How Do Risky Banks Finance Their Assets?

Article excerpt


Our aim is to empirically analyze the moral hazard conflict between banks' stock and debt holders. For this conflict to arise in a bank, two ingredients are necessary: a decreasing capital ratio (i.e., increasing leverage) and a worsening of the risk position. If these two ingredients are present, stakeholders would be reducing their exposure to risk at the expense of debt holders. That is, the former would be transferring risk to the latter. The incentives for risk shifting lie in the fact that the bank's equity has the payoff structure of a call option over the bank's value. Therefore, higher risk increases the upside potential for stockholders, whereas debt holders bear most of the downside risk.

To study this moral hazard problem, we split banks' financial structures into equity, deposits, and other funding sources. This makes it possible to find out not only whether risk shifting exists, but also to whom risk is being transferred. Indeed, we put forward a taxonomy of moral hazard based on the type of debt holders to whom risk is being transferred.

Our econometric model extends that of Shrieves and Dahl (The Relationship between Risk and Capital in Commercial Banks, Journal of Banking and Finance, vol. 16). This allows us to take into account the relationship not only between changes in risk and the assets-to-equity ratio, but also between changes in risk and in any of the ratios of the three funding sources considered. Accordingly, we use three simultaneous systems of two equations where the observed changes in risk and in the ratios of the three funding sources considered are assumed to have two components: a discretionary component and an exogenous, random, independent and identically distributed (iid) shock. The discretionary component is modeled as a partial adjustment process. That is, this component is proportional to the difference between the target value and the value in the previous period of either risk or the ratio of the funding source considered. Since target values are not observed, they are defined as a linear function that depends on the size, profitability, and capital buffer of banks, and on a set of regulatory variables. …