-0.80
-0.20
-0.30
-1.00
-0.40
1
4
7
%
10
13
16
19
22
-0.30
1
Perc. pts
0.60
NEER
4
7
10
13
16
19
Export deflator y/y
4
Perc. pts
0.60
0.50
7
10
13
Import deflator y/y
0.40
0.40
0.00
1
22
0.20
0.20
0.00
-0.50
0.00
-0.20
-1.00
-0.20
-0.40
-1.50
-0.40
-2.00
-0.60
-2.50
-0.80
1
4
Perc. pts
0.40
7
10
13
16
19
22
-0.60
-0.80
-1.00
1
7
13
16
19
Employed persons (BAEL) y/y
0.20
0.00
1
22
4
7
Perc. pts
0.40
0.30
0.20
10
Perc. pts
0.40
Nominal wages y/y
4
10
13
ULC y/y
0.20
0.00
0.10
-0.20
0.00
-0.20
-0.40
-0.10
-0.40
-0.60
-0.20
-0.60
-0.80
-0.30
-0.40
-0.80
1
4
Perc. pts
0.70
0.60
7
10
13
16
19
-1.00
1
22
Perc. pts
0.07
Unemployment
Unemploym
4
7
10
13
16
19
22
Public finance sector balance (% of GDP)
1
10.00
4
7
Perc. pts
10
13
Public debt (% of GDP)
0.02
0.50
-0.03
0.40
-0.08
0.30
-0.13
0.20
-0.18
0.10
-0.23
0.00
-0.28
-0.10
-0.33
-0.20
8.00
6
6.00
4.00
2.00
0.00
-0.38
-0.30
-2.00
-0.43
1
4
7
10
13
16
19
22
1
4
7
10
13
16
19
22
1
4
7
10
13
16
19
22
Source: NBP calculations.
Description: The monetary impulse is defined as a rise in the short-term interest rate by 1 percentage
point for a period of 8 quarters. The simulation has been conducted with an endogenous interest rate (with
Taylor’s Rule included) and with the fiscal policy rule included.
MATERIA¸Y I STUDIA PAPER 36
69
Annex
Fig. 11. Reaction of the economy to the exchange rate impulse
Perc. pts
0.25
Perc. pts
0.15
CPI y/y
Perc. pts
0.05
GDP y/y
Individual consumption y/y
0.04
0.20
0.10
0.03
0.15
0.02
0.05
0.10
0.01
0.05
0.00
0.00
0.00
-0.05
-0.01
-0.02
-0.05
-0.10
-0.03
-0.10
-0.04
-0.15
-0.15
-0.05
-0.20
-0.20
1
4
7
Perc. pts.
0.40
10
13
16
19
-0.06
1
22
Investments y/y
7
Perc. pts
0.30
0.30
4
10
13
16
19
22
1
Exports y/y
0.20
4
7
Perc. pts
0.10
10
13
16
19
22
16
19
22
16
19
22
16
19
22
19
22
Imports y/y
0.05
0.20
0.10
0.10
0.00
0.00
0.00
-0.10
-0.10
-0.05
-0.20
-0.10
-0.20
-0.30
-0.15
-0.30
-0.40
-0.50
-0.40
1
4
7
%
1.40
10
13
16
19
-0.20
1
22
Perc. pts
0.50
NEER
1.20
4
7
10
13
16
19
1
22
0.40
0.40
0.30
0.30
1.00
0.80
0.60
0.20
0.40
0.10
0.20
0.00
4
Perc. pts
0.50
Export deflator y/y
7
10
13
Import deflator y/y
0.20
0.10
0.00
-0.10
0.00
-0.10
-0.20
-0.40
-0.20
-0.60
-0.30
1
4
Perc. pts
0.25
7
10
13
16
19
22
-0.20
-0.30
-0.40
1
0.20
7
10
13
16
19
4
7
Perc. pts
0.40
Employed persons (BAEL) y/y
0.10
1
22
Perc. pts
0.15
Nominal wages y/y
4
10
13
ULC y/y
0.30
0.15
0.10
0.05
0.20
0.05
0.00
0.10
0.00
-0.05
-0.05
0.00
-0.10
-0.10
-0.10
-0.15
-0.15
-0.20
1
4
7
Perc. pts
0.20
10
13
16
19
22
-0.30
1
Unemployment
4
Perc. pts
0.12
0.10
0.15
6
-0.20
-0.20
-0.25
0.10
0.05
7
10
13
16
19
22
Public finance sector balance (% of GDP)
0.08
1.00
0.06
0.50
0.04
0.00
-0.10
0.00
-0.50
-0.02
-1.00
-0.04
-0.15
4
7
10
13
Public debt (% of GDP)
0.00
0.02
-0.05
1
Perc. pts
1.50
-1.50
-0.06
-0.20
-0.08
-0.25
-0.10
1
4
7
10
13
16
19
22
-2.00
-2.50
1
4
7
10
13
16
19
22
1
4
7
10
13
16
Source: NBP model.
