Asiasn review of Financial research

Past Issues

HOMEPast Issues Past Issues

Asian Review of Financial Research Vol.24 No.4 pp.1069-1111
A Study on the Attenuation of Procylicality in the Risk-Based Deposit Insurance System for the Mutual Saving Banks in Korea
Ho Sung Kang Manager, Korea Deposit Insurance Corporation
Kwangwoo Park* Associate Professor, Graduate School of Finance and Accounting, KAIST
Wonho Wilson Choi Assistant Professor, Graduate School of Finance and Accounting, KAIST
Joongho Han Assistant Professor, KDI School of Public Policy and Management
Key Words : Deposit Insurance,Risk-Based Deposit Insurance,Option Pricing Model,Mutual Savings Banks,Procyclicality


Korea is scheduled to adopt the risk-based deposit insurance system starting from 2014. There have been growing concerns about the procyclicality embedded in the risk-based deposit insurance program. Indeed, while deposit insurance premium decreases during the economic upswings, poorly performing financial institutions are liable for greater insurance premium during economic downturns. In this paper, we examine whether procylical property of the risk-based insurance weaken with the application of a moving average deposit insurance contract using mutual savings banks (MSBs) in Korea. MSBs in Korea have been experiencing deterioration of loan quality. In particular, credit quality of loans to project financing (PF) firms in real estate has sharply declined in the distressed housing markets. Even though these MSBs have high risk exposure to real estate-related companies, they typically pay only 0.035% of the insured deposit amount as a fixed deposit insurance premium. If the cost of deposit insurance does not increase along with the default risk, the excess cost of deposit insurer becomes a subsidy to those MSBs. We show the risk-based deposit insurance premium by using the well-known option pricing model in this paper. We then evaluate whether the current deposit in surance premium is under-priced or not. Moreover, when designing a risk-based deposit insurance scheme, we should consider not only how to adapt the financial institution specific risk factors to the deposit insurance premium but also the effect of economy-wide systematic factors such as business cycle on the deposit insurance premiums. In a recessionary period, the risk to the financial institution is much higher than it is in an economic expansion; therefore, higher deposit insurance premium should be levied. This causes an adverse effect on the soundness of the financial institution, and the situation will get much worse. This implies that the risk-based deposit insurance scheme comes with a procyclical property. Pennacchi (2005) argues that if the deposit insurance premium is designed as a long-term moving average contract, procyclicality will be reduced. Our study concentrates on estimating the risk-based fair deposit insurance premium of Korean MSBs which currently pay a fixed deposit insurance premium. This paper contributes in the literature that it is the first attempt to price the risk-based deposit insurance premiums of Korean financial institutions, or to be more specific, privately held Korean MSBs. We have also empirically tested whether procyclicality is reduced if risk-based deposit insurance is designed as a moving average-type premium. The results show that the risk-based deposit insurance premium of MSBs is highly under-estimated when compared to their default risk level. Our paper also confirms that risk-based deposit insurance premiums have a procyclical property, which fluctuates along with the business cycle. This suggests that banks with poorer financial strength will be liable for increased deposit premium, and hence the probability of bankruptcy will be greater under the simple risk-based deposit insurance scheme. Thus, we suggest the adoption of the moving average technique in the risk-based insurance scheme to attenuate its procyclicality property. We further find that with a longer moving average window, the average premium gets larger while its standard deviation becomes smaller. The reduction in the standard deviation indicates that the premium imposed on a financial institution does not change substantially in response to radical changes in business cycles. On the other hand, the increase in the average premium suggests that the premium imposed on the moving average deposit insurance system should be assessed highly in expansions relative to simple differential premiums. Also, with a longer moving average window, we show that the imposed premium gets less sensitive to financial strength measures. This implies that the moving average method mitigates the impact and significance of financial strength measures on premiums, indirectly showing the attenuation of procyclicality in risk-based deposit insurance system. In addition, the attenuating effects are substantial for mutual saving banks with weak capital adequacy, weak asset quality and weak earnings, and in small or medium sizes. As a result, the adoption of the moving average differential deposit insurance contract would be more beneficial to financial institutions with poorer financial strengths that would be severely affected by business fluctuations. The findings of this paper suggest several policy implications. While deposit insurance system plays an important role in attenuating moral hazard problems, it has been well known that the risk-adjusted deposit insurance can also destabilize financial markets by making bank performance procyclical. Based on the results of our analysis, we suggest that it is not necessarily so. Moving average deposit insurance scheme ensures that insurance premium varies with risk exposure of each bank, but the moving average of yearly risk-adjusted risk premium can smooth out the time-series variation of deposit insurance premium at the same time. We do show that the moving average (1 to 5 year moving windows) of yearly risk-adjusted risk premium can mitigate the potential problem of procycality of deposit insurance. However, we do not thoroughly examine how to set the optimal moving average window length, a crucial question for practical purposes. We leave this issue for the future research. As noted earlier, the Korean government has a plan to adopt a risk-based deposit insurance premium system. The main result of our research suggests that we adopt the moving average scheme into the risk-based insurance program, and this idea can be used as a policy reference tool for designing the risk-based deposit insurance system.
Export citation