Retail Bank Interest Rate Pass-Through: The Turkish Experience pptx

International Research Journal of Finance and Economics
ISSN 1450-2887 Issue 28 (2009)
© EuroJournals Publishing, Inc. 2009


Retail Bank Interest Rate Pass-Through: The Turkish
Experience


Bilge Kagan Ozdemir
1

Department of Economics, Anadolu University, Turkey
Yunus Emre Campus, 26470, Turkey
E-mail:


Abstract

This paper intends to investigate the relationship between a money market rate and
banks’ retail rates by empirically examining the pass-through process in the banking system
of Turkey for the period between April 2001 and June 2007. We also aim to highlight the
main factors that influence the price setting behavior of banks. The main findings from
symmetrical and asymmetrical error-correction models suggest that the pass through from
the market rate to deposit and lending rate is complete in the long run, while in the short
run lending rate shows more flexibility relative to deposit rate. In addition, there is greater
rigidity in deposit and lending rate decreases than increases and retail interest rates does not
adjust asymmetrically to an increase or a decrease in money market rate in Turkey.


Keywords: Interest rate pass-through, retail interest rates, Turkish banking system, ECM

JEL Classification Codes: E43, E52, G21

1. Introduction
This study deals with the relationship between a money market rate and banks’ retail rates in the
banking system of Turkey by empirically examining the pass-through that is defined as the degree and
the speed of adjustment of interest rates to money market rate. Most central banks use a short-term
interest rate as their instrument of monetary policy. Changes to this short term interest rate are the first
important stage in the transmission of monetary policy. The transmission mechanism is usually
sluggish and incomplete in the short term but rather complete in the medium and long-term. Thus,
changes in the local currency money market rate, induced by monetary policy decisions or by changing
conditions in the money market, are transmitted to local currency bank retail rates but only with lags.
The speed of response is usually associated to the effectiveness of monetary policy, so measuring the
pass-through is crucial to improving the understanding of the whole process.
The primary concern in this study is twofold: First, I want to investigate empirically the
relationship between a money market rate and banks retail lending rates. Second, I would like to
highlight one of the main factors that influence the price setting behavior of bank in a developing
economics context. For this aim, a single equation error correction models (ECM) is used in order to
quantify the degree and the speed of pass through and estimate symmetric and asymmetric dynamic
adjustment of interest rates in the Turkish banking system both in long and short run.

1 "The first version of this paper was prepared with Professor Ilyas Siklar and presented at Singapore Economic Review
Conference 2007, (2-4 August 2007). The author is deeply indebted to Professor Ilyas Siklar for his superb advice and
contribution regarding methodology and furnishing data. "
International Research Journal of Finance and Economics - Issue 28 (2009) 8
The remainder of the paper is organized as follows; the following part provides a brief review
of related theoretical and empirical literature. Section three presents some relevant background on the
Turkish economy. Section four discusses the methodological issues and presents empirical results that
are obtained. Finally, in section four, includes some concluding remarks.

