Ebook Transmission of Liquidity Shock to Bank Credit: Evidence from the Deposit Insurance Reform in Japan
The question of how exactly liquidity shocks in a banking sector are transmitted to the real economy has long been a subject of active discussion in the field of both finance and macroeconomics. On the one hand, according to the Modigliani-Miller Theorem, even when exposed to periodic negative liquidity shocks, banks should be able to raise sufficient funds from alternative sources swiftly to make up for the temporary funding shortfall and thus be able to finance all profitable lending opportunities (Modigliani and Miller, 1958).
On the other hand, in the presence of informational asymmetry on the value of bank assets (i.e., banks know more about the quality of their own assets than outside investors do), banks will face a lemon premium on external funds, which means that negative liquidity shocks would raise overall funding costs, thereby forcing banks to cut back on loan supply to non-financial sectors (Bernanke and Blinder, 1992; Stein 1998).
The empirical work on the relationship between liquidity and bank lending has explored how bank lending responds to liquidity shocks in aggregate data (Bernanke and Blinder, 1992; Kashyap, Stein and Wilcox, 1993; Bernanke and Gertler, 1995). More recently, the empiricists have moved away from the use of aggregate data and have begun to use disaggregated bank-level data. The motivation for such a shift in the empirical focus is that the analysis of aggregate data suffers from a serious identification problem (i.e., liquidity shocks are likely to coincide with shifts in a loan demand schedule).
Furthermore, the use of bank-level data reveals the exact mechanism of the bank lending channel: the effects of liquidity shocks on loan supply are much larger for smaller, less liquid, and less well-capitalized banks since the funding opportunities of these banks depend critically on the severity of an adverse selection problem (Kashyap and Stein, 2000; Kishan and Opiela, 2000; Jayaratne and Morgan, 2000).
However, even these recent empirical works potentially suffer from a subtle identification problem -- the lending opportunities of banks might not be completely orthogonal to their balance sheet characteristics and/or their deposit flows. For example, a flow of deposits into banks might not be entirely exogenous. Banks with more or better lending opportunities might be willing to pay higher interest rates and attract more deposits so as to finance those lending opportunities. The low level of capitalization or liquidity might not be entirely exogenous either. It is likely to be a symptom of a worsening economic environment that a particular bank is, or has been, facing; i.e., more often than not, banks become poorly capitalized and illiquid as they suffer from a large amount of financial losses on their investments.
This paper examines the transmission of liquidity shocks to loan supply in identification of liquidity shocks by exploiting the removal of a blanket deposit guarantee in Japan as a natural experiment. When the Japanese government lifted a blanket guarantee and imposed a limit on time deposits, Japan’s banking system experienced a clear regime shift. In the old regime, depositors did not face the risk of putting their deposits into “lemon” banks, which in theory must have allowed banks to avoid adverse selection problems all together when raising external funds. In the new regime, the ability of banks, especially weak ones, to raise partially insured time deposits was severely undermined because depositors had incentives to ration funds to risky banks in order to protect themselves from financial losses.
Posted in :
