Remarks from an April 23, 2007 speech by Orin Kramer, general partner at Boston Provident LP, New York, and chairman of the New Jersey State Investment Council, Trenton.
I have been asked to address over the next few minutes the narrow question of the relationship between the unprecedented growth of the derivatives market, hedge funds, financial market conditions and their relationship to the overall economy.
In examining financial market conditions, it is important to acknowledge the substantial successes of the past decade. Despite major shocks the Asian crisis of 1997, Long Term Capital (Management) and Russia in 1998, the bear market in stocks, the corporate bond defaults and accounting frauds, Sept. 11th, Refco and so forth numerous studies have documented declining volatility in GDP growth both in the United States and other developed economies. Similarly, we have experienced a decline in both the rate and volatility of inflation.
The explosive growth of derivative products the new financial architecture has enhanced liquidity flows across all asset classes. Lower liquidity risk has translated into higher asset prices and lower risk premia. Moreover, risk dispersion should reduce the expected frequency of systemic shocks. As volatility measures for equities, interest rates and currencies have shrunk, our experience to date has matched the theory. The persistence of benign credit conditions, the exponential growth in liquidity sources and the absence of financial accidents has produced a secular change in volatility expectations. The evaporation of Amaranth without a hint of contagion effects only reinforced confidence in the new financial universe.
No financial market theory has ever expired in the midst of a winning streak. And no trigger for a financial shock has ever been foreseen, because in markets the anticipated shock is an oxymoron. Today, manufacturers of the widest range of derivative instruments are looking toward record volumes of issuance. If there is a knowable catalyst for a market reversal, it is not visible to the leading private equity shops, investment banks and leading law firms, all of whom believe they are sitting on record pipelines of future deal activity.
When (John Maynard) Keynes famously said that in the long run we are all dead, he prefaced the remark by noting that (the) long run is a misleading guide to current affairs. In the short term, distinctions between secular and cyclical forces are immaterial. We cannot know the precise relationship between accelerating financial engineering, strong financial market conditions and a benign macroeconomic setting. For manufacturers of the new products and traders, efforts at parsing these interrelationships would seem to be a particularly uninteresting existential exercise.
But from a risk management perspective, distinctions between secular and cyclical conditions might matter, because cyclical forces reverse. Unfortunately, measurement of the cyclical component tends to be difficult except in retrospect. For example, we cannot know how much of the credit derivative universe involves hedging vs. risk taking. In fact, since so-called hedges rely on assumptions about persistent relationships between different securities, trades conceived as hedges may mutate into risks. Nor, importantly, can we know the levels of leverage used by some owners of structured products.
This issue of persistent relationships is particularly important in credit derivatives because valuations are highly sensitive to multiple levels of correlation assumptions. Take a CDO constructed from a pool of underlying corporate credits. Valuation is highly sensitive to a set of assumptions regarding the credit performance of the individual underlying credit assets and the correlation between the performance of those multiple assets. The sophistication of the credit derivative community has grown to a point where there is a general recognition that the correlation between assets can demonstrate significant time-variance. But this theoretical recognition doesnt change the fact that while structured credit products and credit derivatives have grown to a point where these markets are now perhaps the dominant force in steering the direction of global credit markets, this development happened so rapidly that we have a paucity of empirical data on its implications.
No quantitative modeling innovation can change the fact that we simply dont know how the enormous growing role that credit derivative products play in the global financial architecture has altered the fundamental correlation assumptions upon which the entire edifice is built. For example, has the capacity of credit originators to hedge exposure changed underwriting incentives so that past credit performance data is less relevant than is priced into modeling assumptions? Will correlations between different credit markets once thought unrelated show greater linkage in an economic downturn now that multiple structured and derivative products tie diverse credit sectors together in new complex and non-transparent ways precisely because they were historically unconnected? Considering the future performance of a complex instrument such as a synthetic CDO-squared where minor variations in any of the multiple layers of correlation variables can lead to very different determinations of value.
From a broader perspective, there is reason to suspect that principal-agent incentive misalignment could lead to potentially problematic outcomes. There are obvious moral hazard problems imbedded in lenders and manufacturers of structured products selling exotic new instruments to yield-hungry investors. In addition, many credit derivative products such as credit default swaps are structured so that a protection seller collects a premium payment under most likely scenarios with a significant payout loss under conditions of extreme volatility. An asset manager can look very smart for a long time selling out-of-the-money protection for a stream of premium payments. But the economics of the capital markets industry create incentives for managers to perceive a personal rationality in selling protection for statistically unlikely events, proper pricing or not. While ultimately destructive of economic value, it can take a very long time for this to reveal itself especially under conditions of high information asymmetry between investors and managers. Worse, some managers may believe that they are actually creating value while really just selling under priced insurance. Most importantly, under current arrangements rating agencies generate fees collaborating with manufacturers of structured products to create instruments which do not appear to offer the risk protection associated with high investment-grade ratings.
