The US Banking Debacle; G. Kaufman

Between 1980 and 1991, when fundamental corrective laws were enacted, some 1,500 commercial and savings banks (insured by the Federal Deposit Insurance Corporation) and 1,200 savings and loan associations (insured by the former Federal Savings and Loan Insurance Corporation) failed and were resolved by the regulatory agencies. These resolutions represented about 10 percent of all banks at the beginning of the period and 25 percent of all S&Ls. In addition, an even larger number of institu tions were in precarious financial condition Dr. Kaufman is the John Smith Professor of Banking and Finance at Loyola University of Chicago, and is Co-Chair of the Shadow Fi nancial Regulatory Committee. This paper is a shortened version ofa longer paper presented at the International Conference on Bad Enterprise Debts in Central and Eastern Europe in Budap est, Hungary on June 6-8, 1994. The author is indebted to Herbert Baer (WorldBank) and Larry Mote (Comptroller of the Currency) for helpful comments and suggestions.

at some time during this period. The costs of the failures were high, not only to the shareholders of the failed institutions, but also to· the surviving institutions, which were required to pay premiums to the de posit insurance agencies, and to U.S. tax payers, who were forced to make good on the losses after the resources of the S&L insurance fund had beeri exhausted. For banks, the loss to the FDIC and thus to other solvent banks was about $40 billion. For S&Ls, the loss was near $200 billion, some $150 billion of which was beyond the re sources of the FSLIC and was therefore charged to U.S. taxpayers. The losses accrued primarily to the fed eral insurance agencies and taxpayers rather than to depositors and other creditors be cause the insurance effectively guaranteed the par value of deposits up to $100,000per account de jure and, except at some small banks, almost any amount of deposits and even borrowings de facto, regardless of the value of the bank's assets. The FDIC and the former FSLIC were funded by premi ums imposed on banks and S&Ls, respec tively, and both had implicit access to the U.S. Treasury that legislators were unwill ing either to challenge or to make explicit until near the end of the debacle.

The crisis ended in the early 1990s,when 254 interest rates declined, the yield curve turned steeply upward sloping, a series of rolling geographic recessions in various re gions of the country came to an end, the aggregate economy slowly expanded, the real estate market bottomed out, and newly adopted legislation increased the cost of poor performance and failure to both the institutions and the regulators. By 1994, both the banking and thrift industries were in their best financial condition since the early 1960s and were realizing record prof its. The number of failed and problem insti tutions declined sharply. II. Background Banking has always been a volatile indus try in the United States, but until the 1930s not an unusual one.! The annual failure rate for commercial banks from 1870 to 1913, before the establishment of the Federal Reserve System, averaged 0.78 percent compared to 1.01 percent for nonbanks. The annual volatility of the failure rate was greater for banks, however. The relatively low failure rate existed despite a banking structure that favored failures by restricting banks to one or at best only a few offices, thus preventing them from reducing risk through geographical and product diversifi cation. As a result, the country had thou sands of independent banks; the number peaked at 30,000 in the early 1920s. The bank failures increased sharply in the 1920s to near 600 per year, but most of the failures were very small banks. Some 90 percent of the banks had loans and investments of less than $1 million, which adjusted for inflation would be equivalent to only about $10 mil lion currently, and would rank them among the very smallest banks. Their failure had no visible effect on national economic activity.

They were primarily located in small agri cultural towns in the midwest. When a recession hit these towns from the rapid fall in farm prices after the post-World War I runup, the local automobile dealer failed, the local drugstore failed, and the local bank failed. But things changed dramatically in the 255 1930s at the onset of the Great Depression. Between 1929 and 1933, the number of banks declined from 26,000 to 14,000, mostly by failure. Indeed, the very first act of newly elected President Franklin D. Roosevelt was to declare a "bank holiday" and close all banks in the country for at least one week in order to prevent depositors from cashing any more of their deposits into currency. The banks were permitted to reopen if the government found them sol vent. Thereafter, banking became a rela tively stable industry through the late 1970s. The number of bank failures averaged only near 10 per year and the number of S&L failures was not significantly greater. Then the picture changed again.

