Transcript Click to add Title

DETERMINANTS OF
NON-GOVERNMENT CREDIT IN
ROMANIA
Student: PĂPURICĂ OANA
Supervisor: Professor MOISĂ ALTĂR
Bucharest
July 2007
Contents

Introduction

Overview of the non-government credit in Romania

Literature review

The empirical model and estimation method

Estimation results

Concluding Remarks
2
Goals

On the background of recent significant
growth of the non-government credit in
Romania, this paper attempts to
identify the determinants of credit to
private non-bank sector during 2003:05
and 2006:12, using Johansen
multivariate cointegration analysis and
error correction model.
3
Literature Review
Authors
Dependent
variable
Explanotory variables
Target group
Methodology
Hofmann (2001)
Real loans
Real GDP, real interest rate, housing
prices
16 developed
countries
VECM for individual
countries
Calza et al. (2003)
Real loans
Real GDP growth, nominal lending
rate, inflation rate
Eurozone
VECM on aggregate
eurozone data
Brzoza-Brzezina (2005)
Real loans
Real GDP growth, real interest rate
POR, IRL, GRE,
HUN, CZE, POL
VECM for individual
countries
Cottarelli et al. (2005)
Credit to the
private
sector (%GDP)
Public debt/GDP, PPP-based GDP,
inflation threshold,
liberalization index, index for
entry restrictions to the
banking sector, accounting
standards and legal origin
15 Central European
and Balkan countries,
out of sample
estimation
Random effect panel
estimation of 24
developed and
nontransition
emerging
countries
Boissay et al. (2006)
Credit to the
private
sector (%GDP)
GDP per capita, real interest rate
(Euribor), quadratic trend
11 Central and
Eastern European
countries
ECM for individual
countries and panel
estimation
Kiss et al. (2006)
Credit to the
private
Sector (%GDP)
GDP per capita, real interest rate
inflation rate
Eurozone
Panel estimation
Duenwald et al. (2005)
Credit to the
private
sector (%GDP)
links with trade balance
BLG, ROM, UKR
Panel estimation
Real loans
Real GDP, eal property prices
Gerlach, S. and W. Peng
(2003)
4
Hong Kong
VECM
Determinants of credit for Romania
real non-government credit =
f( economic activity, interest
rate, property prices)
5
The facts - Romania
Non-government
sustained by:
real credit growth
48%, 26%, 33%, 47%
(2003-2006)
• households new loans dynamics
(360% increase/may03=100) even if their share in
total credit is still low ( 31%-june 06, 23%-dec06)
• consumers loans dynamics (3/4 of households loans)
• foreign exchange denominated loans are preferred
(RON appreciation, lower interest rates,
prices expressed in euro)
• Bucharest is the only place that concentrates a
significant percentage of credit (around 40 %); other
counties less than 4 %.
6
Credit  economic activity
Non-government
credit growth
47.3% (2006)
Supported by evolution of :
Credit Demand
Economic conditions  consumption
and investment  demand for credit
Credit Supply
Changes in economic activity 
firms’ CFs and households
incomes  ability to repay debts
 banks extend credit
Positive interaction between credit and economic activity
7
Credit  interest rate
Interest rate has a negative effect both on credit demand and credit supply:
Credit demand
• Interest rates go up
 loans become
more expensive
 credit demand
reduces
Credit supply
• Monetary tightening (increase in interest
rates) deterioration of financial position
of firms and households  reduced
creditworthiness  credit supply reduces
[balance sheet channel]
• Monetary policy tightening (via reduction
of banking system liquidity)  drain
reserves and loanable funds  reduction
of credit supply [bank lending channel]
8
Credit  property prices
Property prices may have a positive effect on both
credit demand and credit supply:
Credit demand
• Changes in property prices  wealth
effect on credit demand
Credit supply
• Increase in property prices
 increases the value of
collateralisable assets 
increases credit worthiness
 banks extend credit
• Construction activity depends
positively on the ratio of property prices
to construction costs  an increase in
property prices  increases
Remark! a potential two-way causality:
construction activity  leading to an
• increases in credit availability  expand
increase in the demand for credit
the demand for a (temporarily) fixed supply
(Tobin’s q-theory of investment)
of properties  property prices increase 9
Data description
Sample: 2003:05 – 2006:12
Frequency: monthly
Variable
Description
cng_sa
Log of real non-government credit, deflated with CPI (index
May 2003=100)
ip_sa
Log of real industrial output index ( May 2003=100)
ir_l
Nominal aggregate lending rate for non-government credit
(monthly adjusted)
pp
Log of real property prices index ( May 2003=100)
ipc
Consumer price index ( May 2003=100)
High seasonality in December month  cr, ip  Tramo/Seats 
seasonally adjusted time series cr_sa, ip_sa.
10
Data description
6.0
4.80
5.8
4.76
5.6
4.72
5.4
4.68
5.2
4.64
5.0
4.60
4.8
4.6
4.56
2003
2004
2005
2006
2003
2004
CNG_SA
2005
2006
IP_SA
1.6
5.6
1.5
5.4
1.4
1.3
5.2
1.2
5.0
1.1
1.0
4.8
0.9
4.6
0.8
0.7
4.4
2003
2004
2005
IR_L
2006
2003
2004
2005
2006
PP
11
Unit Root Tests (ADF)
Symbol Level
First difference
cng_sa
0.200259 (T)
-11.00894 (C)
ip_sa
-0.967589 (C)
-9.359537 (C)
ir
-0.249589 (C)
-11.24778 (C)
pp
-1.728586 (T)
-2.381833 (N)
C
T
1% critical value
-3.59
-4.21
5% critical value
-2.93
-3.52
10% critical value
-2.60
-3.19
12
VAR Lag Order Selection Criteria

