Speech by SEC Commissioner:
Remarks Before the Conference of Business Economists
by
Commissioner Annette L. Nazareth
U.S. Securities and Exchange Commission
Melrose Hotel
Washington, DC
February 8, 2007
I am delighted to be here this evening. As advertised, I will offer some thoughts on hedge funds and systemic risk post-Amaranth. During the weekend of September 16, 2006, when it became clear that the hedge fund, Amaranth, had suffered significant losses in its natural gas trading, we heard that the losses could be as much as $4 billion. Some of the Commission staff, who follow these issues closely, initially scoffed at this number, thinking it improbably high. In fact, we now know that the losses were in excess of $6 billion.
Some observers, noting the size of these losses and the short time frame in which they occurred, have drawn the conclusion that the episode should be taken as a warning that hedge funds pose a profound risk to regulated entities and to the broader financial system. But other observers argue that the Amaranth episode demonstrates the resilience of the current system and that no new measures by market participants or supervisory authorities are necessary. They note that, despite the rather astounding losses, there appear to have been almost no significant effects on other market participants or markets experienced by Amaranth and its investors. I would like to spend the next few minutes discussing these two views in the context of the current role of hedge funds in global financial markets, with a focus on their potential systemic impact. Before I begin, I must remind you that my remarks represent my own views and not necessarily those of the Commission, my fellow Commissioners, or members of the staff.1
Although the losses were obviously painful for its investors, it appears that a number of things critical from a systemic risk perspective worked well in the several weeks following the Amaranth failure. Amaranth was an active trading operation, which had relationships with a substantial number of banks and securities firms. While founded as a convertible bond fund, Amaranth traded not only the natural gas contracts that figured prominently in producing the outsize loss, but also equities, equity-linked derivatives, credit derivatives, and bank loans, just to name a few. Through these activities, banks and securities firms that functioned as dealers and counterparts in the various markets had credit exposures to Amaranth. Yet, as far as we can tell, they all held adequate collateral to secure these exposures. Further, Amaranth appears to have begun a voluntary liquidation of many positions in markets other than natural gas to bolster their liquidity position and allow them more flexibility in liquidating their natural gas positions. In fact, they invited prospective buyers of assets, including the major banks and securities firms, to Amaranth's offices to review their positions the weekend of September 16, even before the extent of the losses became public.
For all this to be resolved as relatively smoothly as it was, many individual processes at numerous institutions needed to work very well. Banks and securities firms needed to be able to continuously determine the extent of their exposures to Amaranth. This entailed first being able to aggregate positions across the entire spectrum of trading desks, and then to mark each of these to market. For some of the less liquid loans and credit derivatives, the marking process involves the use of models with non-observable inputs about which reasonable people may differ. Yet we are not aware of any disputes over the margin calls that were made based upon these marks. In fact we know of a few cases where banks and dealers returned excess collateral to Amaranth, making liquidity available to meet margin calls by other trading partners. In situations where uncertainty prevails about the size or nature of exposures, counterparties generally decline to return excess collateral.
A number of the products that Amaranth liquidated voluntarily in early September, including corporate loans and mortgages, had extended settlements. Under the usual market convention, transactions involving these assets aren't completed, with the payment by one side and the delivery by the other, until weeks after an agreement is struck. During this period, the parties to the trade bear the risk that the asset which hasn't been delivered will rise in value or the asset which hasn't yet been paid for will decline in value. But the settlements appear to have been completed without any difficulty, even for the products where the lag was considerable.
All of this is testament to the progress made since the failure of Long-Term Capital Management in 1998. Guided by the recommendations developed by the industry group and presented by the Credit Risk Management Policy Group in its report, banks and securities firms have clearly made enormous progress in developing and enhancing risk management capacity since 1998. One piece of this involves building the infrastructure necessary to quantify and monitor exposures that, to varying degrees, are tied to the value of complex financial products. But another and perhaps more difficult task involves maintaining discipline around the many dimensions of credit extension, from obtaining initial margin to establishing the right to close out contracts in the event a counterparty fails to meet its obligations.
But, even in light of these successes, the banks and securities firms, as well as supervisory authorities, cannot afford to become complacent in the wake of what was clearly a good outcome from a systemic risk perspective last September. In particular, reviewing the events surrounding Amaranth's losses with the benefit of hindsight raises a number of questions.