Description: The exchange rate impulse is defined as a rise in the FX risk premium by 1 percentage point
for a period of 8 quarters. The simulation has been conducted with an endogenous interest rate (with
Taylor’s Rule included) and with the fiscal policy rule included.
70
N a t i o n a l
B a n k
o f
P o l a n d
Annex
Fig. 12. Reaction of the economy to the fiscal impulse
Perc. pts
0.40
Perc. pts
0.40
CPI y/y
0.30
0.20
0.10
Perc. pts
0.50
GDP y/y
0.30
0.40
0.20
0.30
0.10
0.20
0.00
0.10
Individual consumption y/y
0.00
-0.10
-0.10
0.00
-0.30
-0.20
-0.10
-0.40
-0.30
-0.20
-0.20
-0.50
-0.40
1
4
7
Perc. pts
0.80
10
13
16
19
22
-0.30
1
Investments y/y
7
Perc. pts
0.30
0.60
4
10
13
16
19
22
1
Exports y/y
0.25
0.40
0.20
0.30
0.40
0.10
0.00
0.05
7
10
13
16
19
22
16
19
22
16
19
22
16
19
22
Imports y/y
0.20
0.15
0.20
4
Perc. pts
0.50
0.10
0.00
-0.10
0.00
-0.20
-0.20
-0.05
-0.40
-0.30
-0.10
-0.60
-0.40
-0.15
1
4
7
%
1.00
10
13
16
19
22
-0.50
1
Perc. pts
0.60
NEER
4
7
10
13
16
19
1
22
0.50
0.40
0.40
0.00
0.20
0.20
-0.50
0.00
0.00
-1.00
-0.20
-0.20
-1.50
-0.40
-0.40
-2.00
-0.60
-0.60
-2.50
-0.80
-3.00
4
Perc. pts
0.50
7
10
13
16
19
22
0.40
10
13
-1.00
1
4
7
10
13
16
19
22
Perc. pts
0.40
Nominal wages y/y
7
Import deflator y/y
-0.80
-1.00
1
4
Perc. pts
0.60
Export deflator y/y
Employed persons (BAEL) y/y
0.30
1
4
7
Perc. pts
0.60
10
13
ULC y/y
0.40
0.30
0.20
0.20
0.10
0.10
0.00
0.00
-0.10
-0.10
0.20
0.00
-0.20
-0.20
-0.40
-0.20
-0.30
-0.60
-0.30
-0.40
-0.50
-0.80
-0.40
1
4
7
Perc. pts
0.60
10
13
16
19
4
Perc. pts
0.90
Unemployment
0.50
1
22
0.80
7
10
13
16
19
1
22
Public finance sector balance (% of GDP)
0.70
-1.00
0.30
0.60
-2.00
0.20
0.50
0.10
0.40
0.00
0.30
-4.00
-0.10
0.20
-5.00
-0.20
0.10
-0.30
0.00
-0.40
-0.10
4
7
10
13
16
19
22
7
10
13
Public debt (% of GDP)
0.40
1
4
Perc. pts
0.00
6
-3.00
-6.00
-7.00
1
4
7
10
13
16
19
22
1
4
7
10
13
16
19
22
Source: NBP models.