2. A Brief Review of Related Theoretical and Empirical Literature
The effectiveness of monetary policy depends on the degree and the speed of interest rate adjustment to
change in policy controlled interest rate. Both theoretically and empirically, the interest rate channel of
monetary transmission has received great attention. A common element found by all researchers is the
stickiness of the retail banks interest rate. One of the most famous attempts to explain interest rate
stickiness is Stiglitz and Weiss (1981). This study examines several information asymmetries that
generate rationing in otherwise competitive market. As such a loan market may be characterized as
credits rationing among observationally identical borrowers some receive loans and others do not. In
this setup expected earnings of banks from loans market are a function of interest rate on loans since
probability of default of borrower is increasing with higher interest rates. Increasing interest rates could
increase the risk of banks’ loan portfolio either by adverse selection or by moral hazard and due to this
increase in riskiness, banks can not increase lending rates even under the case they face higher
marginal costs, i.e. money market rate. Otherwise, safer investors would be discouraged and riskier
ones would be attracted (adverse selection) or borrowers would be induced to invest in riskier projects
(moral hazard) because of higher interest rates. Therefore existence of asymmetric information
between borrowers and lenders in loan market may create an upward stickiness in lending lates.
Hannah and Berger (1991) and Cotterelli and Kourelis (1994) deem the mechanism through
adjustment costs and the elasticity of the demand for loans in order to explain the interest rate
stickiness. They suggest that banking industry, like any industry, faces adjustment costs when prices,
i.e., money market rate, are changed. According to these studies the speed of adjustment of lending
rates to changes in policy interest rate may depend on the elasticity of the demand for bank loans that
depend on the structure of the financial system. Klemperer (1987), in his two period model, shows that
in a mature market switching costs like learning costs, transaction costs or artificial costs imposed by
firms, may cause to stickiness in prices.
The possibility of asymmetric bank interest rate adjustment has been explored, in particular, by
Hannan and Berger (1991) and by Neumark and Sharpe (1992) which was later developed by Scholnic
(1996). These studies both distinguish between symmetrical and asymmetrical rigidities in deposit and
lending rates due to cost of adjustments and they offer two reasons for asymmetrical adjustment:
costumer reactions and collusive pricing arrangements. According to customer reaction hypothesis, one
may expect greater rigidity in deposit rate decreases and in loan rate increases due to customers

unfavorably reactions to unstable rates. Second, especially in concentrated markets, deposit rates may
be more rigid upwards and lending rates may be more rigid downwards due to cost of the breaking
collusive pricing arrangements of banks.
The empirical literature on the interest rate pass through can be divided in to two groups based
on the type of the data used. The first group of scholars, Mojon (2000); Bredin, Fitzpatrick and
O’Reilly (2001); Heinemann and Schüler (2002); Sander and Kleimeier (2002) Cottarelli and Kourelis
(1994) and Borio and Fritz (1995) focus on aggregate interest rate series for individual and a panel of
countries, typically using singe equation error correction models (ECM) in order to quantify the degree
and the speed of pass through. In these studies, differences across countries are related to structural
factors such as level of concentration or financial market characteristics. For example, Cottarelli and
Kourelis (1994) show that the degree of short term pass through is mainly related with the factors like
the level of concentration in the banking system, inflation, volatility of money market rate and bank
costs. The second group of scholars, De Graeve et al. (2004), Sorensen and Werner (2006) and Horvat
et al. (2004) use micro level data with panel method. Horvat et al., for instance, analyze the pass
9 International Research Journal of Finance and Economics - Issue 28 (2009)
through of market conditions to retail bank interest rates in Hungary with a panel of bank deposit and
loan rates. They find out a clear difference in the pricing of household and corporate instruments.


3. Some Background on the Turkish Banking System
At the end of 1999 Turkey adopted 17th IMF supported stabilization program, based on pre-announced
crawling peg system. It was an ambitious program which attempted to solve many problems at the
same time; corruption, privatization, social security, banking supervision, monetary prudence.
According to the program; after eighteen months, from the beginning of the program, the exchange rate
would be allowed to fluctuate in a continuously widening band. However, 12 months after launching of
this stabilization program Turkey experienced liquidity crisis began in November 2000 and ended on
the February 19, 2001.
Turkish banking system, was not only one of the main factors but also the key trigger of the
crises of 2000 and 2001. According to Ozatay and Sak (2002) without a fragile banking system and
triggering factors, high current account deficit and real appreciation of the lira would not be enough to

