Banks and Bank Systems, Volume 5, Issue 2,
2010
Cândida
Ferreira (Portugal)
The credit channel transmission of monetary
policy in the European Union: a panel data approach
Abstract
This
paper seeks to contribute to the analysis of the financial integration, the
importance of bank performance condi-tions and the bank lending channel
transmission of monetary policy in the European Union countries since 1999.
Using pooled panel OLS estimations and dynamic Arellano-Bond GMM estimations
with quarterly data for 26 EU countries for the period from Q1 1999 to Q3 2006
it confirms the high degree of integration between the EU financial systems, as
well as the importance of bank performance conditions to the credit-lending
channel of monetary policy in the EU. Furthermore, it demonstrates not only the
quite high degree of openness of the financial markets but also their
indebt-edness and the dependence of the EU banking institutions on the
financial resources of other countries.
Keywords:
European
integration, bank credit, monetary policy transmission, panel estimates.
JEL
Classification: E4, E5, G2.
Introduction
©
The
introduction of the single currency has acceler-ated the process of
consolidation and financial inte-gration, not only in the Economic and Monetary
Union (EMU), but in the whole European Union (EU), in which the new member
states also have a voice, in spite of the possible heterogeneous nature of
their financial systems.
The
process of financial integration is, on the one hand, a necessary pre-requisite
for the adoption of the single currency and the implementation of the single
monetary policy, with the predominance of the banking intermediation in the
context of the EU. On the other hand, this process raises the potential to
incite liquidity crises, which could become conta-gious and affect the
increasingly integrated Euro-pean financial system.
More
efficient credit sectors should contribute to the economic benefits of the
other sectors and agents which use financial services and they also represent a
necessary condition for the transmission mecha-nism of monetary policy.
According
to the credit and lending view, the effec-tiveness of monetary policy depends
basically on the banking system, since imperfections, such as asym-metric
information and the subsequent phenomena of adverse selection and moral hazard,
exist in the capi-tal markets, which increase the particular difficulties felt
by some economic agents to finance their invest-ment and consumption plans.
Under these conditions, central banks control the supply of money, but the
banking institutions also play an important role in the money-creation process,
as well as in the mobiliza-tion and allocation of financial resources.
In
addition, more efficient banking sectors are gen-erally recognized as a
necessary condition for the transmission mechanism of monetary policy and the
way that banks adapt lending in response to mone-tary policy decisions varies
according to their spe-cific political and economic environment.
However,
there is no agreement on the precise specification of the ways in which
monetary policy influences the economy. Hence, it is an area merit-ing further
investigation (Goddart et al., 2007).
Following
these vectors of research, this paper seeks to contribute to the analysis of
the financial integra-tion, the importance of bank performance conditions and
the bank lending channel transmission of mone-tary policy in the EU countries
since 1999.
The main contributions are to be found
in:
1.
The use of quarterly
data, between Q1 1999 and Q3 2006, for 26 EU1
countries (the only excep-tion is Luxembourg, for which it was not possi-ble to
obtain all the data). This is in contrast with most of the empirical studies in
this area, which analyze only sub-sets of EU countries – all of the EMU, or
some of its more significant members, or some new EU member states – to test
the importance of the credit channel trans-mission of monetary policy;
2.
The adaptation of the
Bernanke and Blinder (1988) model with the introduction of four ra-tios to
represent the bank-performance condi-tions: bank deposits/GDP; bonds and money
market instruments/GDP; foreign assets/GDP; and foreign assets/foreign
liabilities;
3.
The use of panel data
estimations – pooled panel OLS estimations and dynamic Arellano-Bond
©
Cândida Ferreira, 2010.
I
kindly thank the pertinent comments and suggestions of the editors of the
“Banks and Bank Systems” International Research Journal. They were all taken
into account and several sentences and particularly footnotes were inserted in
this version of the paper. However, the usual disclaimer applies.
1 More
precisely, we use the data for Austria, Belgium, Bulgaria, Cy-prus, Czech
Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland,
Italy, Latvia, Lithuania, Malta, Netherlands, Poland, Portugal, Romania, Slovak
Republic, Slovenia, Spain, Sweden and United Kingdom.
Generalized
Method of Moments (GMM) esti-mations – not only to confirm the importance of
the bank lending channel, but also to draw con-clusions on the level of
financial integration of the EU countries.
The
remainder of the paper is organized as follows: Section 1 presents the
contextual setting and the relevant literature; the methodological framework
and the data are presented in Section 2; Section 3 displays the results
obtained; finally in the last Sec-tion, we make our concluding remarks.
1.
Contextual setting and literature
In
recent years and particularly during the last dec-ade, the banking activity has
had to adapt to profound transformations, due to advances in information and
financial technologies and changes in institutional and regulatory conditions,
together with shocks from the socio-economic and financial environment.
In
the EU, the structural changes arising first from the adoption of the single
currency and a common monetary policy and then from the recent histori-cally
remarkable enlargement, which brought the entry of ten countries at the same
time, followed shortly after by two more countries, have had a pro-found
impact, not only in the Euro area but also throughout the entire EU-27, where
the financial sector has experienced an intensification of competi-tion in
banking services.
Some
authors have already analyzed the degrees of integration through the common
trends which may be identified in the context of the pressures of
glob-alization and which affect all the EU countries (not only the EMU members)
with particular intensity, due to the process of disintermediation, new
tech-nologies and increased competition (Belaisch et al., 2001; Gardener et
al., 2002; Melnik and Nissim, 2006).
The
increasingly competitive environment of the EU banking sector and the process
of concentration as well as the decline in the number of banks in almost all EU
countries, did not eliminate much of the ex-cess capacity in the system.
Moreover, there is evi-dence that large banks continue to have efficiency
advantages over the smaller banks (Altunbas et al., 1997; Cabral et al., 2002;
Casu and Molyneux, 2000; Jansen and de Haan, 2003; Molyneux, 2003; Baele et
al., 2004; Romero-Ávila, 2003 and 2007).
In
Barros et al. (2007), the efficiency of almost 1400 commercial banks operating
in the EU be-tween 1993 and 2001 was analyzed. The study confirmed the
importance of country-level charac-teristics and firm-level features to explain
the probability of a bank being a best (worst) per-former. In particular, we
concluded that smaller-
Banks
and Bank Systems, Volume 5, Issue 2, 2010
sized banks with higher loan intensity
and foreign banks from countries upholding common law traditions have a higher
probability of best per-formance.
It is
generally recognized that nowadays special at-tention must be paid to the EU
banking sector follow-ing the most recent enlargements mentioned above, particularly
regarding those countries formerly under the Soviet Union sphere of influence,
given that in a quite short period of time, the banks in these coun-tries moved
from the structure of socialist banking, in which the financial organizations
were used to sup-port the central banking system, to a market economy and the
concomitant decentralization and liberalization of the banking systems.
