Testimony of Professor John C. Coffee, Jr.
Adolf A. Berle Professor of Law,
Columbia University Law School

Before the

Committee on Banking and Financial Services
U.S. House of Representatives
One Hundred and Sixth Congress

May 6, 1999

"The Long-Term Lessons of Long Term Capital Management"

The Long Term Lessons From LTCM

By John C. Coffee, Jr.^(*)

The public policy issues surrounding hedge funds in general, and the failure of Long-Term Capital Management ("LTCM") specifically, fall under four principal headings:

(1) Did the Federal Reserve act wisely in coordinating--indeed, engineering--LTCM's rescue?

(2) Given that banks, other creditors and counterparties ignored for an extended period the high leverage and associated risks at LTCM (and its apparent vulnerability to any increase in the yield spread), does this apparent market failure suggest the need for increased regulatory oversight?

(3) Will the proposals in the Report of the President's Working Group on Financial Markets (the "Report") assure adequate creditor monitoring of hedge funds (and other institutional speculators) in the future?

(4) If not, is additional hedge fund regulation needed to protect investors, counterparties, or creditors?

My answers are in order: (1) Probably, (2) Almost certainly, (3) Probably not--but they are still useful steps; and (4) Probably not--except at the international G-7 level. Although brevity is desirable at these hearings, I can imagine that somewhat more analysis and explanation may be appropriate:

1. The LTCM Rescue. While LTCM's rescue was not a bailout in the sense that federal funds were not invested, it was a rescue engineered by the Federal Reserve. On September 22, 1998, the President of the Federal Reserve called over a dozen top executives of LTCM's creditors to an evening meeting and warned them that "the systematic risk posed by LTCM going into default was 'very real.'"⁽¹⁾ A day later, a 16-member consortium of banks and securities firms agreed to invest over $3.6 billion in return for 90% if the equity in LTCM. The Fed's hand in the LTCM reorganization is undeniable and somewhat downplayed by the Report of The President's Working Group on Financial Markets.

Why did the Fed believe it needed to organize a rescue for a single hedge fund, whose investors were uniquely able to bear their own losses? One factor was, of course, the fragility of markets at that sensitive point in September, 1998 when the Russian default and the Asian financial crisis seemed to threaten system wide risks. Others have questioned this justification and suggested that the Federal Reserve's actions will unfortunately be perceived as extending the "too big to fail" doctrine from large banks to large bank customers. In my judgment, however, this justification of the special vulnerability of the world wide markets is probably alone sufficient to justify the Federal Reserve's limited action in the LTCM debacle.

More interesting, however, is a second justification for the Fed's involvement: Because of LTCM's extraordinary involvement in derivatives contracts, it was particularly important to avoid a formal default by LTCM, because this would trigger formal bankruptcy proceedings (both in the U.S. and, more troublingly, abroad in the Cayman Islands). As this Committee understands, derivative contracts are specifically exempted from the automatic stay provision under the Bankruptcy Code. This fact means that creditors can liquidate any of the debtor's assets under their control (for example, collateral held by the creditor). But here lies the problem: a wholesale liquidation of LTCM's positions by its counterparties and creditors would have been destabilizing given the thinness of these markets (compounded by the sudden loss of liquidity that had temporarily occurred). Had LTCM's creditors begun to liquidate its assets, prices might have collapsed as very large creditors tried to sell large positions into thin markets in competition with each other. Under such circumstances, a race for the exit (like a run on the bank) injures everyone.