precipitate the 2000-2001 crises. They provide a plenty of evidence regarding the risk accumulation in
the banking system in the period preceding the crisis: increase in currency and maturity mismatches
and a rise in non-performing loans. Hence, the banking system was highly vulnerable to capital
reversals. However, risk accumulation was not homogenous throughout the system. There were two
different types of dichotomization: Private versus state banks and within the private banks. While the
state banks were more open to interest rate risk, private ones were more prone to exchange rate risk.
Within the private banking system there were some midsized banks that were heavily concentrated in
government debt instruments business. Moreover, they were carrying these instruments by borrowing
extremely short-term.
Ozkan (2004) point to three sets of vulnerabilities in the Turkish economy that prepared the
ground for the collapse of the Turkish Lira and the resulting 2000-2001 financial crisis. The first
source of vulnerabilities identified was the weak external position caused by excessive levels of debt
repayments. The second has been the weak fiscal position resulting from the record levels of interest
payments on domestic borrowing. When combined with the unfavorable maturity structure of the
existing debt, this resulted in debt servicing placing a considerable burden on the public finances
during this period. Thirdly, her analysis suggests that the weaknesses in the financial and banking
sector have played a major role in preparing the ground for the liquidity squeeze in November 2000
and in aggravating the situation in the wake of the devaluation in February 2001.
Central Bank of the Republic of Turkey (hereinafter CBRT) has gained its independence after the
2001 crisis (in April 2001) as a result of recovery attempts. During the initial process of independence, the
CBRT announced the medium term objective as a gradual move to an inflation targeting framework. As
such at the beginning of 2002 the CBRT began implementing implicit inflation targeting. Eventually, at the
beginning of 2006 the CBRT introduced full-fledged inflation targeting (see Akyürek and Kutan (2006) for
more detailed information).
While high debt level, inflation and macroeconomic instability prevented banks from credit
supply during the pre 2001 crisis era, credit channel has started to work properly in the post crisis era
due to sound macroeconomic policies. During this period, loan portfolio of Turkish Banking System
also showed impressive growth. As of December 2006, the amount of in-cash loans extended by banks
was 219 billion NTL (approximately 156,42 billion USD). Non-performing loans totaled NTL 8,550
billion (approximately 6,110 billion USD) as of December 2006. The amount of provisions set aside

for these loans is NTL 7.665 billion (approximately 5.475 billion USD).
International Research Journal of Finance and Economics - Issue 28 (2009) 10
Table 1: Main Indicators of the Banking Sector

NTL Billion 2005 2006
Total assets 397.0 499.7
Loans 208.5 219.0
Securities portfolio 143.0 158.9
Deposits 251.5 307.6
Source: BRSA Annual Report 2006

The total amount of deposits of the sector as of the end of December 2006 was 307.6 billion
NTL (approximately 219,28 billion USD). 186.3 billion NTL (approximately 133,07 billion USD) of
these deposits are denominated in TL, and 121.4 NTL billion (approximately 86,71 billion USD) in
FX. The fact that 87.3% of the total is held by the 10 large-scale deposit banks shows the high level of
concentration in the sector. The maturities of most of the deposits are short-term. Deposits having
maturities of less than three months constitute about 95.2% of total deposits. Since 2002, total assets of
the Turkish banking system have been growing steadily. As of 2006 total assets reached 499.7 billion
NTL (approximately 321,21 billion USD) (BRSA, 2006).


4. Methodological Issues
4.1. The Data
The data set consist of three interest rate series, called money market rate, lending rate and deposit rate.
Data were collected from the “Electronic Data Dissemination System” of the Central Bank of the
Republic of Turkey for money market rate and deposit rate series and from Turkish Banking
Association reports for lending rate series. Monthly data covering the period April 2001 up to June
2007 is used in the analysis. This period has been chosen for two reasons. First, it is this period when
credit channel of Turkish Banking System (TBS) has started to work properly and total assets of TBS
have been growing steadily. Second, prior to this period the CBRT was not independent. The time

series of variables used in this study are shown in Figure 1.

