유동성위기상황에서 수신고객구조가 은행위험에 미치는 영향
The Impact of Depositor Structure on Bank Risk in the Liquidity Crisis
서정호(한국금융연구원); 박정수(서강대학교); 신현한(연세대학교)
24권 1호, 99~129쪽
초록
Shin(2009), Liu and Mello(2009) 등은 금융기관이 자금조달을 위해 법인, 금융기관 등 대규모신용공여자(large lenders)에 지나치게 의존할 경우 유동성위기상황에서 뱅크런의 위험이 증폭될 수 있다는 새로운 은행실패 가설을 최근 제시하였다. 대규모신용공여자들이 제공한 자금의 상당 부분이단기성일 뿐 아니라, 이들은 시중 유동성 악화에 따른 자신의 재무제표 위험을 선제적으로 통제하기 위해 제공한 자금을 조기에 회수하려는 디레버리징 동기를 갖기 때문이라고 설명하였다. 본 논문에서는 유동성위기상황에서 수신고객구조(depositor structure)가 국내 은행들의 위험에 유의한영향을 미쳤는지를 검증하기 위해 2004년 이후 국내 은행들의 분기별 불균형 패널 데이터에 대해고정효과모형에 기초한 실증분석을 시도하였다. Lehman Brothers의 파산보호신청 등으로 인해 국내 은행들이 급격한 유동성 압박을 경험했던 2008년 하반기를 유동성위기상황으로 정의하고 은행위험의 변동을 살펴본 결과, 주식시장 참여자 뿐 아니라 예금자들도 유동성위기하에서 수신고객구조에 따른 위험에 차별적으로 반응한다는 실증적 근거를 볼 수 있었다. 은행의 특성을 나타내는통제변수들이 감안되었고, 다양한 강건성 검증도 실행하였다. 본 연구는 Shin(2009), Liu and Mello(2009) 등이 제시한 새로운 가설과 이론에 대해 우리가 아는 한 최초의 실증분석 결과를 제시하고 있다. 본 연구의 결과는 법인 등 도매조달(wholesale funding)의 비중이 지속적으로 상승하고 있는 국내 시중은행의 수신고객구조의 추이를 감안할 때 은행경영 뿐 아니라 건전성 감독규제를 설계함에 있어서도 시의성 있는 함의를 던져줄 것으로 판단된다.
Abstract
Shin (2009) and Liu and Mello (2009) presented a new theory on bank risk in which if a bank relies too much on the funding from the large lenders such as financial institutions, corporations, and major credit suppliers, it is more likely to increase the bank risk during the liquidity crisis. A significant portion of credit provided by large depositors is short-term, and it is sensitive to market liquidity because major credit suppliers (large depositors) have an incentive of de-leveraging when liquidity crisis occurs, and withdraw funds from other institutions in order to take a preemptive action against the risk of unfavorable financial situation. In this paper, we assume that capital market participants and bank depositors are concerned about the bank risk caused by depositor structure during the liquidity crisis. In other words, we assume that capital market participants expect potential withdrawals from large depositors during liquidity crisis when a large proportion of bank deposit is made by major credit suppliers. Using the fixed effect model and Korean banks’ quarterly unbalanced panel data since 2004, we investigate whether the depositor structure affects the Korean bank risk. We use the following empirical model in order to test our hypothesis: (1) where is a dependent variable representing the risk of bank failure, is a proxy variable for depositor structure (the proportion of institutional and other types of non-individual deposits), is a proxy variable for liquidity crisis, X is a vector of control variables representing bank characteristics. Dependent variable measures the risk of a bank failure at a point in time. We use stock return volatility (SDR) and percentage changes in deposits (DEP) as a proxy for the risk of bank failure. Stock return volatility is a standard deviation of an individual bank’s stock returns () for a quarter, and it measures how sensitive the capital market participants are to bank risk, while the percentage change in deposit measures how sensitive depositors are to bank risk. We use the second half of 2008 as a period of liquidity crisis because Korean banks experienced a sudden liquidity pressure following Lehman Brothers’ bankruptcy filing. In other words, takes the value of one for the 3rd and 4th quarter of 2008, while it takes the value of zero for all previous periods. We also use CDS premium as a proxy variable for liquidity crisis. During the second half of 2008, domestic banks have experienced liquidity crisis despite abundant liquidity in domestic currency. Because of deteriorating conditions in bank borrowing in foreign currency, bank risk premium has been rising rapidly. Thus, the domestic banks’ degree of stress could be measured more accurately by the CDS premium since it reflects the full risk of the overall economic system caused by foreign borrowing conditions. We find empirical evidence that the reaction to the liquidity crisis by equity market participants as well as depositors is associated with the risk caused by depositor structure. We also performed various robustness tests controlling for variables representing the characteristics of banks. To the best of our knowledge, this paper provides the first empirical evidence consistent with the recent hypothesis and theory provided by Shin (2009), and Liu and Mello (2009). The results of this paper are considered to provide timely implications in the bank management as well as in the design of the prudential regulations given the trend that the proportion of corporate and wholesale funding continues to rise in Korean banks.
- 발행기관:
- 한국금융학회
- 분류:
- 경제학