Credit Growth and Bank Soundness in Emerging Europe Natalia Tamirisa Deniz Igan International Monetary Fund The 13 th Dubrovnik Economic Conference June

Credit Growth and Credit Growth and Bank Soundness in Bank Soundness in Emerging EuropeEmerging Europe

Natalia TamirisaNatalia TamirisaDeniz IganDeniz Igan

International Monetary FundInternational Monetary Fund

The 13th Dubrovnik Economic Conference

June 29, 2007

Rapid credit growth in the Rapid credit growth in the region...region...

Sources: Eurostat; IFS, National Statistical Offices; and IMF staff estimates.

-5

0

5

10

15

20

25

30

35

40

45

50

55

60

Ger

man

yA

ustr

iaM

alta

Ital

yF

ranc

eF

inla

ndP

ortu

gal

Luxe

mbo

Hun

gary

Cze

chC

ypru

sB

elgi

umB

ulga

riaG

reec

eIr

elan

dS

lova

kS

pain

Slo

veni

aP

olan

dR

oman

iaLi

thua

nia

Latv

iaE

ston

ia

-5

0

5

10

15

20

25

30

35

40

45

50

55

60

Real Credit to the Private Sector, 2006(annual percentage change)

GDP per Capita and Current Account Deficit

Austria

Belgium

Finland

France

Germany

Greece

IrelandItaly

Portugal

Slovenia

Spain

Bulgaria

CyprusCzech

Republic

Estonia

Hungary

Latvia

Lithuania

Malta

Poland

Romania

Slovak Republic

-15

-10

-5

0

5

10

0 10000 20000 30000 40000 50000

GDP per Capita, PPP-based 2006

Cur

rent

acc

oun

t def

icit,

200

2-06

-15

-10

-5

0

5

10

...increasingly funded through ...increasingly funded through capital inflowscapital inflows

NMS: Net Capital Flows, 1999-2006(In percent of GDP)

Baltics

-2.0

0.0

2.0

4.0

6.0

8.0

10.0

12.0

1999 2001 2003 2005

-2.0

0.0

2.0

4.0

6.0

8.0

10.0

12.0CEECs

-2.0

0.0

2.0

4.0

6.0

8.0

10.0

12.0

1999 2001 2003 2005

-2.0

0.0

2.0

4.0

6.0

8.0

10.0

12.0PortfolioInvestmentOtherInvestmentFDI

Bulgaria and Romania

-2.0

0.0

2.0

4.0

6.0

8.0

10.0

12.0

1999 2001 2003 2005

-2.0

0.0

2.0

4.0

6.0

8.0

10.0

12.0

Source: IMF, World Economic Outlook.

...concentrated in the ...concentrated in the household sectorhousehold sector

0

10

20

30

40

50

60

70

80

2001

2005

Household Loans(In percent of total outstanding loansto the private sector)

Rising euroisation of Rising euroisation of domestic creditdomestic credit

0

10

20

30

40

50

60

70

80

902001

2005

Foreign Currency Loans(In percent of total outstanding loansto the private sector)

Rapid credit growth reflects financial

deepening...Financial Deepening in Selected Countries, 2001-05

Latvia

Estonia

Lithuania

Australia New Zealand

Mexico

Chile Thailand

Malaysia

Rep. of Korea

Japan

Euro area Slovak Rep.

Slovenia

Czech Rep.

Poland

Hungary

-10

-5

0

5

10

15

20

25

30

35

40

0 20 40 60 80 100 120 140 160 180

Private Sector Credit (In percent of GDP)

Dif

fere

nce

betw

een

Pri

vate

Sec

tor

Cre

dit G

row

th a

nd G

DP

G

row

th (

In p

erce

nt)

Sources: National Banks, International Financial Statistics, and IMF staff estimates.

...and rising financial ...and rising financial integrationintegration

External Assets and Liabilities(In percent of GDP)

0

100

200

300

400

500

600

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

Source: Lane and Milesi-Ferretti (2006).

