As the perception grows that traditional asset classes will not continue to have the same lofty returns of the 1980s, there has been an increased interest in the area of "alternative" investments as both a diversifier and a return-generator. As part of this, there has been a marked increase in interest in the area known as "managed futures."
Managed futures as an industry is not new. Begun in the 1970s, it is now estimated $22 billion is invested in managed futures strategies. However, it has been a predominately retail industry, with only an estimated $1.5 billion coming from the institutional community. Institutions have been reluctant to make investments for a number of reasons, including high fees and the almost universal use of leverage. Another subtle, but significant, issue has been the perceived lack of a sound economic rationale for earning investment returns, other than through manager skill. Without that rationale, a rationale independent of manager skill, widespread and long-term acceptance is unlikely.
Curiously, there has been precisely this kind of economic argument around the academic community for a number of years. However, until recently it has not been widely discussed among investors.
The foundation for this argument is the fact that the fundamental economic purpose of futures contracts is to serve as a tool for commercial hedgers to transfer price risk. Given that, it goes on to postulate that investors, ("speculators" in the parlance of the futures industry), are paid for accepting the price risk that hedgers wish to shed.
Investors become "owners" of price volatility. Thus, investors function akin to an insurance company providing "price insurance" to hedgers in the futures market and earn a commensurate return. This is termed as providing a risk transfer service to hedgers. It is interesting that in the futures industry, this basic function is widely mentioned in literature put out by the futures exchanges as the primary reason for their existence, but there is virtually no mention of this as a source of return. Moreover, it is similarly absent from managers' explanations of their returns.
Instead, the emphasis is placed on trading skill and the pure attraction of the high returns and lack of correlation. Aside from diminishing the skill argument, a major issue up to now with the risk transfer argument has been that the historical tests of the viability of this argument carried out by the academic community have been mixed at best. Recently, though, given more widespread understanding of strategies for trading volatility, it appears these studies have used flawed testing procedures. Thus, it is not a surprise that the conclusions reached have been quite mixed. The problem generally stems from issues in the specification of their return series.
John Maynard Keynes in 1930 originally articulated the theoretical argument for a risk transfer premium in the futures market. In addition to specifying the theory, however, he also specified that he expected the premium to manifest itself as a consistent downward price bias in the futures relative to spot prices. This postulated price bias he termed "normal backwardation."
Not surprisingly, a number of authors - including Lester G. Telser in 1958, Hun Y. Park in 1985, Timothy Bresnahan and Pablo Spiller in 1986, and Mark Rzepczynski in 1987 - have investigated the existence of the risk premium by looking for this systematic price bias. The problems with using this approach are at least two-fold.
First, generally futures prices are set at the margin by arbitrageurs through fairly simple relationships based on the spot price, interest rates and carrying costs for the underlying commodity. Under those conditions, one would expect any significant price bias to be arbitraged away to some extent. The second, and more serious, problem is that this argument postulates a systematic downward price bias. This bias would only hold if the excess demand for hedging were always from hedgers needing to go short to hedge their risks. But it is clear commercial entities may need to go long or short in order to hedge their risks depending on whether they are a producer or consumer of a particular commodity. This means an imbalance in the demand for hedging can be either on the long or short side. Therefore, a priori, there is no reason to expect a particular systematic price bias.
A second technique has been to use a constructed return series using historical prices. Katherine Dusak in 1973, Jennefer Baxter, Thomas E. Conine Jr. and Maurry Tamarkin in 1985, Jacky So in 1987, Michael Ehrhardt, James Jordan and Ralph Walking in 1987, and Emmett Elam and Daniel Vaught in 1988 have all used this approach in their investigations. This use of a constructed return series would be much better if it were not for one major problem.
The return series used typically assumes a long-only position held in futures through time. Clearly, this suffers from the same implicit flaw as the price bias methodology mentioned above. The return series used in testing the risk transfer concept must reflect the fact the demand for hedging can be either on the long or short side. Using a return series constructed with a long-only assumption simply cannot work unless one assumes the excess demand for hedging is always from those who are looking to go short in order to hedge. In our view, that is a completely untenable position.
The final technique generally seen is the use of actual positions for both hedgers and speculators as reported to the Commodity Futures Trading Commission in conjunction with historical prices to determine a return series. The investigations using this methodology - including H.S. Houthhakker in 1957, Charles Rockwell in 1967, Colin A. Carter, Gordan C. Rausser and Andrew Schmitz in 1983, Eric C. Chang in 1985, Michael L. Hartzmark in 1987, and G.S. Maddala and Jisoo Yoo in 1990 - all try to determine whether speculators earn premiums from hedgers over time.
On the positive side, this technique clearly eliminates the long-only assumption because positions taken by these parties can be either long or short. Also interesting, it is this set of papers that provides the most consistent, (but not unanimous), support for the existence of the risk transfer premium.
But there is still a significant problem with this approach. It is well known that many, if not the vast majority, of the speculators operating in these markets take positions based on their own forecasts. In investment parlance, they are "active managers." That means any returns generated, positive or negative, are some mix of both the risk premium and their skill or "alpha."
Unfortunately, these studies provide no breakdown as to how much of the returns are due to either factor. Given that speculators can be long or short at any time, use leverage in varying amounts, and vary the markets in which they take positions, the amount of alpha in the returns can be quite large. This makes the task of sifting out the risk transfer premium from speculator returns extremely difficult. Unfortunately, without such a breakdown, it is impossible to know whether and how much was due to any particular factor, including the risk transfer service. It also means one cannot completely rely on the results of these studies, even though they tend to be supportive.
Interestingly, the John Lintner 1983 study often cited by the managed futures industry as supporting the inclusion of managed futures in a traditional portfolio is not a test for the risk transfer premium. It only attempts to look at the narrow issue of including traditional managed futures funds in portfolios of stocks and bonds, rather than the question of an economic rationale for investing. Thus, its usefulness for institutional investors is limited.
What is truly unfortunate is that these flawed techniques have resulted in inconsistent results that in turn have contributed to institutional investors shying away from what now appears to be a significant new investment opportunity. That these flaws have remained hidden is not that surprising because these are well-accepted techniques for examining traditional investments.
However, traditional investments are inherently involved in providing capital formation, while managed futures is a risk transfer service. Separate economic functions call for separate examination techniques. Still, those involved in trading immediately recognize that the implementation of the risk transfer argument must be through a strategy based on the idea of "owning" volatility. Newer techniques reflecting this notion have shown tremendous results when applied to markets with significant commercial hedging activity.
There is a larger lesson to be learned here as well. As more investors look beyond traditional investments, the techniques used in the evaluation process must appropriately reflect the underlying economic function of those investments. Standard research methodologies, based on buy and hold philosophies, are appropriate for capital formation investments but are inadequate for an independent economic function, such as managed futures. Without the appropriate adaptation of research techniques, investors will be poorly informed as new and innovative investment opportunities are created.
Charles A. "Tony" Baker is a vice president in asset management services at SEI Corp., Wayne, Pa.
