Bank Decision on Capital and Risk Under Capital Regulation

124 p.Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1993.This paper models bank behavior with respect to capital and risk decisions under capital regulation using an option theoretic approach. We assume that regulatory costs are imposed on equity holders if the bank fails to meet minimum capital requirements at the end of period. Incorporating regulatory cost constraints into a contingent claim model of bank equity, equity holder payoffs are derived from an option pricing framework. Linear regulatory costs allow analytic closed from solutions. Numerical simulations suggest that bank decisions on capital and risk depend on initial capital-asset ratios, asset risk, charter value, minimum capital requirements, and regulatory costs. Contrary to previous literature, capital regulation gives two different incentives to banks depending on the extent of capitalization. Most banks have incentives to increase capital-asset ratios and reduce risk-taking strategies. Importantly, however, poorly undercapitalized banks tend to adopt go-for-broke strategies.The second purpose of this paper is to examine empirically the implications of the model. Employing analysis of variance (F test and Kruskal-Wallis test) and simultaneous equation models on 133 bank holding companies during the period 1989-1991, we find that the empirical results support the differential incentive hypotheses. In particular, simultaneous equation models suggest that undercapitalized and risky banks have incentives to increase risk-taking while most banks show a mean reversion in theory risk-taking. Banks first comply with capital requirements to avoid regulatory costs and then adjust risk-taking. Thus, empirical results indicate that most banks pursue sound banking strategies, but undercapitalized banks pursue moral hazard strategies. The model and empirical findings may shed some light on the behavior of banks and provide regulators with guidance for future policy.U of I OnlyRestricted to the U of I community idenfinitely during batch ingest of legacy ETD