Western Europe

Emerging Europe

Emerging Markets

NMS: Macroeconomic Environment and Credit to the Private Sector, 1994-2005(Annual percent change, unless indicated otherwise)

Sources: IMF International Financial Statistics, and staff estimates.1/ CPI-based index with 2000 as base year.

Real Credit Growth

-30

-20

-10

0

10

20

30

40

50

60

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

-30

-20

-10

0

10

20

30

40

50

60

Baltics CEECs

Real GDP Growth

-4

-2

0

2

4

6

8

10

12

1994 1996 1998 2000 2002 2004

-4

-2

0

2

4

6

8

10

12

Consumer Price Index

0

10

20

30

40

50

1994 1996 1998 2000 2002 2004

0

10

20

30

40

50

Real Lending Rate(In percent per annum)

-6

-4

-2

0

2

4

6

8

10

12

14

1994 1996 1998 2000 2002 2004

-6

-4

-2

0

2

4

6

8

10

12

14

Real Effective Exchange Rate 1/

70

80

90

100

110

120

130

1994 1996 1998 2000 2002 2004

70

80

90

100

110

120

130

Macroeconomic and financial conditions have been supportive...• Disinflation

• Improved economic prospects

• EU accession

• Pent-up demand for credit

• Easy global monetary conditions

• Ample global liquidity

...as are supply-side ...as are supply-side factorsfactors

Sources: European Central Bank; and IMF staff estimates.

Share of Five Largest Credit Institutions(In percent of total assets)

0

20

40

60

80

100

120

CzechRepublic

Hungary Poland SlovakRepublic

Slovenia Estonia Latvia Lithuania

Share of Foreign-Owned Banks(In percent of total assets)

0

20

40

60

80

100

120

CzechRepublic

Hungary Poland SlovakRepublic

Slovenia Estonia Latvia Lithuania

• Privatization

• Entry of foreign banks

• Strategic expansion

• High profitability

+ subsidies and tax + subsidies and tax policiespolicies

LiteratureLiterature Financial deepening but “excesses,” credit Financial deepening but “excesses,” credit

booms are a riskbooms are a risk Schadler et al (2004); Cottarelli et al (2005); Egert Schadler et al (2004); Cottarelli et al (2005); Egert

(2007); ECB (2007)(2007); ECB (2007)

Credit growth improves bank soundness, Credit growth improves bank soundness, unless it is “excessive” unless it is “excessive”

Maechler, Mitra, and Worrell (2006) Maechler, Mitra, and Worrell (2006)

FSIs are favorable, but backward lookingFSIs are favorable, but backward looking Hilbers et al (2005); Iossifov and Khamis (2006)Hilbers et al (2005); Iossifov and Khamis (2006)

Foreign banks are more efficient, but loan growth is similar

Aydin (2006); de Haas and van Lelyveld (2005)

Policy Debate

How to manage How to manage macroeconomic and prudential macroeconomic and prudential risks...risks...

...and “not to kill the goose that ...and “not to kill the goose that

lays the golden eggs”?lays the golden eggs”?

Focus of This StudyFocus of This Study

How significant are prudential risks in How significant are prudential risks in the NMS?the NMS?

Has credit growth affected bank soundness? Has credit growth affected bank soundness? Are weaker banks expanding faster? Are weaker banks expanding faster?

Do prudential risks differ across...?Do prudential risks differ across...? CountriesCountries Banks (foreign/domestically owned)Banks (foreign/domestically owned) Purpose of credit (household/corporate)Purpose of credit (household/corporate) Currency of denomination/indexation (foreign/ Currency of denomination/indexation (foreign/

domestic)domestic)

Bank-level AnalysisBank-level Analysis

Bank balance sheet data (Bankscope)Bank balance sheet data (Bankscope) Ugo Panizza’s (IDB) data set, updated Ugo Panizza’s (IDB) data set, updated 217 banks during 1995-2004 in 8 NMS217 banks during 1995-2004 in 8 NMS 7 observations per bank, on average7 observations per bank, on average Unconsolidated data, where availableUnconsolidated data, where available Commercial banks and leasing companiesCommercial banks and leasing companies

Breakdowns of loans by currency and Breakdowns of loans by currency and purpose (supervisory data)purpose (supervisory data)

6 NMS (except Hungary and Latvia)6 NMS (except Hungary and Latvia)

The sample covers most NMS The sample covers most NMS banks...banks...

