Regulatory Reluctance to Fail Banks if Other Banks are also Weak

download report

Transcript Regulatory Reluctance to Fail Banks if Other Banks are also Weak

TOO MANY TO FAIL?
Evidence of Regulatory Reluctance in Bank
Failures when the Banking Sector is Weak
Craig O. Brown
Serdar Dinç
[email protected]
[email protected]
October 2006
Regulation in Banking
Banking is one of the most regulated sectors
(Berger et al. (1995), Barth et al. (2006))
Entry, e.g., restrictions on branching, scope, licensing
Operations, e.g., capital requirements
Exit, e.g., license revocations, government takeovers
Banks can defer failure by issuing new deposits until
regulators take action
Brown & Dinç
2
Regulatory Reluctance &
Intervention in Failing Banking
Regulators seem reluctant to intervene in failing banks if the
whole sector is also weak.
Theory: Mitchell (2001), Acharya & Yorulmazer (JFI, forth.)
S&L Crisis in U.S.: Kane (1989), Kroszner & Strahan (JF, 1996)
Banking Crisis in Japan: Hoshi & Kashyap (2001)
Are these two cases exceptions??
Brown & Dinç
3
Bank Failures in Emerging Markets
 Crisis Literature (country-level): Economic, institutional, and regulatory
framework that makes crises more likely
e.g., Barth et al. (2006), Beck et al. (JBF 2006), Caprio and Klingebiel (2002),
Claessens et al. (2005), Demirguc-Kunt and Detragiache (1998, JME 2002)
 Bank Failures in Asian Crisis (bank-level, 4 East Asian Countries):
Political connections, market’s ability to predict failures
Bongini et al. (JFSR 2001), Bongini et al. (JBF 2002)
 Politics of Bank Failures (bank-level, 21 major emerging markets):
Politicians wait until after the elections before closing/taking over a failing
bank.
Brown & Dinç (QJE 2005)
Brown & Dinç
4
Focus of the Paper
Are the regulations about failing banks implemented regardless of
the health of other banks?
Or,
Are the regulators reluctant to close/take over failing banking if
the whole sector is weak?
Null Hypothesis:
Government takeover or closing of a failing bank does not
depend on the health of other banks (once the macroeconomic
factors are controlled for)
Brown & Dinç
5
Main Result: Preview
Government is less likely to take over or close failing
banks if other banks are also weak!!
This is robust to…
Bank-specific Factors;
Macroeconomic Factors;
IMF Programs;
Election Concerns
Brown & Dinç
6
Sample Construction
(Same sample as in Brown & Dinç (QJE 2005))
Beginning sample: Largest 10 banks in 1993 in each of 21 major
countries
All the emerging countries for which The Economist provides
statistics, except China, and Egypt:
Argentina, Brazil, Chile, Colombia, Czech, Hungary, India, Indonesia, Israel,
Malaysia, Mexico, Peru, Poland, Russia, Singapore, S. Africa, S. Korea,
Taiwan, Thailand, Turkey, Venezuela
Main Data Sources
• Bank Failures: Factiva
• Ownership, Mergers: Bankscope, Factiva, Bankers’ Almanac
• Balance Sheet: Bankscope
Brown & Dinç
7
Banks Failures by Country (Table 1-first half)
Always Government-owned
COUNTRY
Private Banks
Total Number
of
Banks (1993)
Total Number
License revoked
or liquidated
Total Number
Taken over
by the government
License revoked
or liquidated
10
10
10
10
10
10
60
5
2
-2
3
2
14
------0
5
8
10
8
7
8
46
5
--5
-4
14
------0
10
10
10
10
10
10
10
70
2
1
1
2
2
1
1
10
-------0
8
9
9
8
8
9
9
60
-3
-1
3
1
4
12
-------0
Southeast Asia
Indonesia
Malaysia
Singapore
South Korea
Taiwan
Thailand
Total (Southeast Asia)
Latin America
Argentina
Brazil
Chile
Colombia
Mexico
Peru
Venezuela
Total (Latin America)
Brown & Dinç
8
Banks Failures by Country (Table 1 - cont.)