Figure 1: Trends in Money Market, Lending and Deposit Rates

.1
.2
.3
.4
.5
.6
.7
.8
.9
2001 2002 2003 2004 2005 2006
MMR LR DR


4.2. Method of Estimation
In order to analyze dynamic interest rate adjustment between the policy controlled interest rate and the
interest rates on loans or deposits, symmetric error correction model (ECM) is used as a starting point.
11 International Research Journal of Finance and Economics - Issue 28 (2009)
In this model, short run dynamics are linked to long run equilibrium. Our specification of the ECM
takes the following form:
tt
J
j
jtj
I
i
itit

ectmmrrr
ελψβ
++∆+∆=∆
−
=
−
=
−
∑∑
1
01
(1)
where r
t
denotes deposit rate (dr) or lending rate (lr), mmr is the money market over-night interest rate,
and ect is the error correction term estimated from the respective cointegration regressions and ε
t
is the
error term assumed to be normally distributed and not serially correlated. The term ect is one period
lagged deviation from the long run equilibrium. The speed of adjustment parameter is λ, which has a
sensible economic interpretation if it has a positive value. The interpretation of speed of adjustment
value follows this example. If λ estimated of -0.25, this indicates that in a case of a shock to the deposit
or lending rate which changes its value relative to equilibrium value, the one forth of divergence is
eliminate in the following period.
Following Scholnic (1996) and Kleimeir and Sander (2000), asymmetric error correction model
is also estimated in order to find out if interest rates adjust differently to the cases in which they are
below or above their equilibrium levels. In this framework error correction terms from the co-
integration regression is separated into two components (Kleimeir and Sander, 2000, pp.8), such that;
(
)

otherwiseect
ectmeanectifectect
t
tt
0=
>=
+
+
(2)
and
(
)
otherwiseect
ectmeanectifectect
t
tt
0=
<=
−
−
(3)
The positive errors (
+
t
ect
) in equation (2) implies that if the deposit or lending rate is above its
equilibrium value following a decline in the money market rate, then it will start falling in the next
period. Similarly, the negative errors (
−
t

ect
) in equation (3) means that if deposit or lending rate is
below their equilibrium value following an increase in the money market rate, they will start increasing
in the subsequent period. Now the specification of asymmetric short run dynamic equation takes the
following form:
tt
J
j
tjtj
I
i
itit
ectectmmrrr
εααψβ
+++∆+∆=∆
−
−
=
+
−−
=
−
∑∑
12
0
11
1
(4)
The estimated coefficient of α
1

measures the speed of adjustment in response to the previous
period disequilibrium relationship between deposit or lending rates and money market rate when rates
are above their equilibrium level whereas the estimated coefficient of α
2
measures the speed of
adjustment in response to the previous period disequilibrium relationship between variables when rates
are below their equilibrium level. The test of whether retail interest rates adjust asymmetrically is
whether α
1
is significantly different from α
2
, and if, α
1
>
α
2
for deposits, banks are quicker in adjusting
deposit rates downwards than they are to adjust them upwards while α
2
>
α
1
in the loan market implying
that banks adjust lending rates upwards faster than they are to adjust them downward.

4.3. Empirical Results
In the first stage, the order of integration is tested using the Augmented Dickey Fuller (ADF) unit root
tests. Unit Root tests are conducted to verify the stationarity properties (absence of trend and long-run
mean reversion) of the time series data so as to avoid spurious regressions. A series is said to be
(weakly or covariance) stationary if the mean and autocovariances of the series do not depend on time.

If a time series has a unit root, a widespread and convenient way to remove non-stationarity is by
taking first differences of the relevant variable. A non-stationary series which by differencing d times
transfers to a stationary one, is called integrated of order
d
and denoted as I(d) (Charemza and
International Research Journal of Finance and Economics - Issue 28 (2009)
12

Deadman, 1997). In order to investigate the time series properties of the variables the sequential testing
procedure suggested by Holden and Thompson (1992) is used. The main advantage of this procedure is
that it takes into account the full data generating process with drift and trend terms and tests the null
hypothesis of unit root sequentially.

