Risk taking and failure in de novo savings and loans in the 1980s

Abstract

This article examines failure rates in de novo S&Ls that initiated operations during the 1980-1986 period. Overall failure rates are similar to those for existing institutions but are found to vary significantly by location, time of charter, and organizational form. Both univariate tests and results from a probability-of-failure model indicate that inadequate capital, economic stress, poor management of higher risk lending allowed by broader powers, and operating inefficiencies contributed significantly to the likelihood of failure. Use of brokered funds and rapid asset growth are also significantly related to failure likelihood. Interestingly, for those S&Ls which eventually failed, rapid asset growth and high proportions of nonperforming assets actually delayed the timing of their failure. We interpret this to be the result of regulatory forbearance. We also find significant differences between the financial characteristics of de novo and non-de novo S&Ls.

The failure of new entrants in commercial banking markets: a split‐population duration analysis

“…The probability of bank failure, the determinants of that probability, the expected survival time conditional on bank failure, and the determinants of that survival time are estimated for the de novo banks and the established banks using a ''split-population'' duration framework. Schmidt and Witte (1989) developed the split-population approach, and both Cole and Gunther (1996) and Hunter et al (1996) have estimated split-population models of financial institution failure. The central failure concept in these models is the hazard rate, the probability that a bank will fail at time t given that it has survived through all of the previous time periods leading up to time t.…”

“…Cole and Gunther (1996), Thomson (1992), Whalen (1991), Wilson (1995, 2000), and many others have modeled the failure of mature financial institutions, but only a handful of studies have modeled the failure of new financial institutions. 2 Hunter, Verbrugge, and Whidbee (1996) estimated a split-population duration model for de novo savings and loans; they found that credit risk, adverse economic conditions, low capital stocks, and cost inefficiencies all contributed to de novo thrift failure, and they also found evidence consistent with regulatory forbearance at failing thrifts. Santarelli (2000) estimated a proportional hazard model for de novo financial institutions in Italy and found that small start-ups and nonbank start-ups had higher risks of early exit than did large start-ups and bank start-ups.…”

Bank Supervision, Regulation, and Instability During the Great Depression

“…In states that had multiple capital requirements, this paper uses the requirement that applied to the smallest municipality, which was usually for 3,000 people. 17 Hunter, Verbrugge, and Whidbee (1996) provide some empirical support that minimum capital requirements reduced the likelihood of failure for de novo Savings and Loans in the 1980s.…”

Economic capital for nonperforming loans

“…Hu et al (2004) find that, among loans, the rate of nonperforming debt increased in Taiwan during the 1990s. Hunter et al (1996) find that, for the United States in the 1980s, the percentages of nonperforming loans in a portfolio is associated with the portfolio owner's failure probability.…”