Even more importantly, as the prices of LTCM's assets fell, there would have been a serious secondary impact on other financial institutions who held similar positions. They too would have experienced losses and possible defaults causing a potential chain reaction in the derivatives markets. In part, their exposure was aggravated by "copycat" phenomenon: the leading securities firms in the United States were actually mimicking LTCM's "market neutral arbitrage" strategy. During the Third Quarter of 1998, Goldman, Sachs & Co., Merrill Lynch, Bankers Trust and Salomon Smith Barney all reported losses of hundreds of millions of dollars on "relative value" trades similar to those made by LTCM and other hedge funds. Because the giants in the industry all held similar positions and were following a similar strategy to LTCM, a "distress sale" liquidation of LTCM's positions would have hurt them more as investors than it would have helped them as LTCM's creditors. In academic terms, this is an example of a "herding effect" within the financial community: LTCM was a leader and others followed it. As a result, positions became more concentrated, markets became thinner, and a race for the exit would have been even more damaging. "Herding" or copycat behavior is likely to persist. Hence, we must recognize that precisely because of the statutory exemption for derivative transactions from the Bankruptcy Code, which permits such a race to occur, there was an increased justification for the Federal Reserve to play a role corresponding to that of a bankruptcy trustee and coordinate an orderly liquidation proceeding that was in the interest of all creditors.

This situation can (and may well) reoccur. While economists have been predictably dubious of the Federal Reserve's role in the LTCM reorganization on the ground that it creates (or extends) a "moral hazard" problem under which creditors come to assume that they will be bailed out by the Government (and so fail to monitor--much as clearly happened during the 1980's savings and loan crisis), the Federal Reserve's actions (which did not involve any significant use of public funds) can also be viewed as a practical and efficient substitute for a bankruptcy proceeding. The irony is that the more we exempt creditors from the bankruptcy stay, the more they need a substitute coordinating mechanism, which the Fed provided. In the LTCM debacle, the Fed properly acted as the coordinating body that creditors needed to minimize the losses that can result under a "run on the bank" scenario.

2. The Failure of Creditor Monitoring

Why did banks and securities firms that lent to or traded with LTCM have so inadequate a picture of the risks it was facing? Why did they not understand that an institution leveraged at the level of 27:1 (at the end of 1997) was placing a highly risky bet with their funds that a specific yield spread would narrow and not widen? Clearly, banks extended credit without requiring sufficient information. The LTCM debacle points to a recurrent problem that the President's Working Group acknowledges: in a boom market and particularly under competitive conditions, banks often fail to monitor seemingly successful clients. Indeed, LTCM was not a unique case even during the 1998 period. Perhaps more egregious was the $1.4 billion unsecured loan that BankAmerica made to D.E. Shaw, another hedge fund that lacked the spectacular success record of LTCM or its reputational capital and Nobel laureates. On October 14, 1998 BankAmerica announced a $372 million write down of this loan to D.E. Shaw and a trading loss of $529 million. To mitigate further losses, BankAmerica took over D.E. Shaw's $20 billion bond portfolio, thereby in effect bailing out D.E. Shaw. While BankAmerica's losses precipitated some senior executive resignations, neither internal monitoring nor regulatory supervision nor external shareholder oversight seems to have noted or objected to the fact that this $1.4 billion loan was unsecured.

What does this imply? The appropriate legislative focus should be on excessive leverage and the deficiencies in the risk management practices of banks and securities firms, not investor protection. No doubt there has been some self correction in bank monitoring, since the time of the LTCM debacle. But over time, memories dim and crises reoccur. By one estimate, the last twenty years has seen 90 banking crises throughout the world in which the banking system losses have exceeded those experienced by the U.S. banking system in the Great Depression.⁽²⁾ Excessive risk taking by banks has been a common denominator in many to most of these crises.

In this light, reforms proposed in the Report are helpful, but modest. Only an incurable optimist would believe, however, that they are sufficient. Similar crises are likely to reappear--perhaps in hedge funds, perhaps in off-exchange derivatives trading, or perhaps elsewhere in the risk management field.

3. Direct Regulation of Hedge Funds. Although I believe the LTCM debacle exposes serious and systemic problems in creditor monitoring of large institutional borrowers, I do not believe that it supplies any persuasive justification for direct regulation of hedge funds. In overview, the problem with direct regulation of hedge funds is two fold: (1) Investor protection--the traditional primary goal of SEC regulation--does not supply a coherent justification for regulation of hedge funds; and (2) Regulation is likely to drive hedge funds offshore.

