Estimating Credit Demand in Croatia

Katja Gattin-Turkalj, Igor Ljubaj,Ana Martinis, Marko Mrkalj

Overview of the Presentation

1. Credit developments in Croatia

2. The econometric model

Total loans to the private sector

Household loans

3. Concluding remarks

Motivation

GDP growth increases demand for loans via income effect and via wealth effect

improving overall economic conditions, growing optimism by consumers and enterprises and

sharp decline of interest rates, real convergence…

Transition of the financial sector

towards the end of 1990’s …

predominantly supply side effects

demand factors gained greater

importance

Three phases of banks’ loan developments

and the CNB responses

-10

0

10

20

30

40

50

1/9

7

1/9

8

1/9

9

1/0

0

1/0

1

1/0

2

1/0

3

1/0

4

1/0

5

1/0

6

1/0

7

%

Total Loans (yoy growth rate)

"16%"

"12%"

MR fx liquidity (35-32%)

MRR

24%

55%

Decline of interest rates contributed to credit growth

0

5

10

15

20

251

/97

8/9

7

3/9

8

10

/98

5/9

9

12

/99

7/0

0

2/0

1

9/0

1

4/0

2

11

/02

6/0

3

1/0

4

8/0

4

3/0

5

10

/05

5/0

6

12

/06

%

Short-term corporate loans without a currency clause

Short-term household loans without a currency clause

Long-term corporate loans with a currency clause

Long-term household loans with a currency clause

Loans to households driven by housing loans…

0

5

10

15

20

25

30

35

40

453/

05

5/05

7/05

9/05

11/0

5

1/06

3/06

5/06

7/06

9/06

11/0

6

1/07

%

Total credits to households Housing loansCredits for motor vehicles Credits on credit cardOther credits to households

Increasing role of housing loans highlights the risks associated with

house price developments

90

100

110

120

130

140

150

H1

/97

H2

/97

H1

/98

H2

/98

H1

/99

H2

/99

H1

/00

H2

/00

H1

/01

H2

/01

H1

/02

H2

/02

H1

/03

H2

/03

H1

/04

H2

/04

H1

/05

H2

/05

H1

/06

50

150

250

350

450

550

House price index - left Housing loans - right

2. Loans to the private sector - the model

Loans and GDP (gaps) exhibit a high degree of correlation

-3

-2

-1

0

1

2

3

4

5

Q1

/97

Q3

/97

Q1

/98

Q3

/98

Q1

/99

Q3

/99

Q1

/00

Q3

/00

Q1

/01

Q3

/01

Q1

/02

Q3

/02

Q1

/03

Q3

/03

Q1

/04

Q3

/04

Q1

/05

Q3

/05

Q1

/06

Q3

/06

Q1

/07

GDP_GAP LOANS_GAP

),( irgdpfloans

ADF and PP tests

Variable

GDP 0,046 -2,091 0,27 -1,896

d(GDP) -8,317*** -8,348*** -8,317*** -8,434***

IR -2,025 -1,510 -2,897* -1,402

d(IR) -7,161*** -7,521*** -7,161*** -7,455***

LOANS 1,835 -4,908*** -0,297 -1,888

d(LOANS) -2,714* -4,705*** -2,815* -2,763

*denotes rejection of unit root null hypothesis at 10% significance level, ** at 5% significance level,

*** at 1 % significance level

ADF PP

ConstantConstant and

trendConstant

Constant and trend

The model – baseline equation

itt irYloans 210 loglog -0.06 3.09 -

8.75 (-1.29) ***(5.05)

***(-3.52)

GDP coefficient is in line with other studies

Baseline result of GDP elasticity of 3.09

Calza et al. (2001): 1.34 (eurozone) Calza et al. (2003): 1.60 (eurozone) Hofmann (2001): between 1.04 and 2.49 (16 industrial countries) Hülsewig et al. (2001): 1.11 (Germany) Brzoza-Brzezina (2005): between 1.45 and 3.39 (six European countries)

IR coefficients seem to vary more…

Baseline result of real interest rate elasticity of -8.75

Calza et al. (2001): -1.01 (eurozone) Calza et al. (2003): -5.05 (eurozone) Hofmann (2001): between -0.01 and -0.08(16 industrial countries) Hülsewig et al. (2001): -0.69 (Germany) Brzoza-Brzezina (2005): between -4.42 and -10.81 (six European countries)

Bootstrapping… method

11101 xy

22102 xy

33103 xy

44104 xy

............................