Total Bankscope Number Assets

Czech Republic 35 26 74.3 97.6Hungary 36 23 63.9 81.7Poland 60 33 55.0 85.6Slovak Republic 21 20 95.2 83.1Slovenia 22 18 81.8 79.9Estonia 6 5 83.3 94.1Latvia 22 21 95.5 93.2Lithuania 13 9 69.2 93.7

Sources: European Central Bank; Bankscope; and IMF staff estimates.1/ In percent of the total number of banks and total bank assets, respectively.

7.87.98.06.2

7.28.37.67.1

Average Number of Observations per Bank

Proportion of Banks Included in the Sample 1/Number of Banks

Distance to default—a proxy Distance to default—a proxy for insolvency risk for insolvency risk

The number of STD a return realization has to fall The number of STD a return realization has to fall for equity to be exhausted~probability of defaultfor equity to be exhausted~probability of default

DD DD ≡(equity capital+average return)/STD of ≡(equity capital+average return)/STD of return, return,

Bank account dataBank account data

STD deviation for the entire sample periodSTD deviation for the entire sample period

Robustness to alternative ways of measuring Robustness to alternative ways of measuring volatility of returns; NPL ratios; loan loss volatility of returns; NPL ratios; loan loss reservesreserves

Uniformly higher credit growth; Uniformly higher credit growth; stronger, but more heterogeneous stronger, but more heterogeneous

Baltic banksBaltic banks

Variable MeanStandard deviation Mean

Standard deviation Mean

Standard deviation Mean

Standard deviation

Bank credit growth 17.9 40.1 28.7 56.6 27.3 32.7 46.8 43.8Distance to default 14.0 12.5 7.7 9.2 14.8 13.0 12.5 15.3Net interest margin 4.5 2.6 6.1 2.5 3.6 3.1 3.3 1.3Cost-to-income ratio 67.4 99.7 95.5 107.8 71.9 31.8 69.6 19.2Liquidity ratio 17.4 16.1 11.2 9.8 17.2 18.0 17.1 18.0Bank size 6.4 1.3 4.8 1.3 7.0 1.3 5.8 1.3Real GDP growth 2.9 2.4 5.3 3.5 3.3 1.9 8.1 1.2GDP per capita 58.1 23.5 30.9 3.9 70.1 25.7 45.8 10.6Real interest rate 3.2 3.5 -0.5 4.5 2.5 3.7 0.5 1.9Real depreciation 0.2 0.3 -0.1 0.8 -0.4 0.3 -0.5 0.7Foreign ownership 36.2 44.4 31.1 39.7 52.2 46.3 41.1 42.8Public ownership 15.3 33.7 12.5 29.2 6.1 21.5 3.7 15.0

Source: IMF staff estimates.

1995-2000 2001-2004

CEECs BalticsCEECs Baltics

Baseline empirical specification controls Baseline empirical specification controls for macroeconomic and bank-specific for macroeconomic and bank-specific

variables...variables...Equation 1: Bank Credit Growth

),,,

,,,,

,,,,,(

1,

1,1,1,1,

1,1,1,1,1,

ijtijttij

tijtijtijtij

tjtjtjtjtijijt

PublicForeignSize

Liquidity MargintInteresmeCostToIncoefaultDDistanceTo

RERRIRGDPgrowthtaGDPperCapiGrowthBankCreditfGrowthBankCredit

Equation 2: Distance to Default

),,,

,,,,

,,,,,(

1,

1,1,1,1,

1,1,1,1,1,

ijtijttij

tijtijtijtij

tjtjtjtjtijijt

PublicForeignSize

Liquidity rginInterestMameCostToIncoefaultDDistanceTo

RERRIRGDPgrowthtaGDPperCapiGrowthBankCreditfDefaultDistanceTo

where i denotes individual banks, j denotes countries, and t is the year index. BankCreditGrowth is the annual percent change in real bank credit to the private sector. RIR is the real interest rate and ΔRER is the annual percent change in the real exchange rate. CostToIncome and InterestMargin stand for the cost-to-income ratio and the net interest margin. Public and Foreign are measures of public and foreign ownership.