The use of the Johansen procedure implies choosing the appropriate
number of lags in VAR. The optimal number of lags in unrestricted VAR
was based on the information criteria and LR test.
VAR Lag Order Selection Criteria
Endogenous variables: CNG_SA IP_SA IR_L PP
Exogenous variables: C
Lag
LogL
LR
FPE
AIC
SC
HQ
0
192.9377
NA
9.28E-10
-9.446884
-9.277996
-9.385820
1
341.3865
259.7853
1.24E-12
-16.06932
-15.22488*
-15.76400
2
364.7585
36.22667*
8.81E-13*
-16.43792*
-14.91793
-15.88834*
3
375.0378
13.87708
1.25E-12
-16.15189
-13.95635
-15.35805
4
390.2043
17.44150
1.49E-12
-16.11022
-13.23912
-15.07212
The optimal number of lags in unrestricted VAR has proven to be 2
(equivalently 1 lagged difference in VEC).
Diagnostics
13
Johansen cointegration analysis
Unrestricted Cointegration Rank Test
Hypothesized
Trace
5 Percent
1 Percent
No. of CE(s)
Eigenvalue
Statistic
Critical Value
Critical Value
None **
0.597069
64.84857
47.21
54.46
At most 1
0.290309
26.67095
29.68
35.65
At most 2
0.145177
12.26807
15.41
20.04
At most 3 *
0.126490
5.679923
3.76
6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level
Trace test indicates 1 cointegrating equation(s) at both 5% and 1% levels
Hypothesized
Max-Eigen
5 Percent
1 Percent
No. of CE(s)
Eigenvalue
Statistic
Critical Value
Critical Value
None **
0.597069
38.17761
27.07
32.24
At most 1
0.290309
14.40288
20.97
25.52
At most 2
0.145177
6.588145
14.07
18.63
At most 3 *
0.126490
5.679923
3.76
6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level
Max-eigenvalue test indicates 1 cointegrating equation(s) at both 5% and 1% levels
15
The long-run relationship
Cointegrating
Eq:
CointEq1
cng_sa = 0.7069*ip_sa – 0.943*ir_l + 0.1593*pp +2.3049
CNG_SA(-1)
1.000000
IP_SA(-1)
-0.706956
 The long-run elasticity of credit with respect to real
industrial production:
• 1 percent point increase in industrial output implies an
increase of 0.7069 percent points in the real credit;
• the null hypothesis that the change of industrial production
is null in respect to the real credit (B(1,2)=0) is rejected (χ2
(1) = 2.96 [0.0849])
• unit output elasticity (B(1,2)=1) is rejected (χ2 (1)=11.69
[0.0006])
The long run semi-elasticity of credit with respect to
interest rate is significantly negative, in concordance with
economic theory (one percentage increase in the interest
rate triggers a long-run reduction in the real lending of 0.943
percent).
The elasticity of credit with respect to property prices is
significant positive (One percentage increase in the property16
prices has a 0.15 percent increase in the real credit)
(0.35167)
[-2.01026]
IR_L(-1)
0.943059
(0.08551)
[ 11.0293]
PP(-1)
-0.159318
(0.07952)
[-2.00339]
C
-2.304933
The long-run relationship
Error Correction:
D(CNG_SA)
D(IP_SA)
D(IR_L)
D(PP)
CointEq1
-0.247214
0.047480
-0.623764
0.017794
(0.13870)
(0.06760)
(0.12485)
(0.04046)
[-1.78239]
[ 0.70242]
[-4.99626]
[ 0.43976]
The coefficient representing the speed of adjustment of real credit indicates:
-relatively rapid adjustment of real credit to the long-run equilibrium;
-if in the previous month the real credit exceeded the long-run level in the
current month real credit would decrease (negative sign);
- the disequilibria accommodates relatively quickly: 25% from the previous
month disequilibrium is adjusted in the current month 4months.
17
The long-run relationship