We also cannot know how much liquidity flows depend on particular conditions for example, nearly a trillion dollars in annual excess savings from oil-producing and Asian nations (which, incidentally, can be levered multiple times) or carry trades denominated in the Japanese yen or Swiss franc, or the presumption that those currencies cannot appreciate in value, or the sustainability of low nominal and real interest rates.
Risk modeling systems conventionally use past outcomes to predict the distribution of future returns. While current benign conditions may persist, I would suspect that recent history obscures potential risks. In recent years, shorting volatility and owning credit exposure have been winning strategies. Regrettably, narrower spreads and declining volatility require higher leverage to generate a targeted return. As the popularity of the strategy grows, volatility and spreads contract further, heightening the demand for leverage. Non-participants adopting a more defensive posture underperform benchmarks, which entails economic and career risk. In addition, they will appear paranoid.
History suggests that in the not distant future the paranoid, if they still have assets under management, will become fashionable. The reversal will involve a steep rise in risk perceptions, declining prices and higher volatility as players reduce their exposure to losses. The ensuing margin calls and contracting risk appetites will create contagion effects, deepening the spiral. With the large pools of opportunistic capital raised in recent years, investors with liquidity may rapidly take advantage of price dislocations, thus mitigating market disruptions. But the truth is we cannot know how todays untested structure will operate under conditions of severe stress.
The belief that Federal Reserve policy will reliably function as a stop-loss program for investors is an obvious source of moral hazard. But the exponential growth of structured finance products, which has substantially exceeded traditional liquidity measures, raises a question: has the relative power of central banks in determining money supply and managing credit cycles been diminished? While there are multiple possible explanations, it is interesting to note the anomaly of global monetary tightening accompanied by increased liquidity and lower risk premia across credit markets. What are the implications of the apparent disconnect between monetary policy and liquidity? How much of this disconnect is due to credit derivatives weakening the transmission mechanism between central bank monetary policy and economic output?
The presumption is that the disintermediation of credit risk outside the banking system is a stabilizing force. The flip side is: 1) traditional credit intermediaries are now less attentive to loan covenants and other measures of credit risk; 2) the new holders of credit risk are not credit analysts; 3) concentrations and distributions of credit-risk exposure are now unregulated and unknowable; 4) the leverage utilized by holders of credit risk is unknowable; 5) the persistence of abnormally benign credit experience intensifies pressures to manufacture more product and increases leverage; and 6) while the steepness of the decline in credit conditions may occur slowly, the credit cycle has peaked credit performance will not improve.
In the current environment, liquidity appears to be an infinitely renewable resource. In reality, financial engineering cannot alter the dependence of investor confidence on low default rates. Liquidity and credit quality are not independent variables. When perceptions of credit risk deteriorate, liquidity will contract. When liquidity contracts, credit losses will rise. While the effects to date have been entirely benign, the economic payoffs for originators and packagers of structured credit, and in particular the rating agencies, have created a new form of moral hazard. The record deterioration of underwriting standards for home loans, especially in the subprime area, are likely to call the validity of investment grade ratings into question. If that scenario were to unfold, the ripple effects across other levered portfolios could exceed the imagination of some market participants. In evaluating the currently lucrative opportunities created by a transformed financial architecture, we should be vigilant about those so-called outlier tail risks. A period of low interest rates and abnormally benign credit experience has undoubtedly produced substantial leverage among many financial industry players. While the catalyst and timing are never knowable, the illusion of liquidity as a durable feature of capital markets is likely to dissipate.
What are the consequences of the increasingly prominent role of hedge funds in structured finance products? To date, hedge funds have served as a source of liquidity and thus enhanced market efficiency. There are, however, issues. Hedge fund leverage appears to have increased. Moreover, there appear to be increased correlations in performance among hedge funds pursuing investment strategies, which are ostensibly uncorrelated. Whatever the source of increasingly correlated performance, the phenomenon raises the risk of redemptions by hedge fund investors under conditions of market stress. The attraction of pension assets for hedge funds is that pensions and endowments seem like a stable source of funding. But the reality is that for public pension systems, hedge funds are intensely controversial and prime targets for political scrutiny. The hedge fund industry which exists today is larger by orders of magnitude than the industry which existed when we last experienced conditions of financial stress. The argument that hedge funds as a class engage in high-risk activity is inconsistent with all the historical evidence. But it seems unreasonably optimistic to think of hedge funds as entities backed by stable sources of capital. We have a new financial architecture; it is materially dependent upon hedge funds which have assumed a major role in significant corners of the capital markets universe; and the hedge funds may to some degree be financed by investors with a limited understanding of hedge fund exposure and who may themselves be subject to redemptions by other investors. When institutional investors submit redemption notices in the wake of losses, they call it risk management. When hedge fund managers receive these notices, they say the hot money is leaving. It would seem presumptuous to be confident how this dynamic would unfold under conditions of severe market stress.