In
most of these Eastern and Central European coun-tries, forms and programs were
introduced to amend property rights, together with processes of privatiza-tions
of part of the State property. As a result, the importance of the private
sector and firms increased in these countries, as did the particularly relevant
role of their financial intermediaries and banking institu-tions. There is a
fairly strong consensus on the in-creased performance and efficiency of the
banks under the new market conditions in these countries. Several studies
(Holscher, 2000; Winkler, 2002; Backhaus, 2003; Sztyber, 2003; Hanousek and Kocenda,
2003; Stephen and Backhaus, 2003; Tchipev, 2003; Dimi-trova, 2004; Bonin and
Watchel, 2004; Bonin et al., 2005-a, 2005-b; Freis and Taci, 2005; Fries et
al., 2006) confirm the relevant improvements in efficiency of the banking
systems of the new EU members and the effects of ownership, concluding that
foreign-owned banks are usually more cost-efficient.
Other
studies examine how, and to what extent, the banking sectors of the new
member-states have in-tegrated with those of the older EU members and the
process of nominal and real convergence of these countries to EU standards
(ECB, 2004 and 2005; Kocenda et al., 2006).
The
transmission of monetary policy to the non-monetary economic sectors also
requires more effi-cient banking and the way that banks adapt lending in
response to monetary policy decisions varies accord-ing to their specific
political and economic environ-ment. However, in spite of all the theoretical
and empirical advances in this area, there is still no agreement about the
precise specification of the ways in which monetary policy influences the
economy. Thus, it is acknowledged as an area meriting further investigation
(Goddart et al., 2007).
Some
contributions to the explanation of the classic interest-rate channel
transmission of monetary policy (Taylor, 1995; Cecchetti, 1995; Bean et al.,
2002) imply that the influence of interest rates on economic
Banks and Bank Systems, Volume 5, Issue 2,
2010
activity
affects, at least, the components of domestic demand. Nowadays, the traditional
interest-rate channel is not the only transmission mechanism of monetary
policy. Mishkin (1995, 2001) adds an asset-price channel and an exchange-rate
channel, sum-ming up the new different mechanisms as “other asset prices” and
the “credit view”.
This
credit channel may be seen as the development and extension of the conventional
interest-rate effect (also developed by Bernanke and Getler, 1995, as well as
Hubbard, 1995), taking into account the rising evaluation and monitoring costs
for lenders, due to the information asymmetries in credit mar-kets which
provoke adverse selection and moral hazard effects.
According
to this credit view, monetary policy deci-sions will affect not only the credit
demand side, through the balance sheet channel, but also the sup-ply side,
through the bank lending channel. More precisely, for instance, the tightening
of monetary policy, through the balance sheet channel will make external
finance more costly for borrowers with the increase of their interest expenses
and the reduction of their collateral while, through the bank lending channel,
the reduction of the banks’ liquidity will force banking institutions to reduce
lending.
However,
such a reduction also reflects the banks’ characteristics and the environment
in which banks are operating. Lending by smaller and relatively
under-capitalized or illiquid banks is usually more sensitive to interest rate
movements (Kashyap and Stein, 1997, 2000; Kishan and Opiela, 2006).
This paper follows the vectors of
research that adapt and develop the pioneer Bernanke and Blinder (1988) model,
and particularly:
1.
The empirical papers
that recently have tested the existence of a bank lending channel for the
transmission of monetary policies in the Euro zone, obtaining rather similar
conclusions on the relative homogeneity of the behavior of the EU banking
institutions (Erhmann et al., 2001; Fountas and Papagapitos, 2001; Topi and
Vil-munen, 2001; Van Els et al., 2001; Worms, 2001; Altunbas et al., 2002;
Angeloni et al., 2002; Gambacorta, 2004; Gambacorta and Mis-trulli, 2004;
Ferreira, 2007).
2.
The other
contributions that analyze the trans-mission channels of monetary policy in
different EU countries, including the new member-states in Central and Eastern
Europe (Golinelli and Rovelli, 2005; Elbourne and de Haan, 2006; Ferreira,
2008).
2.
Methodological framework and used data
2.1.
The model. The used model is an adaptation of the
Bernanke and Blinder (1988) model.
In
the money market, we will assume that money equals deposits held at banks by
the non-monetary sectors. So, for the demand function, we consider that the
nominal deposits held in banks by the pri-vate sector will depend positively on
the GDP and negatively on the interest rate on bonds:
Depd a
|
0
|
_
a GDP _ a
|
i
|
bonds
|
,
|
(1)
|
|
|
1
|
2
|
|
|
|
where Depd = deposits, d meaning demand; GDP
= Gross Domestic Product; ibonds = interest rate on bonds; a1 > 0; a2 < 0.
On
the other side, money supply will depend not only on the interest rate on
bonds, but also on the influence of monetary policy (represented here by the
relevant monetary policy interest rate, which is defined by the Central Bank):
Deps b
|
_b
i
|
bonds
|
_b
i
|
mon.pol.
|
,
|
(2)
|
|
0
|
1
|
2
|
|
|
|
where Deps = deposits, s meaning supply; ibonds = interest rate on bonds; imon.pol. = monetary policy interest rate; b1 > 0; b2 < 0.
At
the same time, in the credit market, the demand for lending depends positively
on the GDP, nega-tively on the interest rate on lending/borrowing and
positively on the interest rate on bonds:
Lendd
c
|
_
c GDP _
|
c
|
i
|
_
c
|
i , (3)
|
0
|
1
|
2
|
lend
|
3
|
bonds
|
where Lendd = bank lending, d meaning
demand; GDP = Gross Domestic Product; ilend = interest rate on lending; ibonds = interest rate on bonds; c1 > 0; c2 < 0; c3 > 0.
Assuming
the relevance of one or more bank-performance characteristics (Charx)
which may exert either positive or negative influences on lending, we define
the supply in the money market as depending on the deposits of the private
sectors in banks, as well as on the bank characteristics, the interest rate on
lending/borrowing and the interest rate on bonds:
Lends d
|
0
|
_d Dep_d
Car _d i
|
_d i
,
(4)
|
|
|
1
|
2
|
x
|
3 lend
|
4 bonds
|
|
|
|
|
|
|
|
|
|
with Lends =
lending, s meaning supply; Dep = bank deposits of the private
sector; Carx = bank
character-istics (x = 1,..X); ilend =
interest rate on lending; ibonds =
interest rate on bonds; d1 >
0; d2 may be > 0 or <
0 so d2 = ?; d3 >
0; d4 < 0.