Private Banks
Always Government-owned
COUNTRY
Rest of the World
Czech Republic
Hungary
India
Israel
Poland
Russia
South Africa
Turkey
Total (Rest of the
World)
Total (WORLD)
Total Number
of
Banks (1993)
Total
Number
License
revoked
Total Number
10
10
10
10
10
10
10
10
-1
9
2
3
2
1
4
---------
10
9
1
8
7
8
9
6
4
1
---2
-1
2
----4
---
80
22
0
58
8
6
210
46
0
164
34
6
Brown & Dinç
License
Taken over
by the government revoked
9
Bank Failures Around the World
Bank Failures are very common. 24% of all private
banks failed.
Bank Failures are evenly distributed across the
continents.
Take-over of failing banks by the Government is very
common. 85% of all failures.
Brown & Dinç
10
Sample Statistics (Table 2)
Variable Name
Assets (in $B)
Assets / GDP
Total Loans/ Assets
Total Deposits / Assets
Capital Ratio
Operating Income / Assets
Mean
sd.
N
Mean
sd.
N
Mean
sd.
N
Mean
sd.
N
Mean
sd.
N
Mean
sd.
N
Failed Banks
10.048
11.708
140
0.056**
0.069
691
0.588
0.205
138
0.766
0.149
138
0.044***
0.163
140
-0.019**
0.196
137
Brown & Dinç
Other Banks
10.533
13.239
691
0.070
0.092
140
0.574
0.155
684
0.752
0.153
683
0.092
0.054
691
0.015
0.024
684
All Banks
10.451
12.988
831
0.067
0.089
831
0.577
0.165
822
0.754
0.152
821
0.084
0.085
831
0.010
0.084
821
11
Econometric Methodology
Cox Proportional Hazard analysis for the time until bank failure
h(t )  h0 (t ) exp β x it 1   * z i ,t 1 ,
t  t i ,..., Ti
xi: Bank specific (and macroeconomic) variables
z-i: A measure of health for other banks in the same country. (Typically, the
weighted average across all other banks for a bank specific variable
included in xi)
Null hypothesis  g=0 (a failing bank is closed regardless of other banks)
Regulatory reluctance  g>0 (the healthier other banks are, the more likely a
bank is closed by the regulators)
(Errors are clustered at the country level)
Brown & Dinç
12
Regulatory Reluctance to Fail Banks
if Other Banks are also Weak (Table 3)
Total Assets / GDP
Capital Ratio
Income
-0.308
-0.345
(0.266)
(0.291)
-0.121***
(0.022)
-0.114***
(0.023)
-0.430
(0.301)
-0.117***
(0.031)
-0.006
(0.037)
Capital Ratio_OtherBanks
-0.357
(0.314)
-0.162***
(0.041)
-0.043
(0.041)
0.261*
(0.147)
Income_OtherBanks
Bank-years
p-value of W
881
0.000
854
0.000
Brown & Dinç
854
0.000
854
0.000
-0.428
(0.300)
-0.168***
(0.040)
-0.065
(0.045)
0.357**
(0.170)
854
0.000
13
Regulatory Reluctance: Robustness?
Government is less likely to take over or close failing
banks if other banks are also weak!!
Is this effect robust to…
Macroeconomic Factors?
IMF Programs?
Election Concerns?
Additional Bank-specific Factors?