Table 2: Augmented Dickey Fuller Unit Root Test

Variable
Levels First Difference
Lag Order Test Statistics Probability Lag Order Test Statistics Probability
dr 1 -1.04 0.93 1 -5.87* 0.00
lr 4 -0.23 0.99 1 -5.81* 0.00
mmr 1 -0.57 0.98 0 -6.23* 0.00
Note:* Denote rejection of null hypothesis at the 5% and level. Number of lags were chosen in accordance with the
Schwarz info criterion. Critical values are from McKinnon (1996).

The results of the test for all variables are presented in Table 1. The results show that each of
the series is non-stationary when the variables are defined in terms of levels. First differencing the
series removed the non stationary components in all series, concluding that all series are integrated of
order one.
Engel and Granger (1987) argue that, although a set of economic series are not stationary, there
may exist some linear combination of the variables that is stationary. Now, we proceed to test for co-

integration in the data series in order to test whether there is a long run relationship between the
variables. The ADF test on the residuals based on the regression of money market rate and lending
rate, and the money market rate and deposit rate, show that the residuals are indeed I(0), enabling us to
conclude that both the deposit and lending rates are cointegrated with the money market rate. This
study uses the Johansen (1988) approach to test for co-integration. The Johansen co integration
procedure gives the results as reported in Table 3 for lending rate and money market rate relation and
in Table 4 for deposit rate and money market rate relation. The maximum eigenvalue and trace
statistics show that there are one cointegrating vector between lending rate and money market rate, and
deposit rate and money market rate financial development proxies and economic growth at the 5 per
cent level This confirms the existence of an underlying long-run stationary steady-state relationship
between the bank interest rates and the developments in money market rates.

Table 3: Lending Rate – Money Market Rate Equation

Maximun Eigen Value
Null Alternative Statistic %5 Critical Value Probability
r = 0 r =1 34.62 19.39* 0.00
r≤ 1 r = 2 10.08 12.52 0.12
Trace Test Statistics
Null Alternative Statistic %5 Critical Value Probability
r = 0 r ≥ 1 44.69 25.87* 0.00
r ≤ 1 r = 2 10.08 12.52 0.12
Note:* Denote rejection of null hypothesis at the 5% level. Critical values are from Mc Kinnon, Haug, Michelis (1999)

Table 4: Deposit Rate-Money Market Rate Equation

Maximun Eigen Value
Null Alternative Statistic %5 Critical Value Probability
r = 0 r =1 22.82 19.39* 0.02
r≤ 1 r = 2 7.62 12.52 0.28

Trace Test Statistics
Null Alternative Statistic %5 Critical Value Probability
r = 0 r ≥ 1 30.44 25.87* 0.03
r ≤ 1 r = 2 7.62 12.52 0.29
Note:* Denote rejection of null hypothesis at the 5% level. Critical values are from Mc Kinnon, Haug, Michelis (1999)
13
International Research Journal of Finance and Economics - Issue 28 (2009)
The estimates of the normalized cointegrating vector (long run) which shows long run
adjustments for lending rate-money market rate relation and deposit rate-money market rate relation is
shown with following equations;
Normalized cointegrating vector: lr – 1.192(mmr) (5)
(0.052)
and
Normalized cointegrating vector: dr – 1.054(mmr) (6)
(0.050)
Based on equation (5) and (6), it can be said that the long run adjustment of lending rate and
deposit rate with respect to 1 percent change in money market rate is 1,192 and 1,054, respectively
which implies that in long run there is complete pass through from changes in money market rate to
banks’ retail rates.

Table 5: Results of the Error Correction Model (ECM)

Variable
lr dr
Symmetric Asymmetric Symmetric Asymmetric
Intercept -0.01 (2.18) -0.01(1.51) 0.01 (1.56) 0.01(0.48)
∆lrt
-
1
0.74(5.42) 0.26(1.54)

∆dr
t
-
1
0.45 (2.82) 0.23(1.55)
∆mmr
t
-
1
-0.38(2.18) -0.03(0.13) 0.09(0.52) -0.06(0.32)
ecm
t
-
1
-0.35(3.30) -0.12(1.25)
+
−1t
ecm