Description: The fiscal impulse is defined as a fall, for a period of 8 quarters, in a few categories of public
spending — current expenditures, pension spending, unemployment benefits and other social
expenditures - so that their total drop equals 1% of GDP. The simulation has been conducted with an
endogenous interest rate (with Taylor’s Rule included) and with the fiscal policy rule included.
MATERIA¸Y I STUDIA PAPER 36
71
Annex
Annex 2. “Model-experts” iteration method
In forecasting, there are two methods of compiling information from the model
and information from experts’ judgements: the iteration method and the weighted
method. In the process of forecast preparation, the method of model-experts iteration is
used. This refers to the variables in the case of which there are difficulties in estimating
a fully satisfying equation - a special case here is the investment equation.
An expert, having studied forecast results, may propose a correction of the
investment path. His or her information is used to make relevant changes to the constant
in the investment equation. The model is solved again, with the changed equation. The
results are presented to the expert and the correction procedure is repeated, until
obtaining the convergence of the expert’s judgement with the results generated by the
model49.
Thus the final forecast is a result of co-operation of the authors of the ECMOD
model and experts. The co-operation principles guarantee that the mechanism of
interdependencies within the model remain intact, and at the same time external
information is taken into account (the expert adjusts the constant values in the model
equations), the forecast retains its internal cohesion (the model is thus prevented from a
loss of cohesion between the model-generated and the expert-corrected variables), and
6
the comparability of various forecasts or simulations is preserved (constant values that
represent corrections of experts are used to determine all forecast variants).
49
In contrast to the iteration method, the weighted method forecast is generated by combining two
independent forecasts (model forecast and expert forecast) with appropriate weights.
72
N a t i o n a l
B a n k
o f
P o l a n d
Annex
Initial model solution
Expert’s input
Model solution – forecast
Do the model results
converge with experts’
expectations?
no
yes
Final model solution
(forecast)
6
MATERIA¸Y I STUDIA PAPER 36
73
Annex
Annex 3. Long-term equilibrium
The approach of the cointegration analysis used in the ECMOD model enables
us to distinguish
between
the
short-term
and
the
long-term
behaviour
of
macrocategories. In order to examine strictly long-term properties of the model, the
model has been analysed without considering in behavioural equations the part
responsible for the short-term dynamics.
The analysis should take into account the monetary policy rule and the fiscal
policy rule50, which help to achieve stabilisation in inflation, the output gap and the
general government balance in relation to GDP. These variables influence both the
nominal sector and the real sector of the economy. It is assumed that, in the long-term,
Poland’s economy will achieve complete convergence with the euro area economies. It
implies the assumption of equal dynamics of potential product in Poland and abroad as
well as the equal growth rate of the GDP deflator in Poland and abroad.
Let us analyse the nominal variables in the model. The assumed equal dynamics
of the GDP deflator in Poland and abroad, while domestic interest rates remain stable
(interest rates stabilise when the economy achieves long-run equilibrium), contribute
to the stabilisation of the disparity between Polish and foreign interest rates. A constant
6
disparity in interest rates, combined with the assumed equal dynamics of potential
products and GDP deflators in Poland and abroad, leads to the stabilisation of the real
exchange rate. The nominal exchange rate grows at a steady rate, determined by GDP
deflators. Exchange rate stability, combined with stable growth of the GDP deflator in
Poland, leads to a stable growth rate of import and export prices equal to the GDP
deflator growth rate. Due to changes in interest rates, triggered by the reaction of the
monetary policy rule, inflation stabilises. This is additionally supported by the
equilibrium (described below) achieved in the real economy. Stable growth of nominal
wages results from the stable growth of labour productivity and stable inflation growth.