We computed the series of residuals from the long-run equilibrium relationship
and tested the resulting series for stationarity resid01 I(0); the cointegrating
equation represents indeed a long-run run relationship between the specified
variables.
resid01=1*cng_sa-0.706956*ip_sa+0.943059*ir_l-0.159318*pp-2.30493
.2
.1
.0
-.1
-.2
-.3
2003
2004
2005
C o in te g ra tin g re la tio n 1
2006
These deviations from the long-run equilibrium
are stationary and we are going to use them in
an error correction mechanism. These deviations
try to adjust to the equilibrium at the end of the
period, but there is a decrease of these
deviations in September 2005 when came into
force the restrictive provisions of Norm 10 on
mitigating credit risk for credit granted to
individuals
18
Weak-exogeneity tests
Δ cng_sa
A(1,1)=0
Δ ip_sa
A(2,1)=0
Δ ir_l
A(3,1)=0
Δpp
A(4,1)=0
Δ ip_sa, Δpp
A(2,1)=0,A(4,1)=0
χ2 (1) = 3.1293
[0.0768]
χ2 (1) = 0.4611
[0.4970]
χ2 (1) = 18.787
[0.000015]
χ2 (1) = 0.1699
[0.6801]
χ2 (1) = 0.9470
[0.622811]
Note: The null hypothesis is that there is weak exogeneity (in squared brackets - probability)
 The weak exogeneity hypothesis is accepted both separately and jointly for
industrial output and property prices. It is rejected for the interest rate. The
interest rate is not weak exogenous and it adjusts to the real lending
disequilibria from the long term level.
The hypothesis that industrial production deviation form the equilibrium
level does not adjust to the other variables included in the cointegration
relationship is accepted with a probability of 49.7%.
Weak exogeneity hypothesis of property prices suggests that property prices
are determined outside the system, they are not caused by real credit, but they
determine real credit. So is not the rise in real credit that generates an increase
of property prices but real credit increases in order to reach the equilibrium.
19
Short run error correction model
(ECM)