Three-stage Least Three-stage Least SquaresSquares

Commonly used techniqueCommonly used technique Linear models using panel dataLinear models using panel data A relatively short time dimension A relatively short time dimension Lags of dependent variablesLags of dependent variables

Advantages vis-à-vis Arellano-BondAdvantages vis-à-vis Arellano-Bond Two-equation estimationTwo-equation estimation Subsample analysisSubsample analysis

Advantages vis-à-vis 2SLSAdvantages vis-à-vis 2SLS Efficiency gainsEfficiency gains Unbiased in models with lagged dependent variablesUnbiased in models with lagged dependent variables

No apparent specification problemsNo apparent specification problems Unit roots rejectedUnit roots rejected Hausman specification tests inconclusiveHausman specification tests inconclusive Residual analysis validates inclusion of lagged dependent Residual analysis validates inclusion of lagged dependent

variablevariable No multicollinearityNo multicollinearity

Robustness to single-equation estimationRobustness to single-equation estimation

Bank credit growth (lagged) 0.096*** 0.100*** 0.095***[5.83] [3.89] [4.71]

Distance to default (lagged) 0.229** 0.350* 0.147[2.16] [1.94] [1.20]

Real GDP growth (lagged) 2.646*** 2.415*** 2.475***[5.53] [2.92] [4.38]

GDP per capita (lagged) -0.116** -0.301*** -0.057[1.99] [3.19] [0.73]

Net interest margin (lagged) 0.689 1.757** 1.200**[1.47] [2.25] [2.00]

Cost-to-income ratio (lagged) -0.017 -0.037** 0.046[1.13] [1.96] [1.49]

Real interest rate (lagged) -0.558* -0.864 -0.999**[1.65] [1.58] [2.24]

Real depreciation (lagged) -4.911* 14.750** -7.414***[1.95] [2.45] [2.65]

Public ownership -0.178*** -0.153** -0.067[3.73] [2.39] [0.89]

Constant 16.366*** 15.992** 12.721*[3.37] [2.17] [1.87]

R -squared 0.13 0.16 0.15Observations 881 424 457


Baseline Specification: Bank Credit Growth Equation

1995-2004 1995-2000 2001-2004

Notes: Absolute value of z statistics in brackets; * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. The dependent variable in the first (reported) equation is annual percent change in outstanding loans. In the second (unreported) equation, the dependent variable is distance to default.

NMS: Decomposition of Predicted Credit Growth, 1997-2004(In percent per year)


1997-2000

-30

-20

-10

0

10

20

30

40

502001-2004

-30

-20

-10

0

10

20

30

40

50

Bank credit growth Distance to default Real GDP growth GDP per capitaNet interest margin Cost-to-income ratio Real interest rate Real depreciation Public ownership ExplainedActual

Credit growth in the NMS has Credit growth in the NMS has been largely demand-driven...been largely demand-driven...

Bank credit growth (lagged) -0.002 -0.002 -0.001[1.14] [0.76] [0.54]

Distance to default (lagged) 0.896*** 0.854*** 0.927***[85.84] [59.85] [62.15]

GDP per capita (lagged) 0.017*** 0.029*** 0.007[2.83] [3.90] [0.77]

Liquidity ratio (lagged) 0.020*** 0.013 0.027**[2.67] [1.17] [2.55]

Bank size (lagged) 0.311*** 0.240** 0.324**[3.33] [2.07] [2.22]

Foreign ownership 0.008*** 0.012*** 0.003[2.80] [3.28] [0.69]

Constant -2.668*** -2.660*** -2.252**[4.10] [3.18] [2.25]



Baseline Specification: Distance to Default Equation

1995-2004 1995-2000 2001-2004

Notes: Absolute value of zstatistics in brackets; * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. The dependent variable in the first (unreported) equation is annual percent change in outstanding loans. In the second (reported) equation, the dependent variable is distance to default.