Thus, long-term equilibrium in the nominal sector of the economy is achieved.
In the real economy, gross domestic product is the most important variable. The
long-run growth path of the gross domestic product in line with the adopted production
function results from the assumed dynamics of the total factor productivity (exogenous
variable). In the long-term, an assumption has been made with regard to the equalisation
of the marginal product of capital with the real cost of capital to the user51 (the relation
is included in the investment equation). We additionally assume that the supply of
50
51
74
The rules are described in sections 3.1 and 3.2.
The relation complies with the theory describing fixed asset formation in the neoclassical model.
N a t i o n a l
B a n k
o f
P o l a n d
Annex
labour is determined by the labour force52 adjusted for the NAWRU. Thus, the potential
product’s growth is determined and GDP growth adopts itself to the potential products
dynamics.
The steady state is achieved partly due to the monetary policy rule, which —
through adjustments in the output gap — leads to changes in the levels of consumption
and investments, until reaching a constant growth rate. In the general government
module automatic stabilisers exert an anti-cyclical influence, thus helping to achieve
steady growth. At the same time, the fiscal rule ensures the stabilisation of the relation
of the general government balance to GDP and — by changes in the costs of debt
service —the stability of public debt.
Stable growth of import and export prices as well as assumptions concerning the
GDP deflator and the determined growth of the potential product determine a constant
growth rate for exports and imports that is equal to the growth of the potential and thus
the growth of the Gross Domestic Product. This entails stabilisation of the share of net
exports in GDP (one of the components of the domestic product). As a result, the
growth of the current account equals GDP growth. This, in turn, leads to stable growth
of net foreign assets.
Equilibrium of the general government (presented above) results in the same
6
growth of both public revenues and expenditures as GDP growth (by maintaining a
stable balance to GDP ratio). This has a direct impact on determining the growth of
collective consumption at the level of public expenditure growth, i.e. at the level of
growth in Gross Domestic Product. This means a constant share of collective
consumption (another GDP component) in GDP.
Stabilisation of capital growth (at the GDP growth rate), public debt (at the GDP
growth rate) and the balance of foreign assets (at the GDP growth rate) leads to a
constant wealth growth rate (i.e., a constant share of wealth in GDP). Stable growth in
disposable income (equal to the GDP growth rate, since it is determined by growth in
public revenues and expenditures and growth in real wages), combined with stable
growth in wealth, leads to steady increases in individual consumption, equal to the GDP
growth rate. Thus, we have shown that the dynamics of all GDP components, i.e. net
exports, collective consumption, individual consumption and investments (through the
capital relationship described above) stabilise at the level of GDP dynamics. This also
entails the stability of individual components in gross domestic product.
52
We assume that the labour force remains stable in the long-term.
MATERIA¸Y I STUDIA PAPER 36
75
Annex
Annex 4. Cointegration analysis
Equations in the paper are presented in the form of error correction models
(small letters indicate variables in logarithms). Every equation enables us to read both
long-term relations and short-term dynamics:
(i) long-term relation corresponds to the relation in parentheses concerning of
variable levels,
(ii) short-term dynamics are illustrated with other variables (these are usually
variables in first differences, since the variables are non-stationary in most cases).
This annex presents analysis of cointegration and the essence of the error
correction models. The discussion is based on R. Engle and C. Granger (1987),
Cointegration and Error Correction: Representation, Estimation, and Testing,
Econometrica, vol. 55, pp. 251-276.
Process {xt} is wide-sense stationary if three conditions are met:
6
E ( xt ) = ȝ for t=1,2,...
(V.1a)
Var ( xt ) = σ 2 for t=1,2,...
(V.1b)
Cov( xt , xt − k ) = σ k for t=1,2,... and k=1,2,...