Before analyzing the 1980s, it should be noted that both the 1930sand 1980sdebacles occurred after the creation of government institutions intended to correct failings in the system that were believed to have been at the root of the problem, and in order to reduce the likelihood of large numbers of simultaneous failures in the future. The Federal Reserve was established in 1913 in the aftermath of sharp jumps in the number of bank failures in 1894 and 1907 in order to increase flexibility in the system. The Fed was to facilitate the flow of bank reserves from capital surplus to capital deficient ar eas, to provide micro-liquidity through the discount window to individual solvent banks experiencing temporary liquidity problems, and to provide macro-liquidity to the banking system by offsetting outflows of currency and gold. For whatever reasons, not 20 years after it was established, the Fed failed to achieve these objectives suffi ciently to prevent the banking crisis of the 1930s, which was far larger, longer, and costlier than any banking crisis before the establishment of the Fed. Indeed, the Fed appears to have introduced greater rigidities at the time of the Great Depression, e.g., prohibiting the issuance of clearing house certificates and making temporary bank sus pensions more difficult, than existed before its establishment. 2 In large part as a result of the Fed's failure to prevent a recurrence of large-scale bank 256 THE FREEMAN • APRIL 1995 failures, the FDIC was established in 1934.

While the Fed's decisions to provide liquid ity to the banking system in order to offset depositor runs into currency were discre tionary, the FDIC operated by rules that effectivelyeliminated the need for bank runs by unconditionally guaranteeing the par value of insured deposits regardless of the bank's financial condition. This objective was quickly realized and, combined with a more cautious set of bankers and more restrictive regulations imposed by the Bank ing Act of 1933,the number of bank failures dropped equally quickly and remained low for the next 50 years. However, as was true of the Federal Reserve's structure, flaws eventually appeared in the FDIC that in time led to increases in bank failures that matched the conditions in the 1930s before the introduction of deposit insurance. III. The S&L Debacle3 Savings and loan institutions are tradi tional residential mortgage lenders. Before the introduction of deposit insurance in 1934, S&Ls made primarily intermediate three-to-five-year renewable mortgage loans. These loans were effectively variable rate mortgages with sizeable down pay ments. They were financed bytime deposits (legally labeled share capital), which were not necessarily redeemable on demand. As a result, neither the S&Ls' interest rate nor liquidity exposures were very great.

But things changed dramatically after 1934. Public policy encouraged S&Ls to make progressively longer-term (first 20, then 25, and finally30-year)fixed-rate mort gages with progressively smaller down pay ments. At the same time, the new deposit insurance program effectively increased the liquidity and shortened the maturity of their deposits. These changes increased the in stitutions' exposure to interest rate and liquidity risk. Indeed, the large degree of maturity (duration) mismatch by the mid 1970s made the industry a disaster waiting to happen. When interest rates increased sharply in the late 1970s as a result of inflation, the disaster occurred. Be~ween 1976 and 1980, interest rates on three-month Treasury bills jumped from 4 percent to 16 percent and those on long-term Treasury securities from 6 percent to 13 percent. By 1982, an esti mated 85 percent of all S&Ls were losing money and two-thirds were economically or market value insolvent so that, ceteris pa ribus, they would be unable to pay their depositors in full and on time. The negative economic net worth of the industry and the corresponding loss to the FSLIC was gen erally estimated to be about $100 billion,4 although some estimates placed it as high as $150 billion. This figure represents the dif ference between the par value of deposit accounts (the large majority of which were less than the maximuminsured $100,000per account) at insolvent institutions and the market value of the S&Ls' assets. But the FSLIC resolved only a very smallnumber of the insolvencies for a number of reasons, including:5 • It was overwhelmed by the large num ber of insolvencies, and its staff was far too smalland unprepared to deal with the crisis, • It had insufficientreserves to cover the deficits at insolvent institutions and payoff depositors at par, whether the institutions were sold, merged or liquidated, • Formal recognition of the large losses would be a black mark on the agency's record, • Formal recognition of the large losses and number of insolvencies might spread.

