Transcript LENDING BOOMS, FOREIGN BANK ENTRY AND COMPETITION…

LENDING BOOMS, FOREIGN BANK
ENTRY AND COMPETITION:
THE CROATIAN CASE
Evan Kraft
Ljubinko Jankov
Croatian National Bank
*The views presented here are the authors’ alone and do not
necessarily represent the views of the Croatian National Bank.
Outline
• Lending booms, banking and currency crises
• Foreign banks and lending booms
• The Croatian case
– features of the lending boom
– causes: competition, liberalization, stock
adjustment, capital inflows
– consequences
– policy measures
Consequences of lending booms:
financial side
• Credit quality deterioration—looser
underwriting standards (Gavin and Hausmann
1996), dilution of relationships (Niinimaka
2001)
• Financial accelerator followed by crisis
• Financial deepening, with positive long-term
effects on growth (Wachtel 2001, Levine,
Loayza and Beck 2000 )
Policy dilemmas
• Difficult to measure extent of bad asset
problem in real time
• Where is the trade-off between preventing
crisis by slowing down growth and slowing
down beneficial financial deepening?
– speed limits view
– increased capital requirements view
– wait and see view
Consequences of lending booms:
macro side
• Investment and/or consumption boom
• Increased volatility of GDP, recession and
currency crises
What causes lending booms?
• Real business cycle theory—positive
technology or terms of trade shocks. Such
booms would not be problematic at all.
• Financial liberalization
• Capital inflows
• Wealth shocks
Lending booms and banking crises:
the evidence
• Caprio and Klingebiel (1996), Demirgüç-Kunt and
Detragiache (1997), Honohan (1997) and
Eichengreen and Arteta (2000) all find evidence that
rapid lending growth increases the probability of
banking problems
• However, Eichengreen and Rose (1998) do not find
this, and Gourinchas, Valdes and Landerretche (2001)
find that only Latin American lending booms are
strongly correlated with crises.
Foreign banks and lending booms
• Foreign banks less dependent on domestic
funding sources, above all deposits
• Foreign banks subject to strong push factors
• Competition strong among foreign players
• Evidence from Latin America and Central and
Eastern Europe confirms these observations
• Effect on instability: foreign banks grow faster
but are more sound—ambiguous effects
Croatia’s lending boom:
phase 1, 1995-1998
• Liberalization of banking laws in early 1990’s
• Substantial entry by domestic banks
• Funding:
– high deposit growth due to repatriation of deposits
held abroad after ending of hostilities
– strong foreign borrowing after Croatia received an
investment grade credit rating in January 1997
Croatia’s lending boom:
phase 2, 2000-present
• Large scale entrance of foreign banks, late 1999 and
early 2000
• Recovery of household loans begins in second half of
2000
• Recovery of enterprise loans is slower, beginning
slowly in the first half of 2001 and only reaching 1520% growth rates
– improved enterprise liquidity in 2000 may have slowed
loan demand, but also improved balance sheets
Ju
n9
O 5
ct
-9
Fe 5
b9
Ju 6
n9
O 6
ct
-9
6
Fe
b9
Ju 7
n9
O 7
ct
-9
7
Fe
b9
Ju 8
n9
O 8
ct
-9
8
Fe
b9
Ju 9
n9
O 9
ct
-9
9
Fe
b0
Ju 0
n0
O 0
ct
-0
0
Fe
b0
Ju 1
n0
O 1
ct
-0
1
Fe
b0
Ju 2
n0
O 2
ct
-0
2
Fe
b03
Credit growth: households
Graph 1: Lending to households (corrected), yoy growth
100
90
80
70
60
50
40
30
20
10
0
Ju
n9
O 5
ct
-9
5
Fe
b96
Ju
n9
O 6
ct
-9
6
Fe
b97
Ju
n9
O 7
ct
-9
7
Fe
b9
Ju 8
n9
O 8
ct
-9
8
Fe
b99
Ju
n9
O 9
ct
-9
9
Fe
b00
Ju
n0
O 0
ct
-0
0
Fe
b01
Ju
n0
O 1
ct
-0
1
Fe
b02
Ju
n0
O 2
ct
-0
2
Fe
b03
Credit growth: enterprises
Graph 2: Lending to enterprises (corrected) yoy growth
35
30
25
20
15
10
5
0
-5
-10
Causes of lending boom: demand side
• Stock adjustment under communism, economy
was “financially repressed”
– war and transition led to further write-offs and
credit contraction
– but economy is relatively developed, and probably
the equilibrium level of credit/GDP is far above
the actual
• Insider loans (in phase 1)
• Capital inflows
Causes of lending boom: supply side
• Liberalization and competition
– removal of restrictions on entry and interest rate controls
meant to stimulate supply
– increasing competition also should increase supply
• Availability of funding
– deposit growth (especially 1995-97 and 2002)
– foreign borrowing (especially 1997, 2000 on)
• Foreign bank role
– t-tests show that both privatized and de novo (greenfield)
foreign banks increased lending faster than domestic banks
in 2000 and 2001
03.2003.