Is the glass half empty or Is the glass half empty or half full?half full?

No evidence that credit growth has No evidence that credit growth has weakened banksweakened banks

Consistent with FSI analysisConsistent with FSI analysis Not surprising in an upward stage of the credit Not surprising in an upward stage of the credit

cyclecycle

During 2001-04 weaker banks started to During 2001-04 weaker banks started to expand just as fast as sounder banksexpand just as fast as sounder banks

New result, not detectable in aggregate dataNew result, not detectable in aggregate data Some weaker banks are weak in the absolute Some weaker banks are weak in the absolute

sensesense

These results are robust These results are robust to...to...

Including additional macro and bank-level Including additional macro and bank-level variablesvariables

Controlling for year- and country-specific Controlling for year- and country-specific factorsfactors

Using alternative measures of bank Using alternative measures of bank ownershipownership

Using alternative measures of bank Using alternative measures of bank soundness soundness

Controlling for nonlinear effectsControlling for nonlinear effects Assuming faster feedback effectsAssuming faster feedback effects Single equation estimationSingle equation estimation

Bank credit growth (lagged) 0.075 0.089 -0.011[1.43] [1.28] [0.17]

Nonperforming loans (lagged) -0.006 -0.025 0.262***[0.14] [0.47] [2.75]

Real GDP growth (lagged) 2.625*** 3.129** 2.109**[2.58] [2.15] [2.23]

GDP per capita (lagged) -0.280** -0.386** -0.219*[2.16] [2.02] [1.86]

Net interest margin (lagged) 2.097* 2.506 5.397***[1.82] [1.59] [3.04]

Cost-to-income ratio (lagged) -0.085*** -0.099*** 0.099[3.22] [3.09] [1.48]

Real interest rate (lagged) -1.332** -1.674** -1.245*[2.33] [2.00] [1.84]

Real depreciation (lagged) 3.792 13.587 -10.722**[0.68] [1.52] [2.26]

Public ownership -0.204*** -0.198* -0.119[2.63] [1.92] [1.12]

Constant 28.265** 28.711* 0.711[2.55] [1.73] [0.05]



Robustness Analysis: Using a Narrower Measure of Bank Soundness—Nonperforming Loan Ratio

1995-2004 1995-2000 2001-2004

Notes: Absolute value of z statistics in brackets; * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. The dependent variable in the first (reported) equation is annual percent change in outstanding loans. In the second (unreported) equation, the dependent variable is the ratio of nonperforming loans to total loans.


Loan loss reserves (lagged) 11.592*** 11.837*** 59.944*[4.12] [3.83] [1.76]

Real GDP growth (lagged) 2.808*** 1.970** 2.299***[4.86] [2.07] [3.28]

GDP per capita (lagged) -0.085 -0.235** -0.002[1.22] [2.20] [0.03]

Net interest margin (lagged) 0.417 2.607*** 0.242[0.58] [2.62] [0.19]

Cost-to-income ratio (lagged) -0.041** -0.059*** 0.192**[2.11] [2.71] [2.25]

Liquidity ratio (lagged) -0.744* -0.984 -0.839[1.65] [1.49] [1.35]

Bank size (lagged) -2.961 14.165** -2.938[1.02] [2.15] [1.01]

Real interest rate (lagged) -0.230*** -0.177** -0.158*[4.15] [2.40] [1.85]

Constant 21.241*** 16.696* 5.478[3.32] [1.84] [0.51]



Robustness Analysis: Using a Narrower Measure of Bank Soundness—Loan Loss Reserves

1995-2004 1995-2000 2001-2004

Notes: Absolute value of z statistics in brackets; * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. The dependent variable in the first (reported) equation is annual percent change in outstanding loans. In the second (unreported) equation, the dependent variable is loan loss reserves as a proportion of total loans.