So, clearing the
money market – equations (1) and
(2) – we obtain:
Clearing the credit
market – equations (3)
and (4) – we first obtain the expression of the interest rate on lending:
|
Banks
and Bank Systems, Volume 5, Issue 2, 2010
i
|
|
b0 _a0
|
|
_
|
|
a1
|
GDP _
|
|
b2
|
i
|
|
|
Lend
|
_h0 _h2e0 _h3 f0 _ _
_h1 _h2e1 _h3 f1 _GDP_
|
|
|
a2
|
_b1
|
|
a2
|
_b1
|
|
a2 _b1
|
|
|
__h
e
|
|
bonds
|
|
|
|
|
|
|
|
|
mon. pol
|
_h f
|
2
|
_i
|
_h Car
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2 2
|
3
|
|
mon.
pol
|
4
|
x
|
|
or
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
or
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
ibonds
|
e0
|
_e1GDP _e2imon. pol ,
|
|
|
|
|
(5)
|
Lend
|
D0
|
_D1GDP _D2imon. pol _D3Carx , (9)
|
|
with e1 > 0; e2 > 0.
|
|
|
|
|
|
|
|
|
|
|
|
where Lend = bank lending; GDP
= Gross Domestic
|
|
At the same time, if money demand equals money
|
Product; imon.pol. = monetary
policy interest rate; Carx
|
|
supply:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
= bank
characteristics (x = 1,..X); D1 > 0 if h2 > 0;
|
|
d
|
|
|
s
|
a2b0 _a0b1
|
|
a1b1
|
|
|
|
a2b2
|
otherwise D1 may be < 0 ;
so D1 =
?; D2 > 0 if h2 > 0
|
|
Dep
|
Dep
|
|
|
|
_
|
|
|
GDP_
|
|
imon.pol
|
and h2 e2 > h3 f2 ; otherwise D2 may be < 0 ;
so D2 =
?
|
|
|
|
a2 _b1
|
|
a2 _b1
|
a2 _b1
|
|
or
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
; D3 may be > 0 or < 0 so D3 = ?.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2.2. The data. To
build our panel, we use Eurostat
|
|
Dep
|
|
f0
|
_ f1GDP _ f2imon. pol ,
|
|
|
|
|
(6)
|
|
|
|
|
|
|
and International
Financial Statistics (IFS) quarterly
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
data for the
period from Q1 1999 to Q3 2006 (31
|
|
quarters) and 26 EU countries, amounting to
806 observations. As mentioned previously, Luxembourg has been excluded, as it
was not possible to collect all of the necessary data for this country.
i
|
lends
|
d0 _c0
|
|
_
|
d1
|
|
Dep _
|
d2
|
|
Car _
|
|
|
|
|
|
|
|
|
|
|
c2
|
_d3
|
|
|
|
c2 _d
|
|
|
c2 _d
|
|
x
|
|
|
|
|
|
|
|
3
|
|
3
|
|
|
_
|
d4 _c3
|
i
|
_
|
|
c1
|
GDP
|
|
|
|
|
c2 _d3
|
c2 _d3
|
|
|
|
|
|
|
bond
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
or
ilend g0 _g1Dep_g2Carx _g3ibond _g4GDP, (7)
with: g1 < 0; g2 may be > 0 or < 0 so g2 = ?; g3 > 0; g4 > 0.
Using this definition of the
interest rate on lending, and admitting the credit market equilibrium, we get:
Lend
d Lend s
|
|
c2d0 _c0d3
|
_
|
|
c1d3
|
GDP _
|
|
|
|
|
|
|
|
|
|
|
|
|
|
c2 _d3
|
c2 _d3
|
|
|
_
|
c2d4 _c3d3
|
i
|
_
|
c2d1
|
Dep _
|
c2d2
|
Car
|
|
c2 _d3
|
|
c2 _d3
|
|
|
bond
|
|
|
c2 _d3
|
|
x
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
or
Lend h0 _h1GDP_h2ibond _h3Dep_h4Carx , (8)
Now: h1 >
0; h2 > 0 if c2 d4 <
c3 d3 or h2<
0 if c2 d4 >
c3 d3 ; so
h2 = ?;
h3 >
0; h4 may be > 0 or < 0 so
h4 = ?.
Remembering
the expressions of the interest rate on bonds and deposits – equations (5) and
(6)
ibonds
|
e0
|
_ e1GDP _ e2imon. pol ,
|
(5)
|
Dep
|
f0
|
_ f1GDP _ f2imon. pol ,
|
(6)
|
and
introducing these expressions into the equation (8), we obtain the reduced form
of the expression for lending, which is the basis of our estimations:
For
the dependent variable (bank lending) we use the natural logarithm of the ratio
of the domestic credit provided by the banking institutions to GDP. To explain
the growth of this bank lending, we will consider (always in natural
logarithms):
i the real GDP per
capita, representing the mac-roeconomic conditions of the different EU
countries;
i the discount rate
(end of the period) which is the monetary policy interest rate;
i the four ratios
which represent the bank per-formance conditions, more precisely:
- the
ratio of deposits to GDP, that is, the total deposits in the banking
institutions which are important sources of resources for credit lend-ing. For
instance, according to the macroeco-nomic money multiplier mechanism, bank
lending will mainly depend on the collected deposits and the legal minimum
reserves;
-
the ratio of the bonds
and money market in-struments to GDP, as a proxy of the develop-ment of the
financial markets in these coun-tries, which are mostly bank-dominated. Since
healthy financial markets and developed finan-cial institutions are a guarantee
for the direct and indirect financing of the bank clients’ ac-tivities, we may
expect that this ratio will exert a positive influence on bank lending;
- the
ratio of foreign assets to GDP, introducing the influence of the other
countries, more specifically, the financial resources obtained from foreign
partners, represented by the en-try of assets, in particular to pay their debts
and financial obligations, and consequently, more resources to be applied in
the domestic bank lending;
Banks and Bank Systems, Volume 5, Issue 2,
2010
- the
ratio of foreign assets to foreign liabilities, representing the financial
situation of the bank-ing institutions towards other countries, as they may
receive payments from foreign debtors. On the other hand, they also have
financial ob-ligations towards foreign creditors, which im-plies the payment of
debts and obligations to other countries. Therefore, the influence of this
ratio on bank lending will reveal not only the openness of the financial
markets, but mainly the degree of dependence on the other coun-tries’ financial
resources.
In Appendix A, we present the summary
statistics of these series, while the matrix of the correlations is presented
in Appendix B.
2.3.
Unit root tests. The collected data for 26 EU countries
for a time period of 31 quarters (806 obser-vations in total) does not lend
itself to the application of single time series unit root tests. Therefore, we
opt to use panel unit root tests, which are more adequate in this case (see,
among others, Karlsson and Loth-gren, 2000; Wooldridge, 2002; Basile et al.,
2005). These tests not only increase the power of unit root tests due to the
span of the observations, but also minimize the risks of structural breaks due to
possible changes in policy regimes.
Among
the available panel unit root tests, we choose the Levin, Lin and Chu (2002)
test, which may be viewed as a pooled Dickey-Fuller test or as an aug-mented
Dickey-Fuller test when lags are included, and the null hypothesis is the
existence of non-stationarity. This test is adequate for heterogeneous panels
of moderate size, as is the present case, and it assumes that there is a common
unit root process.
According
to the results obtained with the determi-nistic constant and trend up to 3 lags
(see Appendix C), the existence of the null hypothesis may be re-jected for all
the variables, mostly with no lags, except for the monetary policy interest
rate when lags are equal to one or two, while for the ratio of bonds and money
market instruments to GDP the best results are obtained with three lags.
3.
Empirical estimations
Using
the reduced form (equation (9)) of the pre-sented model, and the series
described above, we will explain the response of bank lending to relevant
macroeconomic conditions, as well as to some spe-cific characteristics of the
banking institutions and indicators representing their performance condi-tions,
by the estimation of the following equation (all variables in natural
logarithms1):
(Bank Lending/GDP) it =_M_0 + M 1 real GDP per cap.it + M 2 Interest rate it + M 3 (Deposits/GDP) it +_M 4 (Bonds and Money
Market Instruments/GDP) it + M 5 (Foreign Assets/GDP)
it + M 6 (Foreign
As-sets/Foreign Liabilities) it + Ki + Qt + uit,
where
i = 1,..., 26 (EU countries); t = 1,..., 31 (quarters, between Q1
1999 and Q3 2006); Ki =
country dum-mies; Qt =
time (quarter) dummies; uit =
error term.