(V.1c)
The first two conditions mean that the expected value and the variance of the
process {xt} are constant over time. Condition (1c) indicates that the co-variance
between two variables xt and xt-k is determined only by the distance k, and not by the
choice of the time t. The stationary process is said to be integrated of order 0, noted
xt~I(0). In the event that processs {xt} does not meet the conditions (V.1a)-(V.1c), and
the series in the first differences {∆xt}, where ∆xt = ∆(1) xt = xt − xt −1 , is stationary,
process {xt} is said to be non-stationary, integrated of order 1, which is recorded as
xt~I(1).
Definition of cointegration
Two one-dimensional processes xt and yt are cointegrated of order d, b,
which is recorded as (xt, yt)~CI(d,b), if two conditions are met. First, each process
is integrated of order d, i.e. xt ~ I(d) and yt ~ I(d). Second, there exists a cointegrating
vector [α 1 α 2 ] ≠ [0 0] so that the process {ut}, where
T
T
u t = α 1 xt + α 2 y t
76
(V.2)
N a t i o n a l
B a n k
o f
P o l a n d
Annex
is integrated of order d-b, i.e. ut ~ I(d-b), b>0. Since a large part of economic
time series is integrated of order 1, the econometric model for b=d=1 is particularly
interesting. In this case, the existence of a cointegrating relation implies that, for the
system of the analysed variables, there is a precisely specified long-term equilibrium,
derived from the cointegrating relation (i.e. when ut = 0 ). When there is a disturbance
leading to deviation from the equilibrium (i.e. when ut≠0), the system automatically
returns to the level determined by the cointegrating relation (hence, the cointegrating
relation is also called a long-term relationship).
The above mentioned automatic return of the system to the level determined by
the long-term relationship was recorded in mathematical form by Engle and Granger
(1987), who proposed the following model53:
ª yt −1 º ªε 1t º
ª∆yt −k º ªδ 1 º
ª∆yt º K
«∆x » = ¦ A k « ∆x » + «δ »[α 1 α 2 ]« x » + «ε » ,
¬ t −1 ¼ ¬ 2t ¼
¬ t ¼ k =1 ¬ t − k ¼ ¬ 2 ¼
(V.3)
where [ε 1t
ε 2t ]T is a two-dimensional random component with variance Ȉ , and the
K
moduli of eigenvalues of matrix A = ¦ A k are less than one. In accordance with the
k =1
(V.3) specification, the dynamics of the i- variable (i=1,2) of vector
on past vector changes [xt
[ xt
6
yt ] depend
T
yt ] and on the product of the adjustment parameter δ i and
T
system deviation from the level given in the long-term (cointegrating) relationship in
time t-1. For į = [δ 1 δ 2 ] ≠ 0 the deviation is adjusted through the impact on current
T
changes in the vector
[ xt
yt ] . This phenomenon has been named in the econometric
T
literature as the error correction mechanism. The relation given in equation (V.3)
describes current dynamics of the vector
[ xt
yt ] , and therefore it is frequently called
T
a short-term relationship.
Annex 5. Definitions of variables
Endogenous variables
AVGBENEFIT_N — average quarterly amount of benefit from the Labour Fund
per unemployed person (equation E. 55),
AVGPENSION_N — average amount of retirement and disabled pension benefits
(equation E. 59),
53
A case of two variables.