First, we cannot lose sight of the fact that the soft landing was not in the context of a broader period of market stress. The losses at Amaranth occurred while the markets were awash in liquidity and credit spreads stood at (and remained at) very tight levels. Risk managers at banks and securities firms in this environment could devote their attention fully to the Amaranth positions. They also had the luxury of conducting sufficient due diligence to determine that it was prudent to allow forbearance on certain contractual provisions. This effectively permitted Amaranth to voluntarily and systematically liquidate positions. Had several funds been in distress simultaneously, or had the events occurred during a period of greater uncertainty in the broader markets, it is not clear that the same result would have occurred. In essence, verifying that the brakes bring a car to a controlled stop when the sky is sunny and the pavement is dry is certainly reassuring; but it doesn't necessarily imply a similar outcome in inclement weather.
Amaranth also highlighted, yet again, that the leverage employed by hedge funds may be difficult to control indirectly with any degree of confidence. Following the near failure of Long-Term Capital Management in 1998, the President's Working Group advanced the notion that banks and securities firms applying prudent risk management with respect to the extension of credit to hedge funds and similar market participants, might constrain "excessive" leverage. But in the ensuing years, the various channels through which leverage can be created by hedge funds and others have expanded markedly. A range of highly structured derivatives products exist today that did not exist in 1998 and which can embed enormous leverage, often with respect to fairly illiquid or complex exposures. And multiple prime brokers have become the norm among the largest funds, making it more difficult for any single prime broker to assess the overall leverage employed by a client.
Ironically, however, to the best of our knowledge, Amaranth created its enormous exposure relative to its capital by using fairly mundane forward and futures contracts which were mostly cleared through a single firm. And it managed to do so even though its bank and securities firm counterparties, including its clearing firm, apparently had de minimis uncollateralized exposure. This raises some interesting questions. In the aftermath of the failure of Long-Term Capital Management, the premise was that heightened counterparty risk management by regulated firms would constrain the leverage of unregulated hedge fund counterparties, limiting the risk they could pose to the system as a whole. Certainly the $6 billion melt-down of Amaranth appears to call into question the notion that leverage might be constrained by these practices. Yet the Amaranth failure did not have systemic implications. Was this due to the care exercised by counterparties in collateralizing their own positions? Or were they also uniquely situated to profit as counterparties to Amaranth's colossally wrong-headed bets in the energy market?
The experiences in September also raise questions about the viability of alternatives to an asset-backed lending model for providing credit to hedge funds. The traditional prime brokerage business, which involves providing financing and other services to hedge funds, is rooted in an asset-backed lending model. Such a model entails the prime broker taking no current or potential credit exposure to hedge fund counterparties. In other words, the collateral held by the prime broker is sufficient not only to cover the current loan value, but also to cover increases in exposures that might result from adverse market movements over a set period of time. The hallmark of this model is that, while due diligence may be conducted, the lending decision is much more closely tied to an analysis of the collateral than to an analysis of the creditworthiness of the client.
As hedge funds have become larger and more active market participants, they have moved beyond traditional prime brokerage relationships in executing their trading strategies. In particular, they have become quite active in the over-the-counter derivatives markets, which have a very different history than the traditional prime brokerage business, but which can also offer leveraged exposure to a variety of risk factors.
The over-the-counter derivatives markets have generally involved banks and securities firms transacting with corporate entities and other financial institutions. Such counterparties are typically rated investment grade, and decisions by banks and securities firms about extending credit in the context of such transactions typically entail analysis not only of collateral, but of the credit quality of the counterparty. This assessment focuses on the traditional criteria, including the financials, history, market presence, and personnel of a counterparty. While collateral is typically required, decisions to enter contracts that could over their term generate credit exposure are based on both the collateral terms and the creditworthiness of the counterparty. In practice, this means that in many cases the collateral obtained from some of the largest hedge funds in relation to over-the-counter derivatives trades, while sufficient to cover any current exposure, does not fully cover future exposure.
Thus banks and securities firms are sometimes exposed to potential, although not current, risk to hedge funds in the course of over-the-counter derivatives transactions. I certainly don't want to argue that such exposure is, per se, imprudent. But the type of analysis that typically underlies the rating of counterparties as a precursor to trading doesn't appear to have been particularly useful in understanding the risks at Amaranth, and probably isn't at other hedge funds as well. Amaranth was a very sizable fund with an admirable track record of posting strong quarterly performance, at least until the second quarter of 2006. It had a fairly large group of generally respected traders and back office personnel who had worked at some of the premier banks and securities firms. So this was no one person shop run from someone's garage.
More generally, the quantitative factors considered in rating counterparties, such as leverage and balance sheet strength, may not be particularly useful with hedge funds. The audited financials provide a snap shot once a year of a trading book that may change very rapidly from minute to minute. Monthly balance sheet and net asset value numbers are similarly dated, and may rely on the fairly subjective valuation of complex instruments. All in all, there are probably reasons to question the degree to which information and transparency can function as a partial substitute for the collateral of the traditional asset-backed lending model when hedge fund counterparties are involved.