Therefore,
with a panel of 806 observations, we will use a panel data approach which not
only provides more observations for estimations, but also reduces the
possibility of multi-collinearity among the dif-ferent variables.
To check for the robustness of the results
and the relative importance of the macroeconomic, monetary policy and bank
performance conditions for the ex-planation of the bank lending growth, we will
present the results of three equations: the first including all the explaining
variables; the second excluding the real GDP per capita but including all the
other five explaining variables (monetary policy interest rate and the four
ratios representing bank performance conditions); and the last equation
explaining the bank lending growth only by the bank performance con-ditions. In
our model these bank performance condi-tions are represented by: the deposits /
GDP ratio, the bonds and money market instruments / GDP ratio; the foreign
assets / GDP ratio and the foreign assets / foreign liabilities ratio.
For the estimations, we will use:
i
pooled panel ordinary
least squares (OLS) robust estimates, following Wooldridge (2002); and
i dynamic panel
Generalized Method of Moments (GMM) estimates, following the methodology
developed by Arellano and Bond (1991), Blun-dell and Bond (1998), Windmeijer
(2000) and Bond (2002).
3.1. Pooled panel OLS robust estimations. With pooled
total, ordinary least squares (OLS) robust estimates, we test the degree of
integration assuming a common intercept and a single set of slope coeffi-cients
for all the panel observations.
The obtained results for the three presented
equations are reported in Table 1 and in all situations reveal consistency. In
line with the previously presented unit root tests, the best results were
obtained without any lagged variables2,
indicating the dynamic and imme-diate reaction of bank lending growth to the
real per-capita GDP growth, the monetary policy interest rate and the four bank
performance indicators and condi-tions included in our model.
1 Good
explanations of the advantages and importance of using loga-rithmic
transformation in regression estimates are available, among others in Beauchamp
and Olson (1973) or Bartik (1985).
2 The results of the
estimations including lagged variables are available from the author upon
request.
Table
1. Pooled OLS robust estimations (*)
|
|
EQUATION I
|
EQUATION II
|
EQUATION III
|
|
|
|
|
|
|
Real
GDP per capita
|
|
|
|
|
|
coef.
|
|
.3054466
|
|
|
|
|
T-statistic
|
|
2.73
|
|
|
|
|
P-value
|
|
0.006
|
|
|
|
|
Interest
rate
|
|
|
|
|
|
coef.
|
|
.108883
|
.0944373
|
|
|
|
T-statistic
|
|
3.28
|
2.77
|
|
|
|
P-value
|
|
0.001
|
0.006
|
|
|
|
Deposits/GDP
|
|
|
|
|
|
coef.
|
|
.1937137
|
.2126949
|
.1918622
|
|
|
T-statistic
|
|
3.84
|
4.16
|
3.77
|
|
|
P-value
|
|
0.000
|
0.000
|
0.000
|
|
|
Bonds
and money market instruments / GDP
|
|
|
|
coef.
|
|
.1401866
|
.1427856
|
.159362
|
|
|
T-statistic
|
|
6.78
|
7.02
|
8.20
|
|
|
P-value
|
|
0.000
|
0.000
|
0.000
|
|
|
Foreign
assets / GDP
|
|
|
|
|
|
coef.
|
|
.1706834
|
.1625786
|
.1774548
|
|
|
T-statistic
|
|
4.45
|
4.40
|
4.92
|
|
|
P-value
|
|
0.000
|
0.000
|
0.000
|
|
|
Foreign
assets / Foreign liabilities
|
|
|
|
|
coef.
|
|
-.135372
|
-.1475844
|
-.1393685
|
|
|
T-statistic
|
|
-5.44
|
-6.11
|
-5.68
|
|
|
P-value
|
|
0.000
|
0.000
|
0.000
|
|
|
constant
|
|
|
|
|
|
coef.
|
|
-.5142468
|
.8122658
|
.97971
|
|
|
T-statistic
|
|
-1.10
|
5.93
|
9.23
|
|
|
P-value
|
|
0.270
|
0.000
|
0.000
|
|
|
|
|
|
|
|
|
|
|
|
N = 806
|
N = 806
|
N = 806
|
|
|
|
|
F (61, 744) =
|
F (60, 745) =
|
F (59, 746) =
|
|
|
|
|
1119.72
|
1226.02
|
1237.57
|
|
|
|
|
Prob > F =
|
Prob > F =
|
Prob > F =
|
|
|
|
|
0.0000
|
0.0000
|
0.0000
|
|
|
|
|
R-squared =
|
R-squared =
|
R-squared =
|
|
|
|
|
0.9773
|
0.9769
|
0.9766
|
|
|
Notes:
(*) Time and country dummies were included in the estimations and the obtained
results are available upon request.
According
to the results presented in Table 1, the three models are statistically
acceptable, as not only the values of the R-squares and the F-statistics are
very high but also the t-statistics and the correspondent p-values of all
variables are quite significant.
In
all situations, only the ratio of foreign assets to foreign liabilities has a
negative influence on the bank lending growth, confirming the high degree of
foreign dependence and indebtedness of the EU financial systems during this
period.
All
the other explanatory variables contribute posi-tively to bank lending growth.
In addition, the relative high influence of the ratio of the bonds and money
market instruments to GDP confirms that the EU finan-cial and credit systems
continue to be bank-dominated, since the increase of the bonds and money market
in-struments are in line with the bank lending growth.
Banks
and Bank Systems, Volume 5, Issue 2, 2010
The
positive contribution of the monetary policy interest rate to bank lending is
not a surprise, in view of the fact that during this period, the ECB in
particular, as well as the central banks of the non-EMU member-states,
maintained interest rates at historically low levels, thereby contributing to
the growth of the ratio bank lending to GDP.
3.2.
Arellano-Bond dynamic panel GMM estima-tions. In
addition, we present the results obtained with
dynamic Arellano-Bond panel GMM estimates (two-step difference), which consider
the model as a system of equations, one for each time period. The equations
differ by their individual moment condi-tion sets, since they all include the
endogenous and exogenous variables in first differences as instru-ments with
suitable lags of their own levels. By this use of instruments based on lagged
values of the explanatory variables, GMM controls for the poten-tial
endogeneity of all explanatory variables, al-though only for “weak” endogeneity
and not for full endogeneity, as explained by Bond (2002).
Next,
we check for the quality of the estimations by the Hansen test for
over-identifying restrictions and the Arellano-Bond tests for autocorrelation.
Table 2.