-0.19(0.98) -0.10(0.70)
−
−1t
ecm

-050(2.32) -0.69(3.75)
Wald statistics 0.29 0.03
R
2
0.33 0.29 0.21 0.35
DW 1.86 1.41 1.97 1.48

Note: First numbers are estimates for the coefficients and t-statistics are displayed in parentheses

After employing Enders’ (1995) and Hendry’s (1986) General to Specific Paradigm 2 lags, a
constant and a trend is included in to ECM model. A number of lags for each of the three variables
have been included to capture the short-run dynamic relationship in the ECM system. Table 5 gives the
results of the dynamic short run adjustments estimated for symmetric ECM model and asymmetric
ECM model. In both the lending rate-money market rate relationship and the deposit rate-money
market rate relationship, the coefficients on the error correction terms are significantly negative as
required. The symmetric short run adjustment of the lending rate following a deviation from long run
equilibrium in the previous period is 0.35 for lending rate and 0.12 for the deposit rate.
As stated earlier, according to customer reaction hypothesis, one may expect greater rigidity in
deposit rate decreases and in loan rate increases due to customers unfavorably reactions to unstable
rates. Second, especially in concentrated markets, deposit rates may be more rigid upwards and lending
rates may be more rigid downwards due to cost of the breaking collusive pricing arrangements of
banks. The estimated coefficients of the asymmetric adjustments are in line with the common intuition,
although they are both statistically insignificant at the 5% significance level. According to the
estimated model, the lending rate adjusts downwards by 0.19 following a previous period’s decline in
the money market rate, compared with an upward adjustment of 0.50 and deposit rate adjusts
downward by 0.10 and upward by 0.69 in response to disequilibrium in the money market.
As a final step, a Wald test is used in order to test whether estimated asymmetric coefficients
are significantly different from each other. As Scholnic (1996) states, if α
1
is not significantly different
from α
2
, this implies that there is no significant asymmetry where the retail rate is increasing as
opposed to decreasing An important step in the analysis of asymmetric adjustment is to test whether
International Research Journal of Finance and Economics - Issue 28 (2009)
14

the estimated asymmetric coefficients are equal. From the bottom part of table 5, one can see that the
null hypothesis of equal adjustments cannot be rejected as the Wald statistics are significantly less than
the critical value (χ
2
(1) = 1.94) at 5% level of significance. Therefore, lending and deposit rates adjust
equally to a change in money market rate.


5. Conclusion
In this paper, we investigate the symmetric and asymmetric adjustments of retail interest rates to
money market rate in Turkey for the period between April 2001 and December 2006. This process is
important since the speed and the degree of response is usually associated to the effectiveness of
monetary policy. The empirical results show that Turkish Banking System adjust completely their
interest rates to changes in the market rates in the long run, while in the short run lending rate shows
more flexibility relative to deposit rate. In addition, greater rigidity is find in deposit rate decreases
which is in line with customers’ negative reactions hypothesis and greater rigidity in lending rate
decreases which is in line with in collusive pricing arrangements of Turkish banks due to concentrated
structure of the System. Finally, empirical results presented in this paper suggest that retail interest
rates does not adjust asymmetrically to an increase or a decrease in money market rate in Turkey.


References
[1] Akyurek C. and A. M. Kutan (2006), “Can Inflation Targeting Regimes be Effective in
Emerging Economies?: The Turkish Experience” Southern Illionis University, Unpublished
Working Paper, (accessed on 13/06/2007)

[2] Aydin, H. I. (2007), "Interest Rate Pass-Through in Turkey" the CBRT, Research and Monetary
Policy Department Working Paper No:07/05,
[3] BRSA Annual Report of (2006), Banking Regulation and Supervision Agency, 2006.
[4] Borio, C. E. V., & Fritz, W. (1995). The Response of Short-Term Bank Lending Rates to Policy