Another question involves the increasingly important role of hedge funds to the broader markets. There is just no question that hedge funds have become more important market participants. Commonly cited statistics put the total assets under management at between $1 and $1.5 trillion. But those numbers may actually understate the importance of hedge funds given the direction in which the financial markets have evolved over the last decade. A general theme during that period has been the use of securitization structures to more widely, and presumably more efficiently, spread risk. Notably, several types of risk that used to reside on the asset side of bank balance sheets have now been assumed by investors. Probably the most important examples involve residential mortgages and credit derivatives. Mortgage-backed securities allow investors to receive payment for selectively bearing the risks inherent to mortgages, including credit risk and interest rate risk. Similarly, banks now routinely move credit risk off of their balance sheets to investors in collateralized debt obligations and other structured credit products.
These are clearly positive developments that increase the efficiency of capital markets by both lowering the cost of capital and providing a wide menu of risk and return combinations to investors. And hedge funds have come to play an important role in all of the markets that I've just mentioned. Hedge funds are the purchasers of most of the residual interests from the securitization of mortgages that concentrate the risks that other investors are unlikely to want. Similarly, hedge funds have been critical to the growth of the structured credit markets. They typically purchase certain pieces of the capital structure underlying collateralized debt obligations and similar deals. Without the banks and securities firms being able to sell these pieces to hedge funds, the securitization model could not exist in its present form.
The role of hedge funds as liquidity providers is not limited to markets for highly structured securitized products. A very small number of hedge funds are responsible for a significant proportion of the credit derivatives transactions in the markets today. Funds are also key purchasers in the leveraged loan markets, where loans originated by banks and securities firms to finance corporate acquisitions and recapitalizations are sold to investors rather than held on bank balance sheets.
I cite these examples to make the point that, even if banks and securities firms are prudent in managing their counterparty exposure to hedge funds, they are also dependent on the funds as liquidity providers in certain markets. We saw some evidence after the failure of Amaranth that liquidity in the market for certain mortgage residuals fell sharply. This begs the question as to the potential impact on liquidity if hedge funds more broadly were to cease purchasing certain products that banks and securities firms have grown accustomed to selling into markets. If the banks and securities firms believed the issues were temporary, they might continue to securitize assets and originate loans, even if that led to a ballooning of their balance sheets. To the extent that any withdrawal of liquidity is judged more permanent, they might markedly decrease their activities, which could impact the ability of corporations to raise funding and home purchasers to obtain mortgages. It is worth emphasizing that these are systemic risks, inherent to the current shape of the financial markets, that cannot be mitigated by prudent counterparty risk management practices on the part of banks and securities firms.
And, to some extent, that is my overall point tonight. In September of 2006, a $6 billion loss by a major market participant caused significant losses to its investors, but caused barely a ripple in the broader markets. The latter point should provide tremendous satisfaction to the risk managers at the banks and securities firms. It was they who, in the aftermath of the events of September 1998, formed the Counterparty Risk Management Policy Group and then worked diligently in the following years to implement its recommendations. But as they themselves pointed out in a follow-on report released in 2005, there is still work to be done. The infrastructure must continue to be improved and risk management practices must evolve with market conventions. And, most difficult, they must resist the temptation to relax standards as each year passes without any significant counterparty credit losses at the largest banks and securities firms. We, along with our colleagues at the Federal Reserve, who supervise the systemically important banks and securities firms, must encourage these dealers to continue to invest in their risk management and operations infrastructure.
But the supervisory community must also recognize the limits of such efforts. No matter how diligent, the risk managers at the banks and securities firms may not be able to effectively constrain the overall leverage employed by hedge fund counterparties, particularly as new mechanisms to create leverage proliferate. Nor can they protect against the systemic implications of the largest hedge funds, which function as marginal liquidity providers in certain markets, withdrawing or liquidating in a disorderly manner. To the extent that the size and market footprint of the largest hedge funds raise systemic concerns, some further attention on the part of supervisory authorities may be necessary.
This attention may stop well short of legislative action and formal regulation. For instance, structured and regular dialog between hedge funds and supervisors may be very useful. Some of the largest funds, which would surely be of greatest interest from the systemic risk perspective, have already telegraphed a willingness to interact with Commission staff on a voluntary basis. I am hopeful that this dialog will expand over time and give rise to thoughtful, targeted approaches, crafted with input from the hedge funds themselves, to better address systemic risk concerns.
Endnotes
http://www.sec.gov/news/speech/2007/spch020807aln.htm
