Market Discipline and Bank Subordinated Debt: Note

A lengthy literature in banking has addressed the question of whether the market prices of liabilities respond to individual bank risk-taking activity, that is, whether any ;;market disciplines' exists. However, despite the obvious importance of this question the current record has led to no consensus on the issue. Results have been contradictory. In this Journal alone Hannan and Hanweck (1988) conclude that the market does extract at least some price for risk-taking, while Avery, Belton, and Goldberg (1988) find no such evidence. In this study we argue that this line of inquiry has a serious f law in that it fails to use theoretical models of bank instrument valuation to determine the appropriate testing of market discipline. Therefore, the empirical models have lacked the necessary exactness to isolate the appropriate null hypotheses. To redress this shortcoming and advance the methodology of this area of the literature, we use a pricing model which comes to the banking area from the options pricing literature to look at the yield spread on bank liabilities. Using this model it is clear that the value of large, uninsured, bank liabilities cannot be described as a linear, monotonic, function of risk. Moreover, current bank regulation affects the assumed role risk measures play in the pricing of such debt instruments. Both of these complications mean that simple regressions are not likely to adequately address the relationship between the value of debt and underlying risk. In the final section we empirically investigate the presence of market discipline using contingent claims pricing. We find that accounting measures of bank risk still have little predictive value in explaining yield spreads, and therefore, offer little evidence of meaningful market discipline.