Notes from the Vault
Larry D. Wall
Too big to fail has been an important public policy issue since the 1984 bailout of Continental Illinois National Bank and Trust Company and its parent holding company, Continental Illinois Corp.1 Congress tried to end too big to fail (TBTF) in 1991 with its passage of the Federal Deposit Insurance Corporation Improvement Act (FDICIA).2 TBTF nevertheless persisted during the financial crisis so Congress again tried to end TBTF with the passage of the Dodd-Frank Wall Street Reform and Consumer Protection Act (DFA) in 2010, which provided the supervisors with numerous additional powers. However, doubts remain in the minds of many, including Federal Reserve Bank of Minneapolis President Neel Kashkari, who recently announced an initiative to develop an actionable plan to end too big to fail by year-end.
Some analysts have argued that an essential part of getting rid of TBTF is to return banking back to an earlier era when commercial banks were smaller and limited to providing traditional banking services. The DFA includes three structural changes that go part way in the direction of making banks simpler and smaller. Will these or more rigorous versions end TBTF? One of the few avenues available for assessing any attempt to end TBTF is to ask whether the proposed actions and regulatory changes would have prevented earlier bailouts. This post conducts such an analysis by revisiting the failure and bailout of Continental Illinois to see if the structural measures along the lines of DFA would have prevented the failure.
The 1980s and Continental Illinois
The FDIC's History of the 80s: Volume 1 analyzes some of the major issues the agency faced in the 1980s, including Continental Illinois.3 That analysis observes that Continental Illinois had long been a conservative bank, but that changed when its management decided to adopt a rapid growth strategy in the mid-1970s. In implementing this strategy, Continental's opportunities for in-market growth were sharply constrained by state laws, which, at the time, prohibited interstate banking and limited in-state branching to one drive-up facility, according to Federal Reserve economist Tara Rice and Erin Davis, a Morningstar equity analyst. Thus, it is not surprising that in order to implement this growth strategy, Continental increased its domestic commercial and industrial lending by 180 percent between 1976 and 1981, largely through aggressively competing in the national commercial lending market in terms of loan rates and quality. In order to fund this rapid growth in lending, Continental Illinois increasingly resorted to buying federal funds and issuing certificates of deposit.
The risks in Continental's rapid growth strategy first became apparent in late 1981 with notable deterioration in its corporate loan portfolio. Continental Illinois was also exposed to risky debt from less developed countries (mostly Latin American countries), which started showing weakness in 1982. As other banks were being downgraded, Continental managed to retain its AAA rating from Fitch in March 1982. However, market perceptions of Continental's condition deteriorated abruptly in July 1982 with the failure of Penn Square, a relatively small bank in Oklahoma. The late 1970s saw a sharp upward spike in oil prices, which led to rapid growth in opportunities to lend in the oil-producing states, including Oklahoma. Penn Square had aggressively exploited these opportunities even when doing so meant ignoring the basics of credit risk management.4 The volume of loans generated by Penn Square far exceeded its capacity to fund them, so the bank sold loan participations to other banks, and Continental Illinois was one of the largest buyers.
The failure of Penn Square led market participants to reevaluate the value of those loan participations and the health of banks with significant exposures to them. As a result, Continental found its access to domestic funding markets sharply impaired, forcing the bank increasingly to rely on paying higher rates in foreign money markets. Continental's condition continued to deteriorate through 1983 and into 1984, though it continued to pay some dividends even after Penn Square's failure, and its president and chairman remained in office through almost all of 1983.
Continental became the subject of various rumors in May 1984 and shortly thereafter overseas depositors started to run. In response, the Office of the Comptroller (OCC) took what the FDIC historical review calls the "extraordinary step" of declaring that the agency was unaware of any significant changes in the bank's operations, as reflected in its published financial statements, which would serve as the basis for rumors about financial distress at Continental. The run continued, however, so 16 major banks put together a $4.5 billion line of credit to help fund Continental Illinois. However, this action did not stop the run, so on May 17 the FDIC injected $1.5 billion in capital, the Federal Reserve committed to meet Continental Illinois's extraordinary liquidity needs, and the FDIC guaranteed all of Continental's depositors and general creditors.
These actions allowed Continental to continue operations until a permanent package was put in place in July 1984, which included the purchase of $4.5 billion of loans and the injection of $1 billion in capital by the FDIC. The chapter on Continental in an FDIC (1998) publication Managing the Crisis: The FDIC and RTC Experience, 1980–1994 reports the ultimate cost of the transaction was $1.1 billion, which was, as a percent of assets, a "modest" 3.3 percent (albeit, somewhat higher on a present value basis).5
Would structural changes like those in Dodd-Frank have prevented TBTF at Continental?
