Futures Investment Series S P E C I A L R E P O R T No. 3 The MLM Index Mount Lucas Management Corp.
The MLM Index Introduction 1 The Economics of Futures Markets 2 The Role of Futures Investors 3 Investor Profitability and Efficient Market Theory 4 The Empirical Evidence for Investor Profitability 5 The MLM Index A Return Benchmark 6 Conclusion 11 By Timothy J. Rudderow President Mount Lucas Management Corporation Copyright 1999. All rights reserved. Reproduction with permission only. June 1999 AN INVESTMENT IN FUTURES CAN RESULT IN LOSSES. PAST PERFORMANCE RESULTS ARE NOT NECESSARILY INDICATIVE OF FUTURE PERFORMANCE RESULTS.
The MLM Index I N T R O D U C T I O N Many academic studies have concluded that the returns of investments in futures markets are not correlated with the returns of traditional asset classes. These findings suggest that a managed futures investment is a logical component in a welldiversified portfolio. However, before a prudent portfolio manager can make a specific managed futures allocation, the manager needs to address a series of more general issues concerning the status of managed futures as a legitimate investment area or asset class. Among these issues is whether a futures investment has a clear rationale for profitability that is independent of the particular manager or investment approach under consideration. The distinction between the inherent returns of an asset and the returns earned by active managers investing in that asset is of critical importance. At issue is confidence in continuing profitability. Unless there are demonstrable reasons why a managed futures investment can generate competitive returns without relying solely on the skills of active managers, it cannot be considered a legitimate asset class. This is true regardless of how well particular active managers have performed in the past. The standing of equities, bonds, real estate and other components of modern institutional portfolios as asset classes are validated with indices that quantify the returns of passive investments in those assets. The S&P 500 is a well-known example of a passive index that demonstrates the existence of long run returns to equities. A bona fide index serves as a benchmark both for comparing the performance of an asset with other assets in a diversified portfolio, and for evaluating the performance of active managers of that asset. The purpose of this report is to explain the economic rationale for positive returns to futures investments and to review the development of an appropriate passive index (the MLM Index ) to measure those returns. We also compare the properties of the MLM Index with indices of more traditional institutional asset classes. 1
The Economics of Futures Markets Firms in many industries are exposed to the risk of fluctuating commodity prices in the normal course of their business activity. The futures markets exist as a means to transfer this risk to others, an activity known as hedging. For example, the operator of a grain elevator in the Midwest is in the business of receiving, storing and shipping grain. A competitive profit margin is earned for these services. Normally, the operator buys a large inventory of corn at harvest. If corn prices were to rise subsequently, an unexpected windfall profit would result. However, if corn prices were to collapse instead, the operator might lose his business. Without the ability to hedge, it would be necessary for elevator operators generally to increase turnover margins to compensate for the risk of holding inventory. With the ability to hedge, margins are lower and the corn market is more efficient. offer spreads paid by cash market participants in Treasury Bonds have narrowed sharply since the introduction of interest rate futures. Similar examples apply to all other futures contract markets. In essence, the ability to hedge frees a business to concentrate on its principal activity by minimizing the impact of price fluctuations on operational results. Thus, the futures markets are an insurance medium through which businesses can transfer the risk of price fluctuations to investors who are willing to bear and manage that risk. The futures markets enhance efficiency in the financial services industry as well as in traditional agricultural markets. For example, the existence of the Treasury bond futures contract has allowed dealers to hedge their inventories cheaply and provide clients more accurate pricing information. It is no accident that the bid- 2
The Data Benchmark Futures Data The Role of Futures Investors Since the fundamental economic purpose of the futures markets is to provide the hedge community with a means to transfer price risk, it follows that without hedger participation in a particular futures contract, that market will not thrive. However, the operational efficiency of the risk transfer process also requires a large contingent of market participants who are willing to supply risk capital, i.e., to accept the transfer of price risk from hedgers. These investors are in the business of evaluating price risks. A futures market operates as a zero-sum game. That is, there are an equal number of long and short positions at all times. However, the two major participants, hedger and investor, operate under significantly different agendas. While the investor is motivated solely by profit on futures positions, the hedger is indifferent to futures results since the cash or physical position is hedged. This indifference by the hedgers provides the profit-edge to the investors.¹ In a number of respects, the activity of futures investors is analogous to the insurance industry. An insurance company receives many applications for insurance, but, in general, accepts only those that offer a high probability of profitable returns. Similarly, price fluctuations of the futures markets present profit opportunities to the investor. In a disciplined way, the investor must decide at which prices to accept or reject risks. Unlike the hedger whose market position, long or short, is determined by an underlying physical inventory, the investor is free to assume risks on either side of the market depending on his assessment of reward and risk. ¹We are referring to investors as a group. Obviously, not all investors are profitable any more than all insurance companies are profitable. 3