12.2002.
09.2002.
06.2002.
03.2002.
12.2001.
09.2001.
06.2001.
03.2001.
12.2000.
09.2000.
06.2000.
03.2000.
12.99.3)
09.99.
06.99.
03.99.
12.98.
09.98.
06.98.
03.98.
12.97.
09.97.
06.97.
03.97.
12.96.
09.96.
06.96.
03.96.
12.95.
09.95.
06.95.
Deposit growth
Graph 3: Growth rate of total deposits, yoy, %
70
60
50
40
30
20
10
0
-10
Evidence of increased competition
• Number of banks decreases after 1998, and
Herfindahl index increases, but competition actually
increases
• Narrowing spreads between lending and deposit
interest rates
• Lower variation of market interest rates across banks
• Increased number of banks actually covering the
whole territory of Croatia
• Panzar-Rosse h test
Lending boom and banking crisis:
what is the connection?
• Downgrade incidents:
– definition: greater than 4 percentage point increase
in bad assets (B to E)
– 37 of 43 banks undergoing downgrades in 1998-99
grew faster than 30% yoy in at least 1 quarter prior
to downgrade
– 30 of 40 banks that grew rapidly experienced
downgrades
Lending boom and failure
• Early Warning System (EWS)
– best predictors of failure are deposit interest rates
and liquidity
– loan growth a weaker predictor
– but loan growth may be correlated with other
problems
– cannot conclude that rapid growth leads to failure
Macro side of lending boom, phase 1
Inflation, retail prices, %
Real GDP growth, %
Current Account, % GDP
1995
3,7
6,8
-7,7
1996
3,4
6,0
-5,5
1997
3,8
6,6
-11,6
Sources: Croatian National Bank and Central Statistical Office.
1998
5,4
2,5
-7,1
Macro side of lending boom, phase 2
Inflation, retail prices, %
Real GDP growth, %
Current Account, % GDP
1999
4,4
-0,4
-6,9
2000
7,4
2,9
-2,3
2001
2,6
3,8
-3,8
Sources: Croatian National Bank and Central Statistical Office.
2002
2,3
5,2
-7,1
Policy measures: phase 1
• Tighter monetary policy introduced in mid1997
• Chilean-style capital controls introduced in
April 1998
• Not clear whether capital controls or bank
failures and the Russian crisis slowed down
banks’ foreign borrowing and lending boom
Policy measures, phase 2
• “16% rule”—banks must buy low-interest rate
Croatian National Bank paper if growth of risk
assets exceeds 4% in a given quarter.
• “35% rule”—banks must hold liquid foreign
exchange assets equal to at least 35% of their
total foreign exchange liabilities
• These measures are mainly aimed at slowing
growth, not at preventing asset quality
problems per se
Why not just raise interest rates?
• Transmission mechanism based on fx market
• CNB bills rate would be most likely
instrument
• Raising rates could trigger more capital
inflows
• Rates would have to be raised very
substantially
• Implications for public finance
Prudential measures
• Banks that grow faster than 20% will be required to
form special reserves (0.10% of risk assets)
• Like a temporary increase in capital requirements
• Banks will be exempt if they meet higher capital
standards (15% for growth between 20% and 30%,
20% for growth between 30% and 40% etc)
• Exceptions for new banks (first 3 years).
• Mergers: growth based on sum of merged entities.