11101 nnn xy

nnn xy 10

1101 xy 2

2102 xy 1n

3103 xy 2

4104 xy 3

............................... 1101 nn xy 1

nn xy 10 n

Bootstrapping… results for β1

Empirical Distribution of Coefficient Beta1

0

50

100

150

200

250

300

350

400

450

500

Coefficient beta1

Nu

mb

er

of

ob

se

rva

tio

ns

3.09

p<0,05

mean = 3.06

p<0,05

Bootstrapping… results for β2

Empirical Distribution of Coefficient Beta2

0

50

100

150

200

250

300

350

-2,3

8-2

,03

-1,88

-1,72

-1,58

-1,4

2-1

,28-1

,13-9

,73-8

,25

-6,76

-5,25

-3,76

-2,26 0,

000,

002,

213,

716,

95

Coefficient beta2

Nu

mb

er

of

obse

rvat

ions

- 8.75

p<0.05

m ean = - 8.28

p<0.05

Extensions of baseline specification

Eq. 1 Eq. 2 Eq. 3 Eq. 4 Eq. 5 Eq. 6 Eq. 7 Eq. 8 Eq. 9 Eq. 10 Eq. 11 Eq. 12

(t-stat)Constant -0,06 -0,03 -0,07 -0,02 -0.11 *** -0,05 ** -0,09 -0.08 *** -0,04 0,00 -0,05 -0,01

(-1.29) (-1.97) (-1.56) (-1.33) (-2.10) (-2.17) (-1.46) (-4.33) (-0.92) (-0.32) (-1.07) (-0.43)Loans(-1) 0,76 *** 0,79 *** 0,75 *** 0,83 *** 0,79 *** 0,81 ***

(-13.05) (12.92) (12.7) (17.8) -20 (18.82)GDP 3,09*** 1.3 *** 3,08 *** 1,15 *** 3,34 *** 1,41 *** 3.20 *** 1,26 *** 2,75 *** 0,48 *** 2,95 *** 0,49 *

(5.05) (4.77) (5.24) (3.95) (5.49) (4.80) (4.98) (5.74) (3.60) (2.12) (3.84) (1.99)Interest rate -8,75*** -1.79 -7.59 ***-1.89* -8.36 *** -1,84 -7.81 *** -2.64 *** -8.91 *** -1.82 *** -8.83 ** -1.74 **

(-3.52) (-1.64) (-3.08) (-1.75) (-3.46) (-1.69) (-3.08) (-3.20) (-3.55) (-2.48) (-3.48) (-2.22)Trend 0,06 * -0,03 0,04 -0.15 ***

(1.90) (-1.32) (0.75) (-4.96)CPI 1.27* 0,32 0,57 2,02 ***

(1.80) (1.01) (0.48) (4.85)Exchange -0,51 -1.09 ***rate kuna/USD (-0.75) (-6.11)Exchange -0,22 -1.04 ***rate kuna/euro (-0.32) (-5.38)Number of obs. 35 34 35 34 35 34 35 34 35 34 35 34Adjusted R2

0,43 0,91 0,47 0,91 0,47 0,91 0,46 0,95 0,42 0,96 0,47 0,96AIC -2,07 -3,94 -2,12 -3,93 -2,11 -3,91 -2,07 -4,49 -2,03 -4,71 -2,01 -4,57Schwarz -1,93 -3,76 -1,94 -3,71 -1,93 -3,69 -1,85 -4,22 -1,85 -4,48 1,83 -4,35F-Statistic 13,88 ##### 11,20 83,39 10,97 81,62 8,25 ##### 9,32 ##### 9,03 #####Probabilty 0,00 0,00 0,00 0,00 0,00 0,00 0,00 0,00 0,00 0,00 0,00 0,00