Arellano-Bond dynamic panel GMM two-step difference estimations
|
EQUATION I
|
EQUATION II
|
EQUATION III
|
|
|
|
|
|
Real
GDP per capita
|
|
|
|
|
|
coef.
|
-.1541594
|
|
|
|
|
z
|
-6.01
|
|
|
|
|
P>|z|
|
0.000
|
|
|
|
|
Interest
rate (lag1)
|
|
|
|
|
|
coef.
|
.0530916
|
.0512398
|
|
|
|
z
|
4.97
|
4.30
|
|
|
|
P>|z|
|
0.000
|
0.000
|
|
|
|
Deposits/GDP
|
|
|
|
|
|
coef.
|
.4676554
|
.4839136
|
.5198482
|
|
|
z
|
22.21
|
18.63
|
20.54
|
|
|
P>|z|
|
0.000
|
0.000
|
0.000
|
|
|
Bonds
and money market instruments/GDP (lag3)
|
|
|
|
coef.
|
.2189317
|
.1646729
|
.0797324
|
|
|
z
|
8.16
|
8.69
|
4.13
|
|
|
P>|z|
|
0.000
|
0.000
|
0.000
|
|
|
Foreign
assets/GDP
|
|
|
|
|
|
coef.
|
.0611868
|
.0809159
|
.086716
|
|
|
z
|
3.87
|
4.90
|
8.26
|
|
|
P>|z|
|
0.000
|
0.000
|
0.000
|
|
|
Foreign
assets/Foreign liabilities
|
|
|
|
|
coef.
|
-.1879588
|
-.1997773
|
-.1983791
|
|
|
z
|
-8.67
|
-10.83
|
-25.70
|
|
|
P>|z|
|
0.000
|
0.008
|
0.000
|
|
|
|
|
|
|
|
|
|
N = 702
|
N = 702
|
N = 702
|
|
|
Hansen
test of
|
chi2(129) = 21.30
|
chi2 (130) =
|
chi2 (131) =
|
|
|
24.46
|
22.67
|
|
|
Prob > chi2 =
|
|
|
overid. restrictions:
|
Prob > chi2 =
|
Prob > chi2 =
|
|
|
1.000
|
|
|
|
1.000
|
1.000
|
|
|
|
|
|
|
Banks and Bank Systems, Volume 5, Issue 2,
2010
Table 2.
Arellano-Bond dynamic panel GMM two-step difference estimations
|
EQUATION I
|
EQUATION II
|
EQUATION III
|
|
|
|
|
|
Arellano-Bond test
|
z = -1.88
|
z = -2.30
|
z = -1.93
|
|
|
for AR(1) in first
|
|
|
Pr > z = 0.060
|
Pr > z = 0.022
|
Pr > z = 0.053
|
|
|
differences:
|
|
|
|
|
|
|
|
Arellano-Bond test
|
z = -0.36
|
z = -0.67
|
z = -0.75
|
|
|
for AR(2) in first
|
|
|
Pr > z = 0.719
|
Pr > z = 0.501
|
Pr > z = 0.456
|
|
|
differences:
|
|
|
|
|
|
|
|
Table
2 reports the obtained results with dynamic Arellano-Bond two-step difference
GMM estima-tions for the three presented equations. Now, rein-forcing the
conclusions of the presented unit root tests, the best results in statistical
terms are obtained with lagged values, but only for the monetary policy
interest rate and for the ratio of bonds and money market instruments to GDP.
In
all situations, the Hansen test1
clearly does not reject the null that the instruments are valid and that they
are not correlated with the errors. At the same time, according to the results
of the Arellano-Bond tests, and as required for the validity of the
instru-ments, we may always accept that the residuals are clearly MA (1), but
not MA (2).
Furthermore, except for the growth of
the real GDP per capita2 (included only in equation (1), all the results obtained
with Arellano-Bond dynamic GMM estimates are in line with those obtained with
the pooled panel OLS estimates.
With
regard to real growth of the GDP per capita, we know that while it may be
possible to admit a positive relation between real GDP growth and bank lending
growth, it may also be true that during at least some of the considered time
periods, bank lending was not so directly connected with the pro-ductive
activities. This may be due either to the relatively independent and more
productive financ-ing of the productive activities, or to the channelling of
credit towards less productive activities, such as home buying or private
consumption, with no re-markable future productive multiplier effects.
Concluding
remarks
This
paper confirms the high degree of integration among the EU financial systems,
as well as the im-portance of bank performance conditions to the credit-lending
channel of monetary policy in the EU countries during recent years.
1 The Hansen test is a
test of over-identifying restrictions. The null hy-pothesis for this test is
that the instruments are valid in the sense that they are not correlated with
the errors in the first-differenced equation. Under the null, this test
statistic has a F_2q distribution with q
equal to the total number of instruments minus the number of parameters in the
model.
2 To
check the robustness of these results, we estimate several equations with and
without lags and in all situations with Arellano-Bond GMM estimates (two-step
difference), the real GDP per capita has a negative influence on the bank
lending to GDP. The results are available upon request.
We
contribute to the existing empirical evidence by the introduction into an
adaptation of the Bernanke and Blinder (1988) model not only of the real GDP
per capita or the monetary policy interest rate, but also of some specific
variables, representing the bank performance conditions, to explain bank
lend-ing to GDP, namely, the bank deposits / GDP ratio; the bonds and money
market instruments / GDP ratio, the foreign assets / GDP ratio and the foreign
assets / foreign liabilities ratio.
The
consistency of the obtained results, using pooled OLS and dynamic Arellano-Bond
GMM panel estimations, allows us to conclude that the EU banking institutions
have similar reactions to the variations of the macroeconomic conditions, in
par-ticular to the monetary policy interest rates as well as to the variations
of the bank performance condi-tions. The results also confirm the importance of
these variables to the bank lending growth (more precisely, the growth of the
ratio of the domestic credit provided by the banking institutions to GDP) in
the EU countries.
With
reference to the real GDP per capita, the ob-tained results, although
statistically robust, are in-conclusive as to the positive or negative
influence of this variable on the bank lending to GDP growth during this
period. With OLS robust estimates, which consider a fully integrated panel,
with com-mon intercept and a single set of slope coefficients, we conclude that
a faster growth of the real GDP per capita will contribute to a faster growth
of the bank lending to GDP growth. However, when using Arellano-Bond GMM
estimations, which consider the model as a system of equations, one for each
time period, we found a negative influence of the real GDP per capita growth to
bank lending growth.
Thus,
we may conclude that, in at least some of the considered time periods, bank
lending was not posi-tively related to the real GDP per capita growth. This may
be true in some EU countries, where the historically low levels of interest
rates oriented bank credit to many non-productive activities3.
These results are corroborated with the clear positive con-tributions of the
monetary policy interest rate to bank lending growth.
Furthermore, the results obtained with
the four in-cluded bank performance conditions allow us to state that:
3
Since we are using panel data estimates we can not identify exactly the
countries where bank lending growth is more negatively correlated with GDP
growth. Nevertheless, it is well known that more efficient and well developed
banking institutions should contribute to a more productive use of bank lending
and that during the considered time period the EU coun-tries and their banking
institutions were still adapting to the new market and credit conditions and
particularly to the intensification of competition.
