Accounting and financing choice and capital adequacy: An empirical study of public and private commercial banks

This study investigates a set of accounting and financing choices of public and private banks when the banks are close to and far from capital adequacy constraints. Theories are suggested which state that both public and private banks close to capital constraints will take capital increasing actions when such actions are not constrained by regulators.;This study theorizes that, when close to capital constraints, public and private banks will both take capital increasing actions when those actions are not constrained by regulatory oversight. For banks far from capital constraints, theories are set forth which suggest that the actions of public and private banks will differ. Private banks will continue to be driven by regulatory considerations and will exhibit relatively low agency costs with the result that they will continue to take capital increasing actions. For public banks far from capital constraints, the theories suggest that regulatory considerations are reduced and agency costs become paramount, suggesting that systematic capital increasing or decreasing actions will not be taken by this group of banks.;This study examines the discretionary component of loan loss provisions, loan charge offs, securities gains and losses, and dividends for almost 7000 banks. Tests of various hypotheses are made by comparing the distribution of discretionary amounts between selected groups of interest. In addition, all variables are examined in a multivariate setting using logit.;The results are consistent with the theories presented. Banks with low capital uniformly take capital increasing actions when such actions are within their decision set. For those actions constrained by regulatory oversight, banks take capital decreasing actions, as hypothesized. For high-capital banks, evidence is presented that privately-held and publicly-held banks take different accounting and financing choice actions. The results reported for high capital banks generally support the hypotheses presented.;The logit analysis is consistent with the analysis presented in a univariate setting. This result further supports the theories suggested to motivate the observed behaviors of banks both close to and far from capital constraints.