Weaker Baltic banks are Weaker Baltic banks are expanding faster than other expanding faster than other

banks...banks...Bank credit growth (lagged) 0.096*** 0.095*** 0.094***

[5.84] [3.64] [4.70]Distance to default (lagged) 0.313*** 0.241 0.433***

[2.59] [1.23] [3.01]Distance to default of Baltic banks (lagged) -0.346 0.684 -0.961***

[1.43] [1.46] [3.72]Real GDP growth (lagged) 2.522*** 2.514*** 1.790**

[4.43] [2.85] [2.35]GDP per capita (lagged) -0.118* -0.293*** -0.05

[1.74] [2.84] [0.56]Net interest margin (lagged) 0.703 1.758** 1.467**

[1.50] [2.25] [2.47]Cost-to-income ratio (lagged) -0.018 -0.035* 0.054*

[1.17] [1.87] [1.78]Real interest rate (lagged) -0.477 -0.944 -0.817*

[1.30] [1.53] [1.82]Real depreciation (lagged) -5.078** 15.087** -8.315***

[1.99] [2.49] [3.01]Public ownership -0.178*** -0.155** -0.076

[3.74] [2.42] [1.03]Baltic banks 5.086 -6.839 18.209***

[0.96] [0.81] [2.77]Constant 15.241*** 17.060** 7.321

[2.91] [2.15] [1.04]



1995-2004 1995-2000 2001-2004

Notes: Absolute value of z statistics in brackets; * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. The dependent variable in the first (reported) equation is annual percent change in outstanding loans. In the second (unreported) equation, the dependent variable is distance to default.

Possible explanationsPossible explanations

Real credit growth in the Baltics is Real credit growth in the Baltics is several times higher than in the several times higher than in the CEECsCEECs

Ensuring sound lending decisions and risk Ensuring sound lending decisions and risk management is much more difficultmanagement is much more difficult

Higher degree of foreign participation Higher degree of foreign participation in the Balticsin the Baltics

Additional comfort that the banking system Additional comfort that the banking system can withstand shockscan withstand shocks

More foreign affiliates are branchesMore foreign affiliates are branches Supervision is more challengingSupervision is more challenging

Foreign banks are taking greater risks Foreign banks are taking greater risks than domestic banks...than domestic banks...


Distance to default (lagged) 0.429*** 0.456** 0.398**[3.10] [2.15] [2.26]

Distance to default of foreign-owned banks (lagged) -0.448** -0.380 -0.466**[2.12] [0.96] [1.96]

Real GDP growth (lagged) 2.665*** 2.436*** 2.485***[5.54] [2.89] [4.42]

GDP per capita (lagged) -0.134** -0.293*** -0.094[2.18] [2.88] [1.17]

Net interest margin (lagged) 0.652 1.874** 1.158*[1.36] [2.27] [1.94]

Cost-to-income ratio (lagged) -0.016 -0.036* 0.048[1.01] [1.88] [1.57]

Real interest rate (lagged) -0.539 -0.900 -0.947**[1.60] [1.64] [2.13]

Real depreciation (lagged) -5.123** 15.275** -7.924***[2.02] [2.52] [2.82]

Foreign ownership 8.069** 8.709 5.85[2.09] [1.31] [1.29]

Public ownership -12.897*** -9.254 -5.129[2.75] [1.43] [0.72]

Constant 12.857** 10.782 11.156[2.31] [1.19] [1.51]



1995-2004 1995-2000 2001-2004

Notes: Absolute value of zstatistics in brackets; * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. The dependent variable in the first (reported) equation is annual percent change in outstanding loans. In the second (unreported) equation, the dependent variable is distance to default.

...but commensurate with the strength of parent banks

Lending through Nordic banks Lending through Nordic banks seems the least related to bank seems the least related to bank

soundnesssoundness

Are Banks with Weaker Parents Expanding More Rapidly?

Has Rapid Credit Growth Weakened Banks?