Banks and Bank Systems, Volume 5, Issue 2,
2010
1.
the growth of the
ratio of deposits to GDP exerts a positive influence on the bank lending
growth, confirming the intermediate role of financial in-stitutions and the
fact that the capacity to attract savings (in the form of deposits) is always a
good condition in which to provide credit to those who need financing;
2.
the growth of the
ratio of bonds and money mar-ket instruments to GDP, which can be considered as
a proxy of the development of the financial markets in the EU countries, also
contributes positively to bank lending. This is symptomatic not only of the
fact that the EU financial markets continue to be bank-dominated, but also that
the development of the financial systems is always a good condition for the
direct and indirect financ-ing of the bank clients’ activities;
3.
as expected, the
growth of the ratio of foreign assets to GDP also exerts a positive influence
on the bank lending growth, as the entry of foreign assets received from the
other countries in-creases the resources to concede credit to the domestic
banks’ clients;
References
4.
the growth of the
ratio of foreign assets to for-eign liabilities contributes negatively to the
do-mestic bank lending growth, revealing not only the openness of the financial
markets, but more importantly, their indebtedness and the depend-ence of the EU
banking institutions on other countries’ financial resources.
Finally, it is clear that the total
credit provided by the UE banking institutions depends on the macro-economic
conditions, and particularly on the mone-tary policy decisions. At the same
time, bank lend-ing is an essential transmission channel of mone-tary policy
decisions, but it still depends on the performance conditions of the different
financial institutions.
Taking into account the lessons of the
recent banking and financial crisis, future research in this area will have to
underline the relevance of more efficient banking sectors and their specific
role in the transmission mechanisms of monetary policy.
1.
Angeloni, I., A.
Kashyap, B. Mojon and D. Terlizzese (2002), Monetary Transmission in the Euro
Zone: Where do we Stand?, ECB Working Paper Nº 114, January.
2.
Altunbas, Y., Ph.
Molyneux and J. Thornton (1997), "Big-Bank Mergers in Europe: An Analysis
of the Cost Im-plications", Economica, London School of Economics and
Political Science, 64 (254), pp. 317-329.
3.
Altunbas, Y., O.
Fazylow and Ph. Molyneux (2002), "Evidence on the bank lending channel in
Europe", Journal of Banking and Finance, 26 (11), pp.
2093-2110.
4.
Arellano, M. and S. Bond (1991), “Some
tests of specification for Panel Data: Monte Carlo evidence and an appli-cation
to employment equations”, The Review of Economic Studies, No. 58, pp.
277-297.
5.
Backhaus, J.G. (2003),
“Mass privatization in Central and East European countries: Implications for
corporate governance”, Journal of Economic Studies, 30 (3-4), pp.
196-204.
6.
Baele, L., A.
Ferrando, P. Hordahl, E. Krylova and C. Monnet (2004), Measuring banking
integration in the Euro area, ECB Occasional Paper No. 14.
7.
Barros, C.P., C.
Ferreira and J. Williams (2007), “Analysing the determinants of performance of
best and worse European banks: A mixed logit approach”, Journal of Banking
and Finance, 31 (7), pp. 2189-2203.
8.
Bartik, T. (1985),
“Business Location Decisions in the United States: Estimates of the Effects of
Unionization, Taxes and Other Characteristics of States”, Journal of
Business and Economic Statistics, 3 (1) pp. 14-22.
9.
Basile, R., M.
Constantini and S. Destefanis (2005), Unit Root and co-integration tests for cross-sectional
corre-lated panels. Estimation Regional production functions, CELPE, CSEF
Università di Salerno Discussion Paper 94.
10. Bean,
C., J. Larsen and K. Nikolov (2002), Financial Frictions and the Monetary
Transmission Mechanism: The-ory, Evidence and Policy Implications, ECB Working
Paper Nº 113, January.
11. Beauchamp,
J. and J. Olson (1973), “Corrections for Bias in Regression Estimates After
Logarithmic Transforma-tion”, Ecology, 54 (6), pp. 1403-1407.
12. Belaisch,
A., L. Kodres, J. Levy and A. Ubide (2001), Euro-Area Banking at the
Crossroads, IMF Working Paper No 01/28 Washington, D.C.: International Monetary
Fund.
13. Bernanke,
B. and A. Blinder (1988), "Credit, Money and Aggregate Demand", American
Economic Review, No. 78 (2), pp. 435-439.
14.
Bernanke, B. and M. Gertler (1995),
"Inside the Black Box: the credit channel transmission of monetary
policy",
Journal
of Economic Perspectives, 9 (4), pp. 27-48.
15. Blundell,
R. and S. Bond (1998), “Initial conditions and moment restrictions in dynamic
panel data models”, Jour-nal of Econometrics, No. 87, p. 115-143.
16. Bond,
S. (2002), Dynamic panel data models: A guide to micro data methods and
practice, Institute for Fiscal Studies, London, Working Paper No. 09/02.
17.
Bonin, J. and P.
Watchel (2004), “Dealing with Financial Fragility in Transition Economies”,
Federal Reserve Bank of Chicago, Conference on “Systemic Financial Crises:
Resolving Large Bank Insolvencies”, September 30.
18. Bonin,
J., I. Hasan and P. Watchel (2005-a), “Bank performance efficiency and ownership
in transition econo-mies”, Journal of Banking and Finance, 29 (1), pp.
31-53.
Banks and Bank Systems, Volume 5, Issue 2,
2010
19. Bonin,
J., I. Hasan and P. Watchel (2005-b), “Privatization matters: Bank efficiency
in transition economies”,
Journal
of Banking and Finance, Vol. 29 (8-9), pp. 2155-2178.
20.
Cabral, I., F. Dierick and J. Vesala
(2002), Banking integration in the Euro area, ECB Occasional Paper No. 6.
21.
Casu, B. and Ph.
Molyneux. (2000), A Comparative Study of Efficiency in European Banking, Centre
for Financial Institutions Working Papers No 00-17, Wharton School Centre for
Financial Institutions, University of Pennsylvania.
22. Cecchetti,
S. (1995), "Distinguishing Theories of the Monetary Transmission
Mechanism", FRB of St. Louis Eco-nomic Review, 77, pp. 83-97.
23. Dimitrova,
A. ed. (2004), Driven to Change: The European Union’s Enlargement views from
the East, Manchester University Press.
24.
Elbourne, A. and J. de Hann (2006),
Financial structure and monetary policy transmission in transition countries,
Journal
of Comparative Economics, 34 (1), pp. 1-23.
25. Erhmann,
M., L. Gambacorta,, J. Martinez-Pagés,, P. Sevestre and A. Worms (2001),
Financial Systems and the Role of Banks in Monetary Policy Transmission in the
Euro Zone, ECB Working Paper Nº 105.
26. European
Central Bank (2004), Research network on capital markets and financial
integration in Europe – Research and experience after two years, The Joint
ECB-CFS Research Network, December.
27. European
Central Bank (2005), Banking Structures in the New EU Member States, ECB,
January.
28.
Ferreira, C. (2007),
“Interest Rates, Banking Institutions and Transmission of Monetary Policy in
the Euro-Zone – Some empirical results for Portugal during the last decade”, European
Journal of Finance, 13 (1-2), pp. 181-193.
29. Ferreira,
C. (2008), “The Banking Sector, Economic Growth and European Integration”, Journal
of Economic Studies, forthcoming.
30. Fountas,
S. and A. Papagapitos (2001), “The monetary transmission mechanism: evidence
and implications for European Monetary Union”, Economics Letters, 70,
pp. 397-404.