One proposed structural change that is partially incorporated in DFA is ending commercial banks' involvement in investment banking, such as by reinstating the Glass-Steagall Act as it was originally interpreted. DFA's Volcker Rule goes part way toward reinstating Glass-Steagall with provisions that limit banks' ability to engage in proprietary trading. A second proposed change is to reduce the interconnections among banks by restricting their participation in over-the-counter (OTC) derivatives. DFA goes a long ways toward reducing direct interconnections between individual banks from OTC derivatives by requiring a central counterparty for many OTC derivative transactions. Finally, some observers argue that if some banks are too big to fail, they need to be broken up. DFA does not mandate such a breakup but it gives the supervisors some new authority. Specifically, it requires the largest banks to provide a plan for their orderly and rapid resolution, and gives the FDIC and Federal Reserve the authority to take various actions—including requiring divestiture—if the agencies find a bank's plan is not credible. Would any of the DFA changes in their current or even in a stronger form have made a difference in the case of Continental?
The separation of commercial and investment banking dates back to the passage of the Glass-Steagall Act during the Great Depression. Glass-Steagall prohibited the joint ownership of commercial banks from firms "principally engaged" in certain investment banking activities. The Federal Reserve's initial interpretation of this provision had resulted in the near-complete separation of commercial and investment banking. However, this separation was weakened in the 1980s by Federal Reserve reinterpretation of the phrase "principally engaged," and this barrier was ultimately repealed by the Gramm-Leach-Bliley Act. This deregulation allowed the largest commercial banking organizations to have large investment banking operations prior to the recent financial crisis. The resulting mix of commercial and investment banking could have facilitated increased risk taking by banks prior to the crisis in a direct manner by allowing them to take more risk or had an indirect impact by shifting commercial bankers' attitudes toward a preference for higher-return activities, which were also riskier.
However, a Volcker Rule separation of deposit-taking from proprietary trading would not have helped Continental Illinois since the legal barriers of Glass-Steagall were still intact during this period. The first small regulatory reinterpretation of the phrase "principally engaged" did not happen until 1987 (H. Rodgin Cohen), and the ultimate repeal of the relevant Glass-Steagall section did not happen until 1999. Continental's failure shows that large banking organizations do not need investment banking arms to undertake high-risk activities, activities that could drive and/or hasten financial distress.
Although interconnections due to credit exposure between banks did play a role in the decision to bail out Continental, those connections had nothing to do with the OTC derivatives market. The OTC derivatives market was in its infancy with the first interest rate swap taking place in 1981, according to University of Pennsylvania Professor Robert Litzenberger (1992). Rather, Irvine H. Sprague (page 155) reports that the concern was that Continental's failure might "threaten or bring down" the 50 to 200 small correspondent banks that had deposits in Continental that equaled or exceeded their equity.6
What we can learn about bank size from the Continental Illinois experience is a little more complicated. Continental Illinois was the seventh-largest bank in the United States when it failed and was clearly regarded at that time as being a very large bank. Thus, we need some metric as to what's too large. One such metric was recently provided by MIT Professor Simon Johnson at a Federal Reserve Bank of Minneapolis symposium titled Ending Too Big to Fail. Johnson argued that we should cap individual banks' assets at no more than 2 percent of GDP. As table 1 shows, by this standard many contemporary banks would need to be split into many parts.
However, proportionate to GDP, Continental was much smaller than the largest contemporary banks. Table 2 provides the ratio of assets to GDP for the large bank holding companies as of year-end 1983.7 Continental Illinois would easily have been in compliance with Professor Johnson's test with assets equal to only 1.1 percent of GDP. Indeed, only three banks would have failed the test. However, Sprague says that from the vantage point of 1984, the possibility that a failing Continental would cause two other large banks to fail was "of even more concern" than some correspondent bank failures.
As noted, it was not predominately the direct interconnections between the large banks that caused the concern. Instead, the other large banks were vulnerable in large part because their portfolios of risky Latin American loans were proportionately much larger than Continental's.8 Indeed, in his Minneapolis discussion, Professor Johnson asserts that had the losses on these loans been fully recognized in a timely manner, Citibank would have been insolvent. Citibank and the other very large banks were also heavily reliant on large deposits and nondeposit funding that was almost entirely uninsured. Thus, if Continental Illinois had been allowed to fail with losses to its depositors, large depositors in other banks may have started running themselves.