(1) (2)

Austria Yes? No?

Germany No? Yes?

France Yes? Yes?

Nordic countries Yes Yes?

United States Yes? No?

Italy Yes? Yes?

Belgium Yes? No?

Netherlands No? Yes?

Source: IMF staff estimates.1/ "Yes (?)" indicates a negative and statistically significant (insignificant) coefficient; "No (?)" indicates a positive and statistically significant (insignificant) coefficient. The coefficients correspond to the interaction terms of the parent bank's distance to default and country dummies and measure the marginal effect of bank soundness of parent banks from a given country vis-à-vis the average effect for all other banks. In other respects, the models used for the analysis of the country-specific effects pertaining to parent banks follow the baseline specification.

Summary of Country-Specific Results for Different Foreign Bank Owners, 2001-04 1/

Weaker banks with large foreign Weaker banks with large foreign currency exposures are expanding currency exposures are expanding

fasterfasterBank credit growth (lagged) 0.105*** 0.157*** 0.073**

[4.40] [4.25] [2.51]Distance to default (lagged) 0.346*** 0.422** 0.279*

[2.70] [2.44] [1.68]Distance to default of banks exposed to foreign exchange risk (lagged) -0.680* 0.006 -0.794*

[1.66] [0.01] [1.74]Real GDP growth (lagged) 2.681*** 3.497*** 3.495***

[4.46] [3.91] [4.27]GDP per capita (lagged) -0.123* -0.225*** -0.118

[1.96] [2.77] [1.32]Net interest margin (lagged) 0.881 3.754*** 1.242

[1.28] [4.54] [1.00]Cost-to-income ratio (lagged) 0.011 0.007 -0.037

[0.49] [0.36] [0.52]Real interest rate (lagged) -0.638 -0.133 -1.015*

[1.55] [0.23] [1.94]Real depreciation (lagged) -3.919 26.407*** -4.679

[1.54] [4.98] [1.51]Public ownership -0.160*** -0.075 -0.079

[3.16] [1.28] [0.96]Banks exposed to foreign exchange risk 33.429*** 23.238 29.541***

[3.69] [1.48] [2.80]Constant 9.808* -20.243*** 18.981*

[1.78] [2.85] [1.90]



1995-2004 1995-2000 2001-2004

Notes: Absolute value of z statistics in brackets; * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. The dependent variable in the reported equation is annual percent change in outstanding loans. Banks that are exposed to foreign exchange risk are defined as those with higher-than-average proportion of foreign-currency-denominated loans and higher-than-average rate of growth in the proportion of foreign-currency-denominated loans. The sample is composed of Czech, Estonian, Lithuanian, Polish, Slovak, and Slovenian banks.

Weaker banks with large Weaker banks with large household exposures are household exposures are

expanding fasterexpanding fasterBank credit growth (lagged) 0.127*** 0.180*** 0.086***

[6.42] [5.45] [3.80]Distance to default (lagged) 0.417*** 0.613*** 0.355**

[3.36] [3.33] [2.38]Distance to default of banks exposed to households (lagged) -0.791** -0.886** -1.889***

[2.28] [2.16] [2.86]Real GDP growth (lagged) 2.483*** 3.991*** 3.585***

[4.31] [4.49] [4.64]GDP per capita (lagged) -0.154*** -0.375*** -0.065

[2.59] [4.68] [0.79]Net interest margin (lagged) 0.955 2.318*** 1.830**

[1.62] [3.00] [2.05]Cost-to-income ratio (lagged) 0.023 0.006 0.03

[1.20] [0.30] [1.01]Real interest rate (lagged) -0.725* -0.447 -0.766

[1.84] [0.76] [1.57]Real depreciation (lagged) -2.586 26.291*** -4.706

[1.06] [5.04] [1.61]Public ownership -0.172*** -0.103* -0.059

[3.49] [1.76] [0.76]Banks exposed to households 33.672*** 50.436*** 28.312***

[4.22] [4.35] [2.68]Constant 10.593** -8.001 7.025

[2.02] [1.16] [0.93]



1995-2004 1995-2000 2001-2004

Notes: Absolute value of z statistics in brackets; * significant at 10 percent; ** significant at 5 percent; *** significant at 1 percent. The dependent variable in the reported equation is annual percent change in outstanding loans. Banks that are exposed to households are defined as those with higher-than-average proportion of loans to households and higher-than-average rate of growth in the proportion of loans to households. The sample is composed of Czech, Estonian, Lithuanian, Polish, Slovak, and Slovenian banks.