Investor protection as a justification for enhanced regulation of hedge funds does not work because hedge fund investors are by definition sophisticated and able to fend for themselves. Most are extremely wealthy. For example, LTCM required a minimum investment of $10 million, and the Tiger Fund reportedly requires a minimum investment of $5 million. The statutory exemption under §3(c)(7) of the Investment Company Act of 1940 requires that a "qualified purchaser" who is an individual own not less than $5 million in investments. Moreover, there is no broad problem that requires fixing. Even during 1998, most hedge funds made a profit, and some earned very substantial returns.⁽³⁾ Also, the hedge fund marketplace could not be more competitive. There are currently an estimated 3000 hedge funds competing for investors' funds in a Darwinian competition in which only the fittest survive. For example, sixty percent of all hedge funds do not survive three years--largely because investors withdraw their funds from those hedge funds that do not earn superlative returns. The average life span of a hedge fund is estimated at 40 months; the median life span is 31 months. Those that "survive" appear to earn average annual returns of 18 percent (versus 10.5 percent for those that are closed down).⁽⁴⁾ Finally, even LTCM earned money for its original investors. An investor who started with it in 1994 but who had his capital returned at the end of 1997 would have earned 15 percent per year.⁽⁵⁾

Because attempts to regulate hedge funds may succeed primarily in driving them offshore (as has recently happened to a degree with swaps transactions and other OTC derivatives), one faces a regulatory quandary. The optimal answer to the problem of excessive leverage is probably some coordinated policy among the G-7 nations under which all large institutional investors (including hedge funds) that engage in certain risky speculative strategies would be required to report their positions and exposures to a centralized authority. But such reporting would be private, not public. Bodies such as the Federal Reserve, the Treasury, the SEC and the CFTC would receive such data, but it would not be available to investors or the market.

Why not? Essentially, because information is proprietary. If a hedge fund or other institution is in the business of executing special trading strategies or taking concentrated positions, its positions and strategies may be of concern to regulators seeking to protect the financial system from shocks, but it is not information to which rival traders deserve access. Asking a hedge fund to disclose its positions to the world is a little like asking a poker player to reveal his face down cards to the table.

To sum up: at the G-7 level, it may be feasible and desirable to require reporting by all very large traders of their positions (at least in certain sensitive markets). But this goal should not be pursued under the guise of investor protections. The aim is instead to avoid financial shocks and liquidity crises in an environment where financial fragility is a periodic but recurrent condition. Such a program will also require a coordinated international approach and cannot simply be legislated first and implemented internationally later.

___________________

*. Adolf A. Berle Professor of Law Columbia University Law School. Much of the data cited in this brief testimony comes from a forthcoming article by Professor Franklin R. Edwards. See Edwards, Hedge Funds and the Collapse of Long-Term Capital Management, J. Econ. Perspectives (Spring 1999). Although I benefitted from this article, it should not be assumed that Professor Edwards agrees with any of my recommendations or other views.

1. Michael Siconolfi, "How the Salesmanship and Brainpower failed at Long-Term Capital," Wall Street Journal, Nov. 16, 1998 at p.1.

2. See Charles W. Calomoris, "How To Invent a IMF," The International Economy, January/February 1999.

3. For all of 1998, the average U.S. hedge fund returned 11.7% to investors. See Richard Oppel, "Not A Bad Year for Hedge Funds, a New York Times January 31, 1999, Sec 3, p.9. Global Macro funds did even better, earning more than 21 percent. "Market Neutral" funds (which is the category into which LTCM fell) earned modest but positive returns of 3 percent.

4. For an excellent overview of hedge funds and their recent experience, see Franklin Edwards, Hedge Funds and the Collapse of Long-Term Capital Management, J. Econ. Perspectives (Spring 1999) (forthcoming).

5. See Joseph Kahn and Peter Truell, "Hedge Funds Are Now Estimated to Hold $1.25 Trillion," N.Y. Times Sept. 28, 1998 at p. C-1.