31. Fries,
S. and A. Taci (2005), “Cost efficiency of banks in transition: Evidence from
289 in 15 post-communist countries”, Journal of Banking and Finance,
Vol. 29 (1), pp. 55-81.
32. Fries,
S., D. Neven, P. Seabrigh and A. Taci (2006), “Market entry, privatization and
bank performance in transi-tion”, Economics of Transition, Vol. 14 (4),
pp. 579-610.
33.
Gambacorta, L. (2004), “Inside the bank
lending channel”, European Economic Review, 49 (7), pp. 1737-1759.
34. Gambacorta,
L. and Mistrulli, P.E. (2004), “Does bank capital affect lending behavior?”, Journal
of Financial Intermediation, Nº 13, pp. 436-457.
35. Gardener,
E., Ph. Molyneux, Ph. and B. Moore (ed.) (2002), Banking in the New Europe –
The Impact of the Sin-gle Market Program and EMU on the European Banking
Sector, Palgrave, Macmillan.
36. Goddart,
J., Ph. Molyneux, J.O.S. Wilson and M. Tavakoli (2007), “European Banking: An
overview”, Journal of Banking and Finance, 31 (7), pp. 1911-1935.
37. Golinelli,
R. and R. Rovelli (2005), Monetary policy transmission, interest rate rules and
inflation targeting in three transition countries, Journal of Banking and
Finance, 29 (1), pp. 183-201.
38. Gujarati
D. (2003), Basic Econometrics, McGraw-Hill. Boston
39. Hanousek,
J. and E. Kocenda (2003), “The impact of Czech mass privatisation on corporate
governance”, Journal of Economic Studies, 30 (3-4), pp. 278-293.
40. Holscher,
J. ed. (2000), Financial Turbulence and Capital Markets in Transition
Countries, Palgrave editions.
41. Hubbard,
R.G. (1995), “Is there a credit channel for monetary policy?”, Federal
Reserve Bank of St. Louis Review, 77 (3), pp. 63-77.
42. Jansen,
D-J. and J. de Haan (2003), Increasing concentration in European banking: a
macro-level analysis WO No. 743, Netherlands Central Bank, Research Department.
43. Karlsson,
S. And M. Lothgren (2000), “On the power and interpretation of panel unit root
tests”, Economics Let-ters, 66 (3), pp. 249-255.
44. Kashyap,
A.K. and J. C. Stein (1997), “The role of banks in monetary policy: A survey
with implication for the European Monetary Union”, Economic Perspectives,
Federal Reserve Bank of Chicago, 21, pp. 2-19.
45. Kashyap,
A.K. and J.C. Stein (2000), “What do a million banks have to say about the
transmission of monetary policy”, American Economic Review, 90, pp.
407-428.
46. Kishan,
R.P. and T.P. Opiela (2006), “Bank capital and loan asymetry in the
transmission of monetary policy”,
Journal
of Banking and Finance, 30 (1), pp. 259-285.
47. Kocenda,
E., A.M. Kutan and T.M. Yigit (2006), "Pilgrims to the Eurozone: How Far,
How Fast?" Economic Systems, 30 (4), pp. 311-327.
48.
Levin, A., C. Lin and C. Chu (2002),
“Unit Root Tests in Panel Data: Asymptotic and Finite Sample Properties”,
Journal
of Econometrics,
No. 108, pp.1-24.
49. Melnik,
A. and D. Nissim (2006), “Issue costs in the Eurobond market: The effects of
market integration”, Journal of Banking and Finance, 30 (1), pp.
157-177.
50. Mishkin,
F. (1995), “Symposium on the monetary transmission mechanisms”, Journal of
Economic Perspectives, 9 (4), pp. 3-10.
51. Mishkin,
F. (2001), The Economics of Money, Banking and Financial Markets, Sixth
Edition, Addison Wesley, World Student Series.
52. Molyneux,
Ph. (2003), “Does Size Matter?” – Financial Restructuring under EMU EIFC -
Technology and Fi-nance Working Papers 30, United Nations University, Institute
for New Technologies.
Banks and Bank Systems, Volume 5, Issue 2, 2010
53. Romero-Ávila,
D. (2003), Finance and growth in the EU: New evidence from the liberalisation
and harmonisation of the banking industry, ECB Working Paper, No. 266.
54. Romero-Ávila,
D. (2007), “Finance and growth in the EU: New evidence from the harmonisation
of the banking industry”, Journal of Banking and Finance 31 (7), pp.
1937-1954.
55.
Stephen F.H. and J.G.
Backhaus (2003), “Corporate governance and mass privatisation: A theoretical
investigation and transformations in legal and economic relationships”, Journal
of Economic Studies, 30 (3-4), pp. 389-468.
56. Sztyber,
W. (2003), “Polish mass privatisation scheme: The programme of national
investment funds”, Journal of Economic Studies, 30 (3-4), pp.
205-228.
57. Taylor,
J. (1995), "The Monetary Transmission Mechanism: An Empirical
Framework", Journal of Economic Perspectives, 9, pp.11-26.
58. Tchipev,
P. (2003), “Bulgarian mass privatisation scheme: Implications on corporate
governance”, Journal of Economic Studies, 30, (3-4), pp. 351-388.
59. Topi,
J. and J. Vilmunen (2001), Transmission of Monetary Policy Shocks in Finland:
Evidence from Bank Level Data on Loans, ECB Working Paper No. 73, July.
60. Van
Els, P., A Locarno, J. Morgan and J-P Villetelle (2001), Monetary Policy
Transmission in The Euro Zone: What Do Aggregate and National Structural Models
Tell Us?, ECB Working Paper NO 94, December.
61. Windmeijer,
F. (2000), A finite simple correction for the variance of linear two-step GMM
estimators, Institute for Fiscal Studies, London, Working Paper No 00/19.
62.
Winkler, A. ed. (2002), Banking and
Monetary Policy in the Eastern Europe: the First Ten Years, Palgrave editions.
63. Wooldridge,
J. (2002), Econometric Analysis of Cross Section and Panel Data, the MIT Press.
64.
Worms, A. (2001), The Reaction of Bank
Lending to Monetary Policy, ECB Working Paper No. 96, December.