Equations Variables

Standardized residuals of equations 1 through 12

-.2

-.1

.0

.1

.2

.3

97 98 99 00 01 02 03 04 05 06

Eq.1 : GDP IR

-.08

-.06

-.04

-.02

.00

.02

.04

.06

97 98 99 00 01 02 03 04 05 06

Eq.2 : LOANS (-1) GDP IR

-.2

-.1

.0

.1

.2

97 98 99 00 01 02 03 04 05 06

Eq.3: GDP IR Z

-.08

-.06

-.04

-.02

.00

.02

.04

.06

97 98 99 00 01 02 03 04 05 06

Eq. 4: LOANS(-1) GDP IR Z

-.20

-.15

-.10

-.05

.00

.05

.10

.15

.20

97 98 99 00 01 02 03 04 05 06

Eq. 5: GDP IR CPI

-.06

-.04

-.02

.00

.02

.04

.06

97 98 99 00 01 02 03 04 05 06

Eq. 6: LOANS(-1) GDP IR CPI

-.2

-.1

.0

.1

.2

97 98 99 00 01 02 03 04 05 06

Eq. 7: GDP Z CPI

-.04

-.03

-.02

-.01

.00

.01

.02

.03

.04

.05

97 98 99 00 01 02 03 04 05 06

Eq.8: LOANS(-1) GDP IR Z CPI

-.2

-.1

.0

.1

.2

.3

97 98 99 00 01 02 03 04 05 06

Eq. 9: GDP IR USD

-.06

-.04

-.02

.00

.02

.04

.06

97 98 99 00 01 02 03 04 05 06

Eq.10: LOANS(-1) GDP IR USD

-.2

-.1

.0

.1

.2

.3

97 98 99 00 01 02 03 04 05 06

Eq.11: GDP IR EUR

-.06

-.04

-.02

.00

.02

.04

.06

97 98 99 00 01 02 03 04 05 06

Eq. 12: LOANS(-1) GDP IR EUR

Standardized residuals of equations 1 through 4

-.2

-.1

.0

.1

.2

.3

97 98 99 00 01 02 03 04 05 06

Eq.1 : GDP IR

-.08

-.06

-.04

-.02

.00

.02

.04

.06

97 98 99 00 01 02 03 04 05 06

Eq.2 : LOANS (-1) GDP IR

-.2

-.1

.0

.1

.2

97 98 99 00 01 02 03 04 05 06

Eq.3: GDP IR Z

-.08

-.06

-.04

-.02

.00

.02

.04

.06

97 98 99 00 01 02 03 04 05 06

Eq. 4: LOANS(-1) GDP IR Z

-.20

-.15

-.10

-.05

.00

.05

.10

.15

.20

97 98 99 00 01 02 03 04 05 06

Eq. 5: GDP IR CPI

-.06

-.04

-.02

.00

.02

.04

.06

97 98 99 00 01 02 03 04 05 06

Eq. 6: LOANS(-1) GDP IR CPI

-.2

-.1

.0

.1

.2

97 98 99 00 01 02 03 04 05 06

Eq. 7: GDP Z CPI

-.04

-.03

-.02

-.01

.00

.01

.02

.03

.04

.05

97 98 99 00 01 02 03 04 05 06

Eq.8: LOANS(-1) GDP IR Z CPI

-.2

-.1

.0

.1

.2

.3

97 98 99 00 01 02 03 04 05 06

Eq. 9: GDP IR USD

-.06

-.04

-.02

.00

.02

.04

.06

97 98 99 00 01 02 03 04 05 06

Eq.10: LOANS(-1) GDP IR USD

-.2

-.1

.0

.1

.2

.3

97 98 99 00 01 02 03 04 05 06

Eq.11: GDP IR EUR

-.06

-.04

-.02

.00

.02

.04

.06

97 98 99 00 01 02 03 04 05 06

Eq. 12: LOANS(-1) GDP IR EUR

Extensions of baseline spec.: Eq. 1 to 6

Eq. 1

(t-stat)

Constant -0.06

(-1.29)

Loans(-1)

(-13.05)

GDP 3.09 ***

(5.05)

Interest rate -8.75 ***

(-3.52)

Trend

CPI

Eqations Variables

Eq. 5

-0.11 ***

(-2.10)

3.34 ***

(5.49)

-8.36 ***

(-3.46)

1.27*

(1.80)

Eq. 6

-0.05 **

(-2.17)

0.75 ***

(12.7)

1.41 ***

(4.80)

-1.84

(-1.69)

0.32

(1.01)

Eq. 2

-0.03

(-1.97)

0.76 ***

(-13.05)

1.3 ***

(4.77)

-1.79

(-1.64)

Standard specification tests

Autocorrelation: introducing persistence through lagged dependent variable, improves the fit, although residuals remain "noisy"

Variance inflation factor=1.3 ΔlogYt = β0 + β1Δirt + εi

1/(1 – R2) Test for omitted variables was done for all variables other

than baseline regressors to test the explanatory power of the additional

loanst-1 (Ho: not omitted rejected at 1%), the trend variable z (Ho rejected at 5%), CPI (Ho rejected at 10% significance), and exchange rate (Ho not rejected)

AIC and Schwartz criterion, systematically favor more parsimonious specifications

Recursive coefficients of equation loans =-0.06 + 3.09*gdp+(-8.75)*ir

-.8

-.6

-.4

-.2

.0

.2

.4

.6

2000 2001 2002 2003 2004 2005 2006

Recursive C(1) Estimates± 2 S.E.

