Common/Single frontier methodologies that are used to analyze bank efficiency and performance can be misleading because of the homogeneous technology assumption. Using the U.S. banking data over 1984-2010, our dynamic methodology identifies a few data-driven thresholds and distinct size groups.Under common frontier assumption, the largest banks appear to be 22% less efficient on average than how they are in our model. Also, in the common frontier model, smaller banks seem to be relatively more efficient compared to their larger counterparts. Hence, common policies or regulations may not be wellbalanced about controlling the banks of different sizes on the spectrum.
In this paper we propose a general model that combines the Mixture Hazard Model (MHM) of Farewell (1977Farewell ( , 1982 with the Stochastic Frontier Model (SFM) to investigate the main determinants of the probability and time to failure of US commercial banks during the …nancial distress that began in August of 2007. Unlike the standard hazard model which would assume that all banks in the sample would eventually experience the event (failure), the MHM model distinguishes between healthy (longterm survivors) and at-risk banks. On the other hand, SFM provides a measure of the performance of banks which re ‡ects management quality and potentially plays a key role in their failure, conditional on the usual …nancial ratios and other macroeconomic, structural, and geographical variables that we employ. We consider both continuous-time semi-parametric and discrete-time mixture hazard models which are separately or jointly estimated with the stochastic frontier speci…cation. Joint estimation allows not only the performance to a¤ect the probability and time to failure, but also the former to a¤ect the later. The estimation of these models is carried out via expectation-maximization (EM) algorithm and simulated maximum likelihood (SML) method due to the incomplete information regarding the identity of at-risk banks and to the fact that integration must be carried out in evaluating the likelihood function. In-and out-of-sample predictive accuracy of these models is investigated in order to assess their potential to serve as early warning tools for regulatory authorities, academic practitioners, and bank insiders.JEL classi…cation codes: C33, C41, C51, D24, G01, G21.Key words and phrases: Financial distress, bank failures, semiparametric mixture hazard model, discrete-time mixture hazard model, bank e¢ ciency.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.