Investor Profitability and Efficient Market Theory Conjecture that investors earn positive returns is often mistakenly viewed as inconsistent with efficient market theory. This misconception is based on an incomplete characterization of futures markets as zero-sum games. As we have described, these markets have as their primary function the transfer of risk from hedgers to investors. Well-capitalized investors will accept this risk transfer only if they can earn a long-run return. A long-run return to investors does not imply that futures markets lack efficiency in the sense of not quickly discounting complex fundamental information about the expected equilibrium price. Anyone who has participated in futures markets will confirm that information is indeed discounted very, very quickly. The relevant issue is whether futures markets perform their primary function efficiently. If investors receive a return that is consistent with the involved risks, we say the answer must be yes. This simply means that on a risk-adjusted basis, the returns to futures investing are roughly comparable to the returns of other asset classes - not much higher and not much lower. In other words, the futures markets perform their function efficiently by gaining the participation of investors at fair value. 4
The Empirical Evidence for Investor Profitability Not only is the theory of positive returns to commodity investors compelling, but also the empirical evidence in support of the theory is substantial. For example, numerous studies of the Commitments of Traders in Commodity Futures data have shown that large professional traders as a group are in the long run profitable - at the expense of small traders and hedgers. Moreover, there is a consistent and significant departure from statistical normality in the returns distributions of most futures markets. This departure can explain why businesses need to hedge and why investors can earn positive returns. Specifically, the monthly or quarterly returns distributions of prices typically show many more observations in the Figure 1 tail regions of the distributions than would be expected if the distributions were normal. This characteristic is shown schematically in Figure 1 by the distribution with the solid line. By comparison, the dashed line represents a normal distribution. Note that the distribution typical of futures returns has a larger number of observations in the tail regions. Departures from normality are consistent with observations that commodity prices often exhibit major price trends extending over weeks and even months. These large price changes are usually the result of some major shock to the supply or demand situation of a market, e.g., a typhoon in the Philippines that destroys much of the sugar crop (bullish) or a cancellation of a major Soviet grain purchase (bearish). Events such as these, which can move prices quickly and sharply, represent a threat to the operations of the businesses involved in these markets and make hedging a required, on-going practice. In our view, these periods of rapid price change provide investors with their profit opportunities for accepting the risk transferred by hedgers. If this view is correct, it should be possible to construct a passive mechanism, i.e., a simple algorithm, directed toward capturing the profit potential in these price trends. The returns generated will provide a benchmark for the returns available to investors. The same algorithm will provide an estimate of the risks that are associated with the benchmark returns. This line of reasoning led us to develop the MLM Index. 5
The MLM Index : A Return Benchmark A persistent problem in examining the returns of managed futures investments has been the lack of an appropriate return benchmark. Indices exist which measure the returns of particular futures managers. But since managers use leverage and other skill-based trading approaches they are not an appropriate benchmark for the returns of the underlying asset. Other indices measure the returns of a simple buy and hold strategy. However, the buy and hold strategy, while appropriate for other asset classes, such as equities and bonds, fails to capture the basic economic function of the futures markets. Futures investment returns are derived from bearing the risk transferred by hedgers. As previously noted, this function is performed by investors taking both long and short positions in futures markets (to match the requirements of both long and short hedgers). Thus, a buy and hold index is obviously inappropriate. The MLM Index of futures returns uses a simple trend-following algorithm that signals the inception of possible large price trends. The object of the algorithm is to capture the potential profits represented by such trends. Many of these signals will be erroneous in the sense that price trends do not always occur following these signals. However, if returns are distributed as suggested by Figure 1, price trends will occur frequently enough to assure long-run profitability if the algorithm is used consistently over time. The algorithm is applied to both up trends and down trends in price, i.e., it simulates both long and short trades. It is applied to a broad range of markets since the diversification obtained is an important aspect of the investment benchmark. 6
A graph of the simulated historical MLM Index is shown in Figure 2. For 1961, the Index includes 7 futures markets, primarily agricultural markets. For 1995, it is composed of 25 markets including grains, livestock, currencies, metals, interest rates and industrials. The Index does not include the S&P 500. A logarithmic scale is used in Figure 2 to convey the relative rates of return over the 38-year history of the Index. There is nothing magical about the particular parameters used in the algorithm to generate the MLM Index, i.e., 12 months. The results are extremely robust; a wide array of other parameter choices, 3 months through 18 months, all show similar results. Chart 1 compares the returns of the MLM Index with the returns of indices of more traditional asset classes. The chart shows compound annual returns and standard deviations from 1961, the year the MLM Index begins, through 1999. The results indicate that the returns to futures investing are very similar to the returns of any of the other financial assets. Furthermore, volatility, as measured by standard deviation is also in line with the other assets. The infamous volatility of futures markets is the result of the abuse of leverage by some participants. When viewed in an objective manner through an unleveraged, diversified investment, futures market volatility is very reasonable. Figure 2 The MLM Index 1961 through 1999 (Log Scale) 1000000 100000 10000 1000 Jan-61 Jan-63 Jan-65 Jan-67 Jan-69 Jan-71 Jan-73 Jan-75 Jan-77 Jan-79 Jan-81 Jan-83 Jan-85 Jan-87 Jan-89 Jan-91 Jan-93 Jan-95 Jan-97 7