And what if they had been run and that resulted in their failure? Would this have imperiled the real economy? Sprague doesn't directly address this question. However, one reasonable interpretation is that although these banks were relatively small, they were the primary providers of a service that was essential to the operation of the real economy. The largest banks of that time were the primary providers of commercial banking services to large domestic corporate and institutional customers. If the failure of Continental had led to runs on the other large banks, these services could have been impaired, which would significantly adversely affect large corporations and likely the overall U.S. economy.
What lessons can be drawn from Continental?
One focus of the contemporary debate over ending TBTF is structural changes intended to return banks to the days when they were smaller and focused on traditional commercial banking activities. A way of evaluating the importance of these structural changes to TBTF is to return to the collapse and resulting bailout that helped popularize that phrase, that of Continental Illinois in 1984. We learn from going back to 1984 that a banking crisis with TBTF could happen at a time when banks were focused on traditional commercial banking activities, were not interconnected through OTC derivatives, and were proportionately much smaller than those in the recent crisis. Thus, while reversing the structural changes in banking may (or may not) make TBTF less likely in the future, structural measures by themselves are likely insufficient to end TBTF.
What Continental's bailout does do is remind us that the decision on whether to bail out a bank is not just about the direct fallout from its failure but also on the potential indirect consequences for the stability of funding at other banks. In this respect, Continental's bailout was similar to the recent crisis in which the problem was not so much the weakness of any individual bank but rather the weakness of almost all of the major banks. If we want to end TBTF, past experience says we somehow have to address the problem of having too many (large) banks that fail at the same time.
Barth, James R., Apanard (Penny) Prabha, and Phillip Swagel. "Just How Big Is the Too-Big-to-Fail Problem?" Journal of Banking Regulation 13, no. 4 (2012): 265–299.
James, Christopher. "The Losses Realized in Bank Failures." The Journal of Finance 46, no. 4 (1991): 1223–1242.
Kaufman, George G. "Are Some Banks Too Large to Fail? Myth and Reality." Contemporary Economic Policy 8, no. 4 (1990): 1–14.
Litzenberger, Robert H. "Swaps: Plain and Fanciful." The Journal of Finance 47, no. 3 (1992): 831–851.
Sprague, Irvine H. Bailout: An Insider's Account of Bank Failures and Rescues. Beard Books, 2000.
Stern, Gary H., and Ron J. Feldman. Too Big to Fail: The Hazards of Bank Bailouts. Washington, DC: Brookings Institution Press, 2004.
Wall, Larry D., and David R. Peterson. "The Effect of Continental Illinois' Failure on the Financial Performance of Other Banks." Journal of Monetary Economics 26, no. 1 (1990): 77–99.
Zweig, Phillip L. Belly Up: The Collapse of the Penn Square Bank. New York: Crown Publishers, 1985.
Larry D. Wall is the executive director of the Center for Financial Innovation and Stability at the Atlanta Fed. The author thanks Paula Tkac for helpful comments. The views expressed here are the author's and not necessarily those of the Federal Reserve Bank of Atlanta or the Federal Reserve System. If you wish to comment on this post, please email firstname.lastname@example.org.
1 See Federal Reserve Bank of Minneapolis President Gary Stern and Vice President Ron Feldman (2004).
3 Most of this summary of Continental's collapse is taken from Chapter 7, titled "Continental Illinois and 'Too Big to Fail'."
4 Wall Street Journal reporter Phillip Zweig (1985) documents numerous violations of good lending practice. My favorite was an "innovation" of Penn Square that Zweig (page 120) said "could have been called the 'negotiable cocktail napkin.'"
5 Modest, that is, relative to typical FDIC losses as a percent of the failed bank's assets during that period. University of Florida Professor Chris James (1991) found a ratio of losses to assets of 30.5 percent on average for his sample of FDIC failure resolutions between 1985 and mid-1988.
6 Loyola University of Chicago Professor George Kaufman (1990) observed that these figures provided a misleading estimate of the risk that Continental posed these banks. The failure of Continental would not have wiped out all of these deposits; indeed, estimated recoveries were 97 percent to 98 percent of deposits.
7 The bank holding company asset data for these calculations is taken from Auburn University Professor James Barth, Milken Institute economist Apanard Prabha, and University of Maryland Professor Phillip Swagel (2012), who start with the 11 largest banks that were labeled as too big to fail by Comptroller of the Currency C.T. Conover.
8 Florida State University Professor David Peterson and I found that adverse news about Latin American debt was being released in May 1984 and did have an effect on bank stock returns.