Brown & Dinç
14
Regulatory Reluctance: Robustness to
Macroeconomics (Table 4A)
Total Assets / GDP
-0.314
(0.246)
Capital Ratio
-0.149***
(0.043)
Income
-0.032
(0.038)
Capital Ratio_OtherBanks 0.276**
(0.132)
GDP Growth
-0.072**
(0.031)
GDP per capita
-0.257
(0.280)
-0.160***
(0.042)
-0.034
(0.042)
0.303**
(0.137)
-0.366
(0.324)
-0.161***
(0.040)
-0.041
(0.042)
0.253*
(0.151)
-0.329
(0.311)
-0.166***
(0.043)
-0.041
(0.042)
0.317*
(0.169)
-0.374
(0.313)
Currency depreciation
0.042
(0.073)
Inflation rate
1.147
(1.327)
Real Interest rate
Bank-years
p-value of W
-0.387
(0.331)
-0.162***
(0.040)
-0.037
(0.046)
0.249*
(0.153)
854
0.000
854
0.000
844
0.000
Brown & Dinç
854
0.000
-0.000
(0.000)
805
0.000
15
Regulatory Reluctance: Robustness to
Macroeconomics (Table 4B)
Total Assets / GDP
Capital Ratio
Income
Income_OtherBanks
GDP Growth
-0.391*
(0.235)
-0.155***
(0.042)
-0.053
(0.040)
0.381**
(0.155)
-0.079***
(0.030)
GDP per capita
-0.333
(0.274)
-0.168***
(0.041)
-0.058
(0.045)
0.419***
(0.157)
-0.437
(0.306)
-0.167***
(0.040)
-0.063
(0.045)
0.349**
(0.174)
-0.412
(0.301)
-0.174***
(0.042)
-0.068
(0.045)
0.449**
(0.192)
-0.437
(0.297)
Currency depreciation
0.046
(0.069)
Inflation rate
1.531*
(0.916)
Real Interest rate
Bank-years
p-value of W
-0.451
(0.315)
-0.167***
(0.039)
-0.059
(0.049)
0.343*
(0.177)
854
0.000
854
0.000
844
0.000
Brown & Dinç
854
0.000
-0.000
(0.000)
805
0.000
16
Regulatory Reluctance: Elections & IMF
Lending (Table 5)
Total Assets / GDP
-0.401
(0.297)
Capital Ratio
-0.172***
(0.037)
Income
-0.064
(0.043)
Capital Ratio_OtherBanks 0.287**
(0.116)
Income_OtherBanks
BeforeElection
-1.575***
(0.594)
IMF loans / GDP
Bank-years
p-value of W
-0.354
(0.317)
-0.154***
(0.040)
-0.033
(0.040)
0.418**
(0.169)
854
0.000
0.415
(0.297)
854
0.000
Brown & Dinç
-0.474
(0.297)
-0.169***
(0.037)
-0.077*
(0.044)
-0.495
(0.341)
-0.155***
(0.039)
-0.064
(0.042)
0.329*** 0.554***
(0.127)
(0.178)
-1.467**
(0.575)
0.472*
(0.282)
854
854
0.000
0.000
17
Regulatory Reluctance: Additional BankLevel Controls (Table 6)
Total Assets / GDP
Capital Ratio
Income
-0.439
(0.328)
-0.118***
(0.037)
-0.005
(0.041)
Capital Ratio
_OtherBanks
-0.338
(0.348)
-0.161***
(0.042)
-0.045
(0.046)
0.115
(1.153)
854
0.000
-0.228
(0.248)
-0.235***
(0.050)
-0.009
(0.061)
0.349**
(0.152)
(0.137)
-0.218
(0.911)
0.373*
(0.191)
-0.567
(0.969)
Lending Margin
Bank-years
W
-0.328
(0.234)
-0.164***
(0.056)
0.034
(0.068)
0.263*
Income_OtherBanks
Loans
-0.380
(0.334)
-0.166***
(0.040)
-0.072
(0.052)
854
0.000
854
0.000
Brown & Dinç
-0.321
(0.238)
-0.248***
(0.049)
-0.040
(0.066)
0.492***
(0.162)
-0.456
(0.752)
842
0.000
-0.218
(0.658)
842
0.000
-0.143
(0.580)
842
0.000
18
Main Result: Summary
Government is less likely to take over or close failing
banks if other banks are also weak!!
This is robust to…
Bank-specific Factors;
Macroeconomic Factors;
IMF Programs;
Election Concerns
Brown & Dinç
19