MATERIA¸Y I STUDIA PAPER 36
77
Annex
AVGRELIEF_N — average quarterly amount of social benefits (equation E. 57),
CAB — current account balance in zloty (equation E. 24),
CAB_INC — balance of revenues (equation E. 28),
CAB_NT — balance of goods and services (equation E. 25),
CAB_TRANS — balance of transfers,
CONP — individual consumption (equation E. 7),
CONP_N — nominal individual consumption,
CPI — consumer price level in period t (E. 13),
CPIe — expected inflation,
EMP — number of employees in thousands, by BAEL (equation E. 4),
FINACC — share of enterprises’ disposable income54 in national disposable income,
FINV — gross fixed capital formation (equation E. 11)
GAP — output gap (equation E. 6)
GCE_N — current government spending, comprising purchases of non-investment
goods (about 30% of total expenditure) and wagefund (about 70% of total expenditure)
(equation E. 52),
6
GCIT_N — corporate income tax revenues (equation E. 42),
GCT_N — real estate tax revenues (equation E. 49),
GDEBT_N — nominal volume of public debt (equation E. 65),
GDEF_N — general government balance (equation E. 64),
GDP — Gross Domestic Product (equation E. 1),
GDP_EXP — exports of goods and services in constant prices, according to the national
accounts (equation E. 31),
GDP_EXP_N — nominal value of exports of goods and services, according to the
national accounts (equation E. 26),
GDP_IMP — volume of imports of goods and services, according to the national
accounts (equation E. 32),
GDP_IMP_N — nominal value of imports of goods and services, according to the
national accounts (equation E. 27),
GDP_N — nominal GDP value,
GDP_POT — potential GDP (equation E. 5)
54
By the term “enterprises” we understand two institutional sectors, distinguished in the national
accounts: the enterprise sector and the sector of financial institutions .
78
N a t i o n a l
B a n k
o f
P o l a n d
Annex
GDS_N — costs of debt service (equation E. 60),
GEXP_N — general government expenditures (equation E. 51),
GEXT_N — excise tax revenues (equation E. 39),
GGAM_N — gaming tax revenues (equation E. 40),
GIDT_N — indirect tax revenues (equation E. 37),
GINC_N — general government revenues (equation E. 36),
GINV_N — public investment (equation E. 61),
GLT_N — revenues from health and social insurance contributions (equation E. 43),
GLT_CORP_N — social insurance contributions paid by employers (equation E. 45),
GLT_EMP_N — social insurance contributions paid by employees (equation E. 44),
GLT_HC_N — health insurance contributions (equation E. 47),
GLT_SB_N — social insurance contributions paid by business owners and freelance
professionals (equation E. 46),
GPIT_N — personal income tax revenues (equation E. 41),
GRE_N — other expenditures (equation E. 63),
GRT_N — other revenues (equation E. 50),
6
GTAR_N — customs revenues (equation E. 48),
GTR_CORP_N — subsidies to enterprises (equation E. 62),
GTR_N — transfers to the population (social benefits, unemployment benefits,
pensions) (equation E. 53),
GTR_RELIEF_N — transfers from social welfare (equation E. 56),
GTR_RETIRED_N — pension transfers (equation E. 58),
GTR_UNEMP_N — unemployment benefits (equation E. 54),
GVAT_N — value-added tax revenues (equation E. 38),
I_RENTP — average quarterly value of the average weighted interest rate on loans
to enterprises from commercial banks (E. 23),
I_W3M — quarterly nominal average value of WIBOR3M (equation E. 66),
I_W3MR_CPI — quarterly real average value of WIBOR3M deflated by the CPI,
I_W3MR_PGDP — GDP-deflated real WIBOR 3M,
I3MR_EXT — GDP-deflated real foreign 3M rate (equation E. 19),
INF — CPI inflation (equation E. 13),
MATERIA¸Y I STUDIA PAPER 36
79
Annex
INV — gross capital formation
K — average capital KEP in the period (equation E. 3),
KEP — capital stock at the end of the period, equel to cumulated investment value
(equation E. 2),
NETCPI — index of the core inflation price level, net of food and fuel prices (E. 12
NFA — net foreign assets (equation E. 29),
P_EXP — export deflator according to the national accounts (equation E. 33),
P_IMP — import deflator according to the national accounts (equation E. 35),
P_IMP* — import prices adjusted for customs duties and import taxes (in effect until
the mid-90s): P_IMP*=P_IMP*(1+podatek imp.*cło),
P_IMP_NOIL — import deflator according to the national accounts, adjusted for oil
price movements (equation E. 34),
PGDP — GDP deflator (equation E. 15),
RUCC — real cost of capital (interest rates on loans to enterprises, GDP-deflated,
adjusted for capital depreciation and corporate income taxes),
S_EUR_PLN — nominal PLN/EUR exchange rate — growth denotes zloty
depreciation (equation E. 22),
S_NEER — nominal effective PLN exchange rate — growth denotes zloty depreciation
(equation E. 20),
6
S_REER — real effective PLN exchange rate (weights Tab. 2) — growth denotes zloty
depreciation (equation E. 18),
S_USD_PLN — nominal PLN/USD exchange rate (equation E. 21),
ULC — unit labour costs,
UNEMP — BAEL unemployment rate,
WAGE — average monthly wage in nominal terms (adjusted for the transition to gross
wages in 1999 — equation E. 16),
WAGEFUND_N — nominal quarterly wage fund in the economy,
WEALTH — wealth,
YD — disposable income,
YD_N — nominal gross disposable income (equation E. 8),
YD_NO — nominal operational surplus of the households’ sector (equation E. 9),
YD_WL — nominal property income in households’ sector (equation E. 10),
YD_POZOSTALE — other disposable income.