The deviation of real credit from its long-run
level is stationary, so we can use it in an
error correction mechanism (the residual
series will be used as error correction term in
dynamic model)
D(CNG_SA)= C(1)*D(CNG_SA(-1))+
C(2)*D(IP_SA(-1)) + C(3)*D(IR_L(1))+
C(4)*D(PP(-1))+C(5)* RESID01(-1) +C(6)
20
Short run error correction model
(ECM)

Following the general-to-specific approach, we can obtain a
parsimonious model:
Dependent Variable: D(CNG_SA)
Method: Least Squares
Sample(adjusted): 2003:07 2006:12
D(CNG_SA)= C(1)*D(CNG_SA(-1))+ C(2)*D(IP_SA(-1)) + C(5)* RESID01(-1)
+C(6)
Coefficient
Std. Error
t-Statistic
Prob.
C(1)
-0.335399
0.127261
-2.635529
0.0121
C(2)
-1.017459
0.327678
-3.105055
0.0036
C(5)
-0.216727
0.129821
-1.669434
0.1033
C(6)
0.035997
0.009107
3.952909
0.0003
R-squared
0.428313
Mean dependent var
0.022608
Adjusted R-squared
0.383180
S.D. dependent var
0.069881
S.E. of regression
0.054883
Akaike info criterion
-2.876841
Sum squared resid
0.114461
Schwarz criterion
-2.711348
Log likelihood
64.41365
Durbin-Watson stat
2.263962
21
Short run error correction model
(ECM)



In the short-run one lag changes in interest rates and
property prices are not significant for the current real
credit growth.
The error correction term has a negative sign but is
significant at 10 percent level. This sign suggests that
in the current month real credit adjusts as a result of
previous month disequilibrium from the equilibrium
level.
When credit departs from its long-term trend, the
adjustment towards equilibrium implies not only a
change in credit, but also a change in industrial
production. More specifically, when lending is above
(below) its long-run level, restoring equilibrium is
achieved via reductions (increases) in lending, but also
a contraction (expansion) of industrial output.
22
Concluding remarks




Cointegration analysis reveals that there is a stationary
long run relationship between real non-government
credit, industrial production as a proxy for the
economic activity, nominal interest rate and real
property prices.
An important finding of this paper is that property
prices can be considered a determinant of credit in the
long-run.
Property prices  weak exogenous : are determined
outside the system, they are not caused by real credit,
but they determine real credit.
The coefficient representing the speed of adjustment
of real credit indicates relatively rapid adjustment of
real credit to the long-run equilibrium ( aprox 4
months).
23
Concluding remarks


In the short-run real credit is not influenced by
changes in property prices and interest rates and when
credit departs from its long-term trend, the
adjustment towards equilibrium implies not only a
change in credit, but also a change in industrial
production.
Limitations:
- The estimated elasticities must be used cautiously, as
it is difficult to interpret them as true long-run
elasticities given the short time series available (44
observations).
- The inexistence of an official property price index.
- The use of industrial output as a proxy for economic
activity.
- The monthly series.
24
Further research


For further research, we should consider an
analysis on the components of nongovernment credit and include a bigger
number of determinants (as unemployment
rate for credit to individuals, exchange rate,
consumption, wages, etc).
An important aspect to be considered for the
further research could be the new regulatory
framework starting with 2007 imposed by
Regulation Nr. 3/2007 which comes with
relaxing credit conditions for households.
25
References