Appendix A. Summary statistics
VARIABLES
|
Mean
|
Std.
dev.
|
Min
|
Max
|
Observations
|
|
|
|
|
(all
in natural logarithms)
|
|
|
|
|
|
|
|
|
|
Bank
lending/GDP:
|
|
|
|
|
|
|
|
overall
|
.9634144
|
1.106034
|
-3.23828
|
3.39354
|
N = 806
|
|
|
between
|
|
1.10247
|
-2.791806
|
3.356673
|
n = 26
|
|
|
within
|
|
.2305816
|
.0820338
|
3.117834
|
T = 31
|
|
|
Real
GDP per capita:
|
|
|
|
|
|
|
|
overall
|
6.051168
|
2.678176
|
1.34968
|
12.66796
|
N = 806
|
|
|
between
|
|
2.72726
|
1.443205
|
12.42524
|
n = 26
|
|
|
within
|
|
.1089511
|
5.524108
|
6.514988
|
T = 31
|
|
|
Interest
rate:
|
|
|
|
|
|
|
|
overall
|
1.481935
|
.56964
|
-.02703
|
3.55535
|
N = 806
|
|
|
between
|
|
.4792346
|
.7142648
|
3.06961
|
n = 26
|
|
|
within
|
|
.3215321
|
.4369553
|
2.581846
|
T = 31
|
|
|
Deposits/GDP:
|
|
|
|
|
|
|
|
overall
|
1.295129
|
1.519575
|
-2.77394
|
6.04847
|
N = 806
|
|
|
between
|
|
1.528612
|
-2.488646
|
5.997196
|
n = 26
|
|
|
within
|
|
.2439945
|
-.3845842
|
1.981864
|
T = 31
|
|
|
Bonds
and money market instruments/GDP:
|
|
|
|
|
|
|
|
overall
|
-.0795288
|
1.750138
|
-5.39641
|
2.28638
|
N = 806
|
|
|
between
|
|
1.695878
|
-3.744695
|
1.986973
|
n = 26
|
|
|
within
|
|
.5423645
|
-2.622679
|
1.495851
|
T = 31
|
|
|
Foreign
assets/GDP :
|
|
|
|
|
|
|
|
overall
|
-.080594
|
2.21202
|
-10.41371
|
3.23734
|
N = 806
|
|
|
between
|
|
2.240099
|
-9.21917
|
2.771957
|
n = 26
|
|
|
within
|
|
.2489938
|
-1.275133
|
.6851366
|
T = 31
|
|
|
Foreign
assets/Foreign liabilities :
|
|
|
|
|
|
|
|
overall
|
-.0051242
|
.7618599
|
-2.47735
|
2.88475
|
N = 806
|
|
|
between
|
|
.6818787
|
-1.203865
|
2.336299
|
n = 26
|
|
|
within
|
|
.3644169
|
-1.446609
|
2.090331
|
T = 31
|
|
|
Appendix
B. Correlation matrix (*)
|
Real lending/
|
Real GDP
|
Interest
rate
|
Deposits/
|
Bonds and money
|
Foreign as-
|
Foreign as-
|
|
|
|
|
|
|
market instruments/
|
sets/Foreign
|
|
|
|
GDP
|
per capita
|
GDP
|
sets/GDP
|
|
|
|
|
GDP
|
liabilities
|
|
|
|
|
|
|
|
|
|
|
Bank
lending/GDP
|
1.0000
|
|
|
|
|
|
|
|
|
Real
GDP per capita
|
-0.1951
|
1.0000
|
|
|
|
|
|
|
|
Banks and Bank Systems, Volume 5, Issue 2,
2010
Appendix B (cont). Correlation matrix (*)
|
Real lending/
|
Real
GDP
|
Interest
rate
|
Deposits/
|
Bonds
and money
|
Foreign as-
|
Foreign as-
|
|
|
market instruments/
|
sets/Foreign
|
|
|
GDP
|
per
capita
|
GDP
|
sets/GDP
|
|
|
|
GDP
|
liabilities
|
|
|
|
|
|
|
|
|
Interest
rate
|
-0.4227
|
0.1853
|
1.0000
|
|
|
|
|
|
Deposits/GDP
|
0.7154
|
-0.1843
|
-0.3777
|
1.0000
|
|
|
|
|
Bonds
and money market instru-
|
0.4828
|
-0.4132
|
-0.3314
|
0.4144
|
1.0000
|
|
|
|
ments/GDP
|
|
|
|
|
|
|
|
|
|
|
|
Foreign
assets/GDP
|
0.8005
|
-0.2019
|
-0.5605
|
0.6140
|
0.5878
|
1.0000
|
|
|
Foreign
assets/Foreign liabilities
|
0.2235
|
-0.1555
|
-0.2109
|
0.4341
|
0.1835
|
0.3939
|
1.0000
|
|
Notes:
(*) Several of these correlations seem rather high and, in order to
reduce the multicollinearity problems, we could have tried an orthogonalization
test, but, following among others, Gujarati (2003) these correlations can be
considered in an acceptable range.
Appendix C. Panel unit root tests –
Levin-Lin-Chu
VARIABLES
|
Lags
|
Coefficients
|
T-value
|
T-stat.
|
P>t
|
N
|
Bank
lending / GDP
|
0
|
-0.85254
|
-48.179
|
-43.23521
|
0.0000
|
750
|
|
1
|
-0.50974
|
-15.206
|
2.11907
|
0.9830
|
725
|
|
2
|
-0.40864
|
-10.955
|
9.39903
|
1.0000
|
700
|
|
3
|
-0.38976
|
-11.328
|
10.91595
|
1.0000
|
675
|
Real
GDP per capita
|
0
|
-1.01649
|
-28.060
|
-18.99302
|
0.0000
|
750
|
|
1
|
-1.57624
|
-38.559
|
-26.68914
|
0.0000
|
725
|
|
2
|
-1.89295
|
-26.221
|
-7.30147
|
0.0000
|
700
|
|
3
|
-0.37484
|
-8.712
|
25.39089
|
1.0000
|
675
|
Interest
rate
|
0
|
-0.16644
|
-8.404
|
0.48152
|
0.6849
|
750
|
|
1
|
-0.22246
|
-14.416
|
-5.64454
|
0.0000
|
725
|
|
2
|
-0.26835
|
-15.240
|
-5.20633
|
0.0000
|
700
|
|
3
|
-0.29185
|
-13.809
|
-1.49730
|
0.0672
|
675
|
Deposits
/ GDP
|
0
|
-0.40334
|
-13.622
|
-5.38483
|
0.0000
|
750
|
|
1
|
-0.38278
|
-11.697
|
-2.25471
|
0.0121
|
725
|
|
2
|
-0.30752
|
-9.013
|
1.43541
|
0.9244
|
700
|
|
3
|
-0.24927
|
-7.173
|
4.77273
|
1.0000
|
675
|
Bonds
and money market instruments /
|
0
|
-0.20377
|
-8.980
|
-0.24074
|
0.4049
|
750
|
GDP
|
|
|
|
|
|
|
|
1
|
-0.22969
|
-9.423
|
-0.19688
|
0.4220
|
725
|
|
2
|
-0.20166
|
-7.782
|
2.50132
|
0.9938
|
700
|
|
3
|
-0.34266
|
-12.507
|
-2.97402
|
0.0015
|
675
|
Foreign
assets / GDP
|
0
|
-0.29999
|
-11.244
|
-2.56597
|
0.0051
|
750
|
|
1
|
-0.29557
|
-10.280
|
-0.78186
|
0.2171
|
725
|
|
2
|
-0.28142
|
-8.924
|
1.69569
|
0.9550
|
700
|
|
3
|
-0.31657
|
-9.217
|
2.43607
|
0.9926
|
675
|
Foreign
assets / Foreign liabilities
|
0
|
-0.17329
|
-9.362
|
-1.78288
|
0.0373
|
750
|
|
1
|
-0.19161
|
-9.696
|
-1.77454
|
0.0380
|
725
|
|
2
|
-0.20652
|
-9.886
|
-1.47377
|
0.0703
|
700
|
|
3
|
-0.25318
|
-11.463
|
-2.60665
|
0.0046
|
675
|
240