The impact of federal deposit insurance on savings and loan failures.

• Author: Cebula, Richard J.

1. Introduction

   The failure rate of federally insured Savings and Loans (S&Ls) has reached proportions unprecedented since the Depression years prior to 1934. From the beginning of 1934 through the end of 1979, the S&L industry experienced comparatively few failures. Indeed, there were some 14 years during this period when not a single federally insured S&L failed. In addition, there were some eight years over this time span that witnessed only one failed federally insured S&L. Moreover, the number of failed federally insured S&Ls reached double figures only three times over this entire period: 13 failures in 1941 and ten failures each in 1966 and 1970. Unfortunately, this record became blemished in the turbulent 1980s, when, according to Barth |2~, some 525 federally insured S&Ls failed.

   Barth |2, 100-101~ has argued that federal deposit insurance has encouraged the S&Ls to take on additional risk and that in so doing federal deposit insurance has significantly contributed to the S&L crisis. Barth |2, 100~ charges that ". . . the very availability of such insurance enabled many inadequately capitalized savings and loans to engage in high-risk activities and to gamble for resurrection." Barth |2, 101~ in fact alleges that ". . . the federal deposit insurance system was the unifying cause of the savings and loan disaster." In point of fact, Barth's arguments are in principle consistent with the arguments in many other studies |4; 21; 23; 24; 33~.

   The purpose of this paper is to test empirically the hypothesis that federal deposit insurance has in fact acted significantly to raise the S&L failure rate. Section II below develops the basic model. The model includes a variety of factors in addition to federal deposit insurance that are alleged by Barth |2~ and others to have induced the S&L failure problem. Section III provides the empirical analysis. The focus is on two measures of the S&L failure rate: (1) the percentage of federally insured S&Ls failing and (2) the number of federally insured S&Ls failing. Given certain limitations on data availability, the time period studied runs from 1965-1989. In addition, of necessity, the basic data used in this study are annual. Conclusions are provided in section IV.

2. The Basic Framework

   We begin the analysis by focusing on the potential impact of federal deposit insurance on S&L failures. The basic argument regarding federal deposit insurance and S&L risk-taking incentives is actually a rather simple one. We begin by noting that, in point of fact, S&Ls that are technically insolvent are oftentimes not closed (or merged with another institution) until their net worth is very significantly negative. For one thing, since it is often very difficult to assess the financial viability of various S&L assets, especially real estate loans, an institution's insolvency is very difficult to determine. In addition, federal agencies, such as the (now-defunct) FSLIC, lack (lacked) the authority to immediately close S&Ls judged to be insolvent. Indeed, the agency that chartered the S&L in question had to determine institutional insolvency before the FSLIC could take action. Clearly, given the existence of federal deposit insurance, S&Ls near insolvency, becoming insolvent, expecting insolvency, or actually insolvent were provided a strong incentive to take greater risks in the pursuit of greater rates of return because they were, in effect, gambling for financial resurrection using the FSLIC's money and little if any of their own |2; 4; 21; 23; 24; 33~. As Kane |24, 219~ so clearly points out, ". . . the FSLIC subsidizes its clients' risk-taking . . . ." If the riskier undertakings proved unprofitable or resulted in a major loss, it was the federal insurer's funds, not the S&L's own capital, that was at risk. In other words, federal (FSLIC) deposit insurance enabled and actually encouraged S&Ls that desperately needed profits in order to avoid outright failure to take greater risks in the hope of a greater rate of return; if the venture failed, it was the FSLIC that would be liable for depositor losses, not the S&L. Naturally, by inducing greater risk-taking activities on the part of S&Ls, the federal deposit insurance exposed the S&Ls to a greater probability of failure.

   The ceiling on FSLIC insurance per account was $5,000 from 1934 through 1949. The ceiling was then increased to $10,000 in 1950, where it remained through 1965. It was again increased in 1966 to $15,000, and then quickly raised again, effective in 1969, to $20,000. From 1974 through 1979, it was set at $40,000 (for non-government accounts). Finally, as of 1980, it was set at $100,000. Table I provides the ceiling amount per account of FSLIC deposit insurance, expressed in 1982 dollars, for the years 1963 through 1989. The surges in the real FSLIC insurance ceiling clearly correspond to the surges in the nominal FSLIC insurance ceiling over this period. It is hypothesized here, based on the aforementioned literature and arguments, that the higher the ceiling level of FSLIC deposit insurance, the greater the incentive for S&Ls to take risk and hence the greater the likelihood of S&L failures.

   A number of additional factors are also to be included in the analysis. For one thing, there was the 1981-1982 recession. This recession was quite severe as well as protracted |2; 13; 15; 18; 30~. Among others, Barth |2~ argues that this recession severely hurt the real estate industry, which was already at the time beleaguered by high mortgage rates. As a consequence, S&L profitability and solvency were both severely compromised. Another potential factor in the S&L crisis was the price of crude petroleum. The decline in crude oil prices from 1981 through 1985, combined with the marked price decline of crude petroleum between 1985 and 1986, appears to have hurt the real estate industry, especially in the Southwest |20; 2~, and to have thereby significantly reduced S&L profitability and solvency. The empirical analysis in section III explicitly allows both for the 1981-1982 recession (with a binary variable) and for the price of crude petroleum.

   Another set of factors relates to interest rates...