DETERMINANTS OF DEPOSIT-INSURANCE ADOPTION AND DESIGN Asli Demirgüç-Kunt, World Bank Edward J. Kane, Boston Colle ge Luc Laeven, World Bank Prepared by 郭郭郭 R94723019 郭郭郭 R947 23045
DETERMINANTS OF DEPOSIT-INSURANCE
ADOPTION AND DESIGN
Asli Demirgüç-Kunt, World Bank
Edward J. Kane, Boston College
Luc Laeven, World Bank
Prepared by
郭威伶 R94723019
林雪芳 R94723045
吳昱穎 R94723077
Factors that determine a country’s decision to adopt explicit deposit insurance :
1.Outside influence
2.Internally political system
Data : 170 countries over 1960-2003 period
Introduction
Every country offers implicit deposit insurance.
Adopting a system of explicit deposit insurance does not eliminate implicit guarantees.
Many countries do not have an explicit deposit-insurance scheme (EDIS).
How cross-country differences in political systems affect decisions to adopt and design an EDIS ?
Two perspectives on political decision-making :
1.public-interest perspective
2.privative-interest perspective
Public-interest perspective - government interventions promote th
e interest of the public at large (Joskow and Noll, 1981)
- protect small and uniformed depositors and assure the stability of the banking system (Diamond and Dybvig, 1983)
Privative-interest perspective
- Regulatory decisions as the outcome of interest-group competition (Stigler, 1971, Peltzman, 1976, and Becker, 1983)
- In the United States, lobbying for deposit insurance and generous design features shapes up historically as rent-seeking behavior.(Kroszner, 1998)
Deposit insurance is likely to be favored by riskier banks
In financial-services industry, political competition is strong.
Financial institutions pressure politicians for regulations that increase their franchise values (Kroszner and Stratmann, 1998)
- Differences in political systems might impact EDIS adoption
- More democratic →EDIS adoption and generous design
In poor institutional settings an EDIS with generous design features tend to destabilizes the banking system and undermines market discipline (Demirgüç-Kunt and Detragiache, 2002)
Installing an EDIS in weak institutional environment appears to retard financial development (Cull, Senbet and Sorge, 2004)
Crisis countries :
- Blanket deposit-insurance guarantees increases the fiscal cost of resolving distress without reducing the ultimate output loss or duration of the crisis (Honohan and Klingebiel, 2003)
During crisis periods, outside influences are important in the deposit-insurance adoption decision.
全體金融機構風險差別費率分級表 - 九等三級費率
\檢查資料評等綜合得分 A 級 B 級 C 級
資本適足率
資本良好機構 第一組 第二組 第三組
資本適足機構 第四組 第五組 第六組
資本不足機構 第七組 第八組 第九組
本國、外國銀行在台分行、信託投資公司及信用合作社適用
\檢查資料評等綜合得分
65 分以上
50 分以上至未達 65分
未達 50分
自有資本佔風險性資
產比率
12% 以上 萬分之 5 萬分之 5 萬分之 5.5
6% 以上至未達 12% 萬分之 5 萬分之 5.5 萬分之 6
未達 8% 萬分之 5.5 萬分之 6 萬分之 6
農、漁會信用部適用\檢查資料評等綜合得
分 65 分以上 50 分以上至未達 65分
未達 50分
淨值佔放款比率
10% 以上 萬分之 5 萬分之 5 萬分之 5.5
6% 以上至未達 10% 萬分之 5 萬分之 5.5 萬分之 6未達 6% 萬分之 5.5 萬分之 6 萬分之 6
A Logit Models of the Adoption Decision
Logit model:
In models where “Y” is qualitative, our ob-jective is to find the probability of some-thing happening.
<comparison>: Dummy variable model:
In models where “X “is qualitative.
A Logit Models of the Adoption Decision
Pi = E( Y=1 | Xi ) = 1 / (1+e-(β1+β2Xi) )
Y = 1 means something happening.
Li = ln[ Pi / (1-Pi )] = β1+β2Xi
If L is positive, it means that the variable increases the probability of Y happening, vice versa .
A Logit Models of the Adoption Decision
Political variable
A Logit Models of the Adoption Decision
Emulation
A Logit Models of the Adoption Decision
A Logit Models of the Adoption Decision
Different!!
A Logit Models of the Adoption Decision
A Logit Models of the Adoption Decision
A Logit Models of the Adoption Decision
Factors affect adoption of EDIS: Macroeconomic: GDP per capita(+) External pressure(+) Financial crisis experience(+) Bank ownership(+) Polity score(+) Executive constraint(+)
A Logit Models of the Adoption Decision1.Original model
2.Including year dummies
3.The same as 1 but includes the external pressure variable
4.The same as 1 but adds the executive constraint variable
5.Add the executive constraints variable to 3
6.Re-estimates 5 restricting the sample to the post-1970 era
A Logit Models of the Adoption Decision
7.Fit 5 to the post-1980 era
8.Fit 5 and increase the sample size by excluding three macroeconomic explanatory variable
9.Re-estimates 8 but drops observations after deposit insurance is adopted in the country.
10.Present the marginal effects and their standard errors of 8
★1-8 assume that a country make each year a decision about its deposit-insurance status. However, once EDIS is in plan, countries rarely jettison it.
Hazard Models of the Adoption Decision
Hazard rate is an interpretation of the probability of country’s leaving state (N) in year t, given that it was in state N when year began.
The Cox model
)exp()()( ii xtt
Where x is any specified vector potential explanatory variables. The exponential procedure impose on (1) the restriction that
)(t
(1)
Example:GDP per capita is denominated in thousand of US dollars. If GDP per capita increase by one unit(i.e., by US1000) then the hazard rate for adopting deposit insurance increases by exp(0.069) = 1.071 fold (7%). This suggests us that countries with higher GDP per capita are more likely to adopt EDIS sooner.
Weibull Model
Where determines where the hazard rate is increasing(>1), decreasing(<1) or constant (=1) over time.
Example:
1)( tt
Government power sharing variable (Executive constraint) also confirms that democratic countries are more likely to adopt an EDIS.
In table 11 it shows the prediction of a country who should or will adopt the EDIS.
Example for Armenia, the predicted adoption year 2006 which is 4 years away from 2002.
For Zimbabwe the model predicts adoption in year 2021 but in reality Zimbabwe adopted deposit insurance prematurely in 2003.
Democratic countries likely to adopt EDIS
Adoption proves more likely after crisis. Richer and more institutionally
developed countries are more likely to adopt explicit deposit insurance.
Conclusion