Rates: A Cross-Country Perspective. In Financial Structure and the Monetary Policy
Transmission Mechanism. Basle: BIS.
[5] Bredin, D., Fitzpatrick, T., & O’Reilly, G. (2001). Real Interest Rate Pass-Through: The Irish
Experience. Technical Paper 06/RT/01, Central Bank of Ireland.
[6] Cottarrelli, C. and Kourelis, A, (1994) “ Financial Structure, Bank Lending Rates, and the
Transmission Mechanism of Monetary Policy”, IMF Staff Paper, December, 41, No. 4 587-623
[7] De Graeve, F., De Jonghe, O. and Vander Vennet, R. (2004), "The Determinants of Pass-
Through of Market Conditions to Bank Retail Interest Rates in Belgium", Ghent University,
Working Paper, 261.
[8] Enders, W. (1995) Applied econometric time series, New York: John Wiley & Sons.
[9] Hannan, T. and A. Berger (1991), “The rigidity of prices. Evidence from the banking industry,
American Economic Review, September, pp 81: 938-945
[10] Hendry, D.F. (1986) Econometric Modeling With Cointegrated Variables: An Overview,
Oxford Bulletin of Economics and Statistics, 48, 201–212.
[11] Hendry, D. F. (1995): Dynamic Econometrics, Oxford University Press, Oxford.
[12] Heinemann, F. and M. Schüler (2002), Integration Benefits on EU Retail Credit Markets -
Evidence from Interest Rate Pass-through, Discussion Paper No. 02-26, Centre for European
Economic Research (ZEW), Mannheim.
[13] Horvath, C., Kreko, J., A. Naszodi, (2004), "Interest Rate Pass-Through: the Case of Hungary",
Magyar Nemzeti Bank, Working Paper, 8.
[14] Klemperer, P. (1987). Markets with consumer switching costs. Quarterly Journal of Economics,
102(2), 375–394.
15
International Research Journal of Finance and Economics - Issue 28 (2009)
[15] Johansen, S. (1988). “Statistical Analysis of Cointegration Vectors” Journal of Economic
Dynamics and Control, Vol: 12, pp. 231-254.
[16] Mc Kinnon, J.G., Haug, A.A., Michelis, L. (1999), “Numerical Distribution Functions of
Likelihood Ratio Tests for Cointegration,” Journal of Applied Econometrics, Vol: 14, pp. 563-
577.
[17] Mojon, B., (2000) “ Financial Structure and the Interest Rate Channel of ECB Monetary

Policy”, November, European Central Bank Working Paper Series, No. 40
[18] Neumark, D. and Sharpe (1992) “ Market Structure and the Nature of Price Rigidity” The
Quarterly Journal of Economics, May 1992
[19] Ozatay, F. and G. Sak. (2002) “The 2000-2001 Financial Crisis In Turkey”, Prepared For
Brookings Trade Forum.
[20] Ozkan, F. G., (2004) “Currency and Financial Crises in Turkey 2000-2001: Bad Fundamentals
or Bad Luck” University Of York, (accessed on 13/03/2006)

[21] Sander, H. and Kleimeier, S. (2002). Asymmetric adjustment of commercial bank interest rates
in the euro area: Implications for monetary policy. Cologne: Mimeo, University of Applied
Sciences Cologne.
[22] Sander, H. and Kleimeier, S. (2004). Convergence in Eurozone retail banking: What interest
rate pass-through tells us about monetary policy transmission, competition and integration.
Journal of International Money and Finance, 23, 461–492.
[23] Scholnick, B.(1996), “Asymmetric Adjustment of Commercial Bank Interest Rates” Evidence
From Malasia and Singapore” Journal of International Money and Finance
[24] Sorensen, C.K. and T. Werner, (2006), "Bank Interest Rate Pass-Through in the Euro Area: A
Cross Country Comparison ", ECB Working Paper, 580.
[25] Stiglitz, J. E., & Weiss, A. (1981). Credit rationing in markets with imperfect information.
American Economic Review, 71(3), 917–926.
[26] Turkish Banking Association Reports