On the Estimation of the Risk of Financial Intermediaries

Abstract

In this paper we reconsider the formal estimation of the risk of financial intermediaries. Risk is modeled as the variability of the profit function of a representative intermediary, here a bank, as formally considered in finance theory. In turn, banking theory suggests that risk is determined simultaneously with profits and other bank- and industry-level characteristics that cannot be considered predetermined when profit-maximizing decisions of financial institutions are to be made. Thus, risk is endogenous. We estimate the new model on a panel of US banks, spanning the period 1985q1-2010q2. The findings suggest that risk was fairly stable up to 2001 and accelerated quickly thereafter and up to 2007. We also establish that the risk of the failed banks is quite higher than the industry’s average and this risk peaks one to two-years before the default date. Indices of bank risk commonly used in the literature do not capture these trends and/ or the scale of the increase in bank risk. Thus, we provide a new leading indicator, which is able to forecast future solvency problems in the banking industry.