Investigating the Relationship between Supervisory Change and Community Bank Failure in a Competing Risk Framework

Abstract

Bank failures began to rise in 2007 and more than 350 were closed by the end of 2011. A number of observers have asserted that the ability of banks to switch among competing supervisors was an important contributing factor, but the issue has not been carefully investigated in empirical work. This study explores this issue using a sample of almost 6500 community banks that existed at the end of 2007. A competing risk hazard model developed by Fine and Gray (1999) is employed to estimate the effect of supervisory changes during the 2004-2007 period on bank failure over the 2008-2011 time period. The model explicitly recognizes that merger as well as subsequent supervisory change represent possibly important competing risks. The evidence indicates that community banks that changed their primary federal supervisor during the four year period ending on 12/31/2007 were more likely to fail over the following 4 year interval than those that did not make such a switch. This effect is evident for banks that switched to all three primary federal supervisors. Another interesting finding is that state banks that changed only their primary federal supervisor and not their charter were also more likely to fail all else equal. No significant differences in failure likelihoods are apparent when the banks continuously supervised by each primary federal supervisor are compared to one another. These results are evident in both simple comparisons of failure rates, and in the estimated multivariate competing risk failure models. These findings taken together suggest that differences in agency funding sources or organizational responsibilities are not the primary cause of the higher failure rates of the switching banks.