The details of the MLM Index construction 1. Choice of futures markets and contracts. The MLM Index is based on daily closing prices of the nearby contract month of a portfolio of the most active futures markets. Only highly liquid U.S. futures markets are currently included in the Index; the choice of markets for a calendar year is made in the December preceding the start of the year, and markets are not added to or deleted from the Index during a year. If a commodity is traded on more than one futures exchange, only the one with the largest open interest is included in the Index. For example, Chicago Board of Trade wheat has larger open interest than Kansas City Board of Trade wheat; consequently, Chicago Board of Trade wheat is included in the Index but Kansas City Board of Trade wheat is not. 2. Determination of long or short futures position for each market. The rate of return of an individual market depends on whether the market position is long or short. Since a futures contract eventually expires, the MLM Index is based on the unit asset value of a market, rather than on the actual futures price. This month s unit asset value of a futures market is determined by multiplying last month s value by 1 plus the percentage change in this month s nearby futures price. The market position is long during the current month if the market s closing value on the next-to-last trading day of the prior month is greater than or equal to the market s 12-month moving average of closing values; otherwise, the market position is short. A market s 12-month moving average of closing values is defined as the average of the 12 closing unit asset values for the next-to-last day of the 12 months immediately preceding the beginning of the current calendar month. For example, to find the 12-month moving average for June 1992, first find the unit asset value for the next-to-last trading day of each of the prior 12 months, June 1991 through May 1992. The average of those 12 values equals the 12-month moving average. 8 3. Calculation of the monthly rate of return for each market. If the market position is long, then the market monthly rate of return equals the percentage change in the market price during the month, i.e., the market monthly rate of return (%) equals the closing price of the current month divided by the closing price of the prior month, minus 1, times 100. If the market position is short, then the market monthly rate of return (%) equals -1 (minus one) times the percentage change in the market price during the month, i.e., the market monthly rate of return equals the closing price of the current month divided by the closing price of the prior month, minus 1, times -100 (minus 100).
4. Calculation of the monthly rate of return for the MLM Index. The monthly rate of return of the Index equals the simple average of the individual market monthly rates of return plus the T-Bill rate of return. 5. Determination of the MLM Index value. The value of the MLM Index is computed by compounding the Index monthly rates of return. The beginning value of the Index is defined to be 1000 in January 1961. Each month thereafter, the Index is changed by the monthly rate of return. That is, each month s Index value is determined by multiplying the prior month s value by 1 plus the current percentage monthly rate of return. The MLM Index Delivery Month Selection Rules Index Position For Established On the Close Moving Average Based On Index Position Determination Uses Rate of Return Based On Current Month Last trading day of prior month Next-to-last trading day of each of the prior 12 months Nearest contract at least two months* from date Index position is established Nearest contract at least two months from current month An Example Using Wheat August July 31 August 30 (last year) September Contract December September 29 (last year) (used through July 31) Contract July 30 (this year) September August 31 September 29 (last year) December Contract December October 30 (last year) (used through October 31) Contract August 30 (this year) *For the following commodities, rolling takes place at the end of the month that is immediately prior to the delivery month, e.g., rolling for a September delivery takes place on the last day of August. Australian Dollars British Pounds Canadian Dollars Euro Currency Japanese Yen Swiss Francs 9
Chart 1 * Return and Volatility for Selected Asset Classes (1961-1999) MLM Index S&P 500 Small Stocks Government Bonds Corporate Bonds High-Yield Bonds Treasury Bills Return STD Deviation 16.08 10.29 13.67 14.65 17.32 20.66 7.41 9.75 8.12 8.68 8.87 7.82 6.10 0.76 0 5 10 15 20 25 Compound Annual Percentage Return As noted earlier, one of the attractions to investing in a futures asset is the low correlation between futures and traditional assets. Chart 2 shows the correlation coefficient between each of the major asset classes and the MLM Index. The MLM Index shows little or no correlation with most asset classes. No other asset combines such strong diversification potential with the liquidity and ease of valuation of a futures investment. Chart 2¹ Correlation Matrix (1961-1999) S&P SMALL GOVT CORP HI-YLD TRSY MLM 500 STOCKS BONDS BONDS BONDS BILLS INDEX S&P 500 1.00 SMALL STOCKS 0.78 1.00 GOVT BONDS 0.32 0.16 1.00 CORP BONDS 0.38 0.24 0.91 1.00 HI-YLD BONDS 0.51 0.50 0.48 0.61 1.00 TRSY BILLS -0.06-0.07 0.10 0.08 0.03 1.00 MLM INDEX -0.06-0.09-0.04-0.04-0.03 0.14 1.00 ¹The data used to develop Charts 1 and 2 are described in Mount Lucas Management s Special Report No. 1. 10
Conclusion The futures markets exist to facilitate the transfer of price risk from hedgers to investors. The business of futures investment is to assume and manage that price risk. The mean and standard deviation of returns to futures investment, as demonstrated by the MLM Index, are similar to other financial assets and the returns of the investment are not correlated with the returns of traditional assets. Thus, managed futures are an asset that every prudent manager should evaluate as a potential addition to an institutional portfolio. 11
Mount Lucas Management Corporation 47 Hulfish Street Suite 510 Princeton, New Jersey 08542 609.924.8868 http://www.mtlucas.com