The negative correlation between The negative correlation between bank soundness and credit growth bank soundness and credit growth is the highest in household creditis the highest in household credit

NMS: Effect of Bank Soundness on Credit Growth, 2001-04 1/(In percent per year)

Baltics

CEEC

Household loans

Corporate loans

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

Source: IMF staff estimates.1/ The effect of a one-unit increase in distance to default on bank credit growth, corresponding to the coefficient on distance to default in the credit growth equation. Distance to default is measured by the number of standard deviations a return realization would have to fall for equity to be depleted.

Foreign-ownedbanks

Dom. ownedbanks

Local currency

loans

Foreign currency

loans

Regional Policy Regional Policy ImplicationsImplications

Weaker banks in the NMS have recently Weaker banks in the NMS have recently started to expand at least as fast as sounder started to expand at least as fast as sounder banks (but credit growth per se has not banks (but credit growth per se has not weakened banks)weakened banks)

Forward-looking and risk-based supervisionForward-looking and risk-based supervision Supportive market infrastructure (credit bureaus)Supportive market infrastructure (credit bureaus) Sufficient disclosure of informationSufficient disclosure of information Financial sector surveillance and analysisFinancial sector surveillance and analysis

Weaker banks’ expansion is most pronounced Weaker banks’ expansion is most pronounced in household and foreign currency lendingin household and foreign currency lending

Closer monitoring of risk exposures and lending practices Closer monitoring of risk exposures and lending practices in these marketsin these markets

Foreign banks are taking on greater risks, Foreign banks are taking on greater risks, consistent with parent banks’ strengthconsistent with parent banks’ strength

Effective cross-border cooperation between supervisorsEffective cross-border cooperation between supervisors

Calibrating Policy Response to Calibrating Policy Response to Country-Specific CircumstancesCountry-Specific Circumstances

In the Baltics, weaker banks are expanding faster In the Baltics, weaker banks are expanding faster ((Latvia, LithuaniaLatvia, Lithuania) or credit growth has weakened ) or credit growth has weakened banks (banks (EstoniaEstonia))

In In the Czech Republic, Hungarythe Czech Republic, Hungary, and , and SloveniaSlovenia, , weaker banks are expanding as fast as sounder banksweaker banks are expanding as fast as sounder banks

In In PolandPoland and and the the Slovak RepublicSlovak Republic, stronger , stronger institutions are growing fasterinstitutions are growing faster

Different intensity of risk-based policy instruments (Hilbers et al, Different intensity of risk-based policy instruments (Hilbers et al, 2005)2005)

Country-specific regulatory framework (e.g., supply-side Country-specific regulatory framework (e.g., supply-side measures)measures)

Basel II, EU Capital Requirements Directive, IFRSBasel II, EU Capital Requirements Directive, IFRS

Concluding RemarksConcluding Remarks

Probabilistic conclusionsProbabilistic conclusions Quality of banks’ lending decisions and risk Quality of banks’ lending decisions and risk

managementmanagement Macroeconomic conditionsMacroeconomic conditions

Cross-country econometric analysis Cross-country econometric analysis using publicly available datausing publicly available data

Not a substitute for country-specific stress Not a substitute for country-specific stress tests using supervisory datatests using supervisory data

A complement because it draws on a A complement because it draws on a regional set of information in a systematic regional set of information in a systematic mannermanner

Documents

Credit Growth and Bank Soundness in Emerging Europe Natalia Tamirisa Deniz Igan International Monetary Fund The 13 th Dubrovnik Economic Conference June