fear among the public and ignite a run on all institutions that would spill over to com mercial banks and·even beyond to the mac roeconomy. Further, • Many of the losses were "only" unrec ognized paper losses; and, because interest rates are cyclical and there was a high probability that they would decline again in the not very distant future, it was hoped that waiting would restore the associations to economic solvency. Therefore, regulators publicly denied the magnitude of the problem, argued that the problem was a liquidity rather than a sol vency problem, introduced creative ac counting measures to make the industry's THE U.S. BANKING DEBACLE OF THE 1980s 257 net worth appear higher even than the al ready overstated book value levels (i.e., they covered up the evidence), delayed imposing sanctions on insolvent and near insolvent institutions, and encouraged insti tutions to reduce their interest rate exposure by using newly permitted variable-rate mortgages and shorter-term loans to reduce their maturity mismatch. And the regulators and the industry lucked out. Interest rates declined sharply from ·1982 through 1986.

This reversal in rates caused the industry's net worth to rise and by 1985 its estimated negative net worth was only about $25 billionand was expected to improve further, ceteris paribus. But ceteris did not remain paribus for many institutions. A substantial number incurred increases in credit risk that offset the decline in interest rate risk and either prevented their net worth from increasing greatly or actually caused it to decline fur ther. The assumption of credit risk was either unintentional, arising from severe local and regional economic recessions, or intentional, arising from calculated gambles to regain solvency. The first and most severe regional reces sions started in the mid-1980sin Texas and the neighboring energy-producing states in the Southwest following the collapse of world oil prices. This area had experienced a strong economic surge based on sharply rising oil prices and expectations of contin ued price increases. Employment, income, and real estate values all increased sharply and stimulated both a rapid immigration of people in search of employment and a build ing boom, particularly in commercial real estate. Much of this boom was financed by local S&Ls. When oil prices not only failed to increase further after 1981, but declined sharply from $30 a barrel in 1985to near $10 in 1986, the bubble burst.6 As incomes and real estate values dropped, borrowers de faulted on loans, and collateral values fell too fast for many lending S&Ls to protect the value of all their loans. As a result, many S&Ls became insolvent.

At the same time, a number of institu tions, particularly those that had only recently converted from mutual ownership (which was the prevailing form of owner ship) to stock ownership in order to raise additional capital more easily, became tempted to "gamble for resurrection." Be cause these institutions had little if any market value capital of their own to lose, this was a logical strategy. If the high-risk bets paid off, the institution won and possi bly regained solvency. Ifthe institution lost, the FSLIC bore the loss. That is, heads the institution won, tails the FSLIC lost! Some S&Ls placed progressively larger bets on the table by offering above market interest rates on deposits so that their deposit size grew rapidly. Such gambling was often ac companied by fraud, either ex-ante deliber ate or ex-ante inadvertent through excessive carelessness in extending and monitoring loans. Particularly at the more rapidly grow ing associations, loan documentation was frequently incomplete or even nonexistent, record keeping casual at best, and loan collection was sporadic and done with little enthusiasm. Some of the new owners were land developers, who are gamblers almost by nature. They used greatly overinflated values of their personal properties as the base for their institution's capital, and the resources of the institution as their personal "piggy banks" to finance their ventures.

Losses were often not recognized on the institutions' books on a complete or timely basis, so that the institutions gave false appearances of solvency. The National Commission appointed in 1992to identify and examine the origins and causes of the S&L debacle concluded that: "It is difficult to overstate the importance of accounting abuses in aggravating and obscuring the developing debacle. It would have been difficult for the process to con tinue for so long in the absence of an information structure that obscured the ex tent of the mounting losses.,,7 The FSLIC economic deficit (computed as the differ ence between the par value of insured de posits at economically insolvent S&Ls and the market value of their assets), which had declined from some $100 billion in 1982 to near $25 billion in 1985, climbed back up to 258 THE FREEMAN • APRIL 1995 above $100 billion in 1989, almost entirely due to losses from credit risk exposure.