80
N a t i o n a l
B a n k
o f
P o l a n d
Annex
Exogenous variables
CGNBP_N — NBP profit transfer,
CONGOV — collective consumption,
DEPR — quarterly depreciation rate,
dl — dummy variable, quantifying the impact of the growth rate of the “net” core
inflation price index higher than the GDP deflator price index,
FOOD_CPI — food and alcohol price level,
FUEL_CPI — fuel prices for private means of transport,
GCIT_TR — effective CIT rate,
GCT_TR — effective real estate tax rate,
GDEBT_EUR — share of euro-denominated debt in total debt,
GDEBT_IR — effective interest rate of public debt,
GDEBT_PL — share of domestic debt in public debt,
GDP_EXT — weighted GDP level for Poland’s major trading partners (weights Tab.
2),
GDP_EXT_POT — weighted potential GDP abroad (weights Tab. 2),
6
GEXP_UE_N — contribution to the EU budget,
GEXT_TR — effective excise tax rate,
GGAM_TR — effective gaming tax rate,
GINC_UE_N — special cash-flow facility from the EU budget,
GLT_CORP_TR — effective rate of revenues from social contributions paid by
employers,
GLT_EMP_TR — effective rate of revenues from social contributions paid by
employees,
GLT_HC_TR — effective rate of revenues from health insurance contributions,
GLT_SB_TR — effective rate of revenues from social contributions paid by business
owners and freelance professionals,
GPIT_TR — effective PIT rate,
GPRIV_N — revenues from privatisation (adjusted for compensation from the budget),
GTAR_TR — customs duty effective rate,
GVAT_TR — VAT effective rate,
I3M_EXT — weighted nominal M interest rate abroad (weights in Tab. 2),
MATERIA¸Y I STUDIA PAPER 36
81
Annex
INF_CEL — inflation target (i.e. 2,5%),
LF — labour force in thousands, according to BAEL,
NAWRU — Non-Accelerating Wage Rate of Unemployment
OIL — world oil prices in US dollars,
PGDP_EXT — weighted (weights in Tab. 2) GDP deflator abroad,
RETIRED — number of pensioners,
S_RISK_PR — valuation of risk premium,
S_USD_CZK — nominal CZK/USD exchange rate,
S_USD_EUR — nominal EUR/USD exchange rate — growth denotes depreciation of
the euro,
S_USD_GBP — nominal GBP/USD exchange rate,
S_USD_SEK — nominal SEK/USD exchange rate,
TFP — total factor productivity,
TR_WPR_CON — revenues obtained within the Common Agricultural Policy,
allocated for consumption.
6
82
N a t i o n a l
B a n k
o f
P o l a n d
Keep reading this paper — and 50 million others — with a free Academia account
Used by leading Academics
Selliah Sivarajasingham
University of Peradeniya
zerife yıldırım
Harran University
E.i.abdul Sathar
University of Kerala
Srinivasa Rao Gadde
The University of Dodoma, Tanzania