-2

0

2

4

6

8

10

2000 2001 2002 2003 2004 2005 2006

Recursive C(2) Estimates± 2 S.E.

-50

-40

-30

-20

-10

0

10

20

2000 2001 2002 2003 2004 2005 2006

Recursive C(3) Estimates± 2 S.E.

Results:

the baseline specification seem to satisfactory explain the observed developments of credit;

the extension of the baseline equation, did not significantly change the results

credit growth during the lending boom of the late 1990's and in 2006 remains above the fitted line in all specifications and during the recession that followed the first lending boom credit plunged well below the fitted line;

inclusion of the lagged dependent variable "smoothes" the curve and improves the fit, but even then actual growth remains slightly above the fitted line until 1998Q1 and in 2006.

Results:Loans (yearly growth rates) - baseline

specification

-.2

-.1

.0

.1

.2

.3

-.1

.0

.1

.2

.3

.4

98 99 00 01 02 03 04 05 06

Residual Actual Fitted

3. Household loans - the model

Correlation between household loans and explanatory variables

Interest rate_H Consumption Wages House prices Loans(-1)

4log -0.48 0.32 -0.44 -0.13  0.92

log -0.98 0.94 0.96 0.86 0.99Loans to

households

Wages seem to move in the opposite direction from loans…

Wage and households loans gaps (normalized)

-4

-3

-2

-1

0

1

2

3

Q1

/97

Q4

/97

Q3

/98

Q2

/99

Q1

/00

Q4

/00

Q3

/01

Q2

/02

Q1

/03

Q4

/03

Q3

/04

Q2

/05

Q1

/06

Q4

/06

Loans to Households Wages

…as well as house price index - possibly due to low data frequency

YoY growth rates of house prices and household loans (normalized)

-3

-2

-1

0

1

2

3

4

Q1

/98

Q4

/98

Q3

/99

Q2

/00

Q1

/01

Q4

/01

Q3

/02

Q2

/03

Q1

/04

Q4

/04

Q3

/05

Q2

/06

House price index Loans to households

Equations for household loansEq. 1_h Eq. 2_h Eq. 3_h Eq. 4_h Eq. 5_h Eq. 6_h Eq. 7_h Eq. 8_h

(t-stat)

0.15 *** 0.00 0.19 *** 0.00 0.16 *** 0.06 *** 0.25 *** 0.07 ***

(3.79) (-0.02) (4.38) (-0.29) (2.91) (3.82) (3.90) (3.31)

Loans to 0.83 *** 0.83 *** 0.78 *** 0.77 ***

households (-1) (21.68) (20.42) -17.89 (16.55)Interest rates on -11.11 *** -0.46 -7.67 *** -0.51 -10.52 *** -1.88 * -6.67 * -1.70

loans to households (-2.87) (-0.49) (-2.13) (0.59) (-2.87) (-1.79) (-1.95) (-1.54)

-0.21 0.87 *** 0.06 0.87 ***

(-0.35) (6.12) (0.11) -5.90

-0.20 -0.90 *** -0.93 -0.99 ***

(-0.25) (-4.15) (-1.20) (-4.08)

-0.58 0.06 -0.78 -0.11(-1.28) (0.48) (-1.66) (-0.74)

Number of obs. 35 34 35 34 35 34 35 34

Adjusted R20,16 0.95 0,15 0.94 0,16 0,92 0,19 0,92

AIC -1.32 -4.33 -1.53 -4.24 -1.35 -3.99 -1.58 -3.9

Schwarz -1.19 -4.15 -1.35 -4.01 -1.22 -3.82 1,40 -3.67

F-Statistic 4.32 194.00 2.84 136.89 4.41 135.78 3.45 95.45

Probabilty 0.02 0.00 0.06 0.00 0.02 0.00 0.03 0.00

Equations Variables

House price index

Bas

elin

e re

gres

sors

Constant

Consumption

Wages

Concluding remarks

the behavior of loans can be explained mainly by the developments of real GDP and real interest rates

GDP captures most important forces behind the loan demand

somewhat unexpectedly, house price index did not contribute to explaining household credit demand

Q?