Boissay, F., O. Calvo-Gonzalez and T. Kozluk (2006), “Is Lending in Central and Eastern Europe Developing too
Fast? “ Paper presented at the Conference.Finance and Consumption Workshop: Consumption and Credit in
Countries with Developing Credit Markets., Florence, 16-17 June 2006.
Brzoza-Brzezina, M. ( 2005), “Lending booms in Europe.s periphery: South- Western lessons for CentralEastern members”, ECB Working Paper no. 543.
Calza, A., C. Gartner, and J. Sousa (2003), “Modelling the Demand for Loans to the Private Sector in the Euro
Area”, Applied Economics, 35, pp. 107-117.
Cottarelli, C., G. Dell.Ariccia and I. Vladkova-Hollar (2005), “Early birds, late risers and sleeping beauties: Bank
credit growth to the private sector in Central and Eastern Europe and in the Balkans”. Journal of Banking and
Finance. 29, 83-104.
Dickey, D and W Fuller (1981), “Likelihood ratio statistics for autoregressive time series with a unit root”,
Econometrica, 60, pp 423-33.
Duenwald, C., N. Gueorguiev, and A. Schaechter (2005). “Too much of a good thing? Credit booms in transition
economies: The cases of Bulgaria, Romania, and Ukraine.” IMF Working Paper no. 128.
Egert, B. , P. Barke, and T. Zumer (2006), “Credit growth in Central and Eastern Europe. New (Over) Shooting
stars?”, European Central Bank Working Paper No. 687
Engle, R. and C. Granger (1987), “Cointegration and Error-correction: Representation, Estimation and Testing”,
Econometrica, 55, 252-276.
Ericsson, N. R. (1992), “Cointegration, exogeneity, and policy analysis: an overview”, Journal of policy
modeling 14(3): 251-280
Gerlach, S. and W. Peng (2003), “Bank Lending and Property Prices in Hong Kong” HKIMR Working Paper No.
12/2003.
Greene, W. (1993), “Econometric Analysis”, second edition, Macmillan Publishing Company, New York, N.Y.
Hall, S., (2001) “Credit channel effects in the monetary transmission mechanism” Bank of England Quarterly
Bulletin Q4. 442.448.
Hendry, F. D., and K. Juselius (2000), “Explaining Cointegration Analysis: Part II”, Energy Journal 21
26
References

















Hofmann, B (2003), “Bank Lending and Property Prices: Some International Evidence”,The Hong Kong
Institute for Monetary Research Working Paper, No. 22.
(2001), “The determinants of private sector credit in industrialized countries: Do property prices matter?”
BIS Working Paper 108.
Hamilton, J (1994), “Time series analysis”, Princeton University Press, Princeton, New Jersey.
Johansen S., Juselius K., (1990), “Maximum Likelihood Estimation and Inference on Cointegration - with
Applications to Simultaneous Equations and Cointegration”, Journal of Econometrics, 69
Johansen, S (1988), “Statistical analysis of cointegration vectors”, Journal of Economic Dynamics and
Control, 12, pp 231-54.
(1991), “Estimation and hypothesis testing of cointegration vectors in gaussian vector autoregressive
models”, Econometria, 59, pp 1551-81.
Kiss, G., M. Nagy and B. Vonnák (2006), “Credit Growth in Central and Eastern Europe: Trend, Cycle or
Boom?” Paper presented at the conference Finance and Consumption Workshop: Consumption and Credit in
Countries with Developing Credit Markets., Florence, 16-17 June.
MacKinnon, J.G. (1996), “Numerical Distribution Functions for Unit Root and Cointegration Tests,” Journal
of Applied Econometrics, 11, 601-618.
Martinez-Carrascal, C. and A. del Rio (2004), “Households borrowing and consumption in spain: a VECM
approach”, Banca de Espania, Documentos de Trabajo no. 0421
Osterwald-Lenum, M (1992), “A note with quantiles of the asymptotic distribution of the maximum likelihood
cointegration rank test statistics”, Oxford Bulletin of Economics and Statistics, 54, pp 461-72.
Sims, C (1980), “Macroeconomics and reality”, Econometrica, 48, pp 1-48.
Terrones, M. and E. Mendoza (2004), “Are credit booms in emerging markets a concern?” IMF, World
Economic Outlook, Chapter IV. 147-166.
Tobin, J (1969), “A general equilibrium approach to monetary theory”, Journal of Money, Credit and
Banking, 1, pp 15-29.
*** ECB Monthly Bulletin 11/2006
*** NBR Monthly Bulletin 2003-2006
*** NBR Annual Reports 2003-2006
27
*** NBR Financial Stability Report 2006-2007

Thank you!
28