We occasionally select for retrospective review an all-time “best selling” research paper from the past few years from the General Financial Markets category of the Social Science Research Network (SSRN). Here we summarize the February 2005 paper entitled “Facts and Fantasies About Commodity Futures” (download count over 12,000) by Gary Gorton and Geert Rouwenhorst. Commodity futures are derivative, short-maturity claims on real assets. Many commodities have pronounced price/volatility seasonality. In this paper, the authors compare and contrast the basic properties of commodity futures, equities and corporate bonds. Using monthly returns for stocks (the S&P 500 index), corporate bonds and a broad equally-weighted index of for commodity futures over the period July 1959 through December 2004, they conclude that:
- Both commodity spot and futures prices outpace inflation. Future prices substantially outperform (but correlate strongly with) spot prices, accruing a risk premium for holding period uncertainties. The average historical commodity futures annual risk premium of about 5% is roughly equal to that of stocks.
- Commodity futures historically offer about the same return (0.89% per month) and Sharpe ratio as equities. (See the chart and first table below.)
- Returns for commodity futures correlate negatively with those of both equities and bonds, especially over long horizons, thereby offering diversification benefits. (See the second table below.) When stocks perform worst, commodity futures generate above average returns.
- Returns for commodity futures correlate positively with inflation, unexpected inflation and changes in expected inflation, especially over long horizons. Opposing reactions to unexpected inflation is an important source of negative correlation between commodity futures and stocks.
- The negative correlations between commodity futures and stocks/bonds depends substantially also on different behaviors over the business cycle. Commodity futures outperform (underperform) late (early) in expansions and early (late) in recessions. (See the third table below.)
- Commodity company stocks are not a close substitute for positions in commodity futures. The returns of these stocks lag those of futures, and the returns of these stocks correlate better with those of other stocks than with those of futures.
The following chart, taken from the paper, shows the inflation-adjusted cumulative performance of stocks (S&P 500 index), corporate bonds and an equally-weighted commodity futures index over the sample period. The average annualized return for commodity futures is similar to that of stocks, and both of these asset classes outperform the less volatile corporate bonds.
The following table, also from the paper, provides summary statistics for monthly returns of commodity futures, stocks and corporate bonds over the sample period. The equally-weighted commodity futures index has only slightly lower returns but more significantly lower volatility than stocks (S&P 500 index), so commodity futures have a slightly higher Sharpe ratio. The returns on all three asset classes deviate from a normal distribution, displaying skewness and kurtosis (fat tails). Returns for commodity futures (stocks) are positively (negatively) skewed, meaning that commodity futures (stocks) have more weight in the right (left) tail of the return distribution. Commodity futures have relatively high kurtosis, meaning that they generate extreme returns (in the fat tails) more frequently than expected from a normal distribution. The slightly lower variance and positive skewness for commodity futures indicates relatively less downside risk than for equities.
The next table, also from the paper, summarizes the correlations of commodity futures returns with stocks, corporate bonds and inflation over the sample period. (* indicates significance at the 5% level.) Over all horizons except monthly, the total return for the equally-weighted commodity futures index correlates negatively with those for stocks and long-term bonds, suggesting that commodity futures effectively diversify an equities/bonds portfolio. Diversification benefits grow with the investing horizon. Returns for commodity futures correlate positively with inflation, and this correlation also is stronger at longer horizons.
The final table, also from the paper, compares average annualized returns across the business cycle for commodity futures, stocks and corporate bonds (seven full business cycles over the sample period). On average, commodity futures and stocks and generate similar returns during expansions and recessions. However, during the early part of recessions, commodity futures (+3.74%) substantially outperform both stocks (-18.64%) and bonds (-3.88%). During the late part of recessions, relative performance reverses and both stocks and bonds substantially outperform commodity futures. Diversification benefits of commodity futures are milder during expansions.
The authors also describe the theory and mechanics of speculating in commodity futures.
In summary, commodity futures in aggregate offer a long-term return comparable to that of stocks, with less downside risk and a substantial diversification benefit for a stock/bond portfolio.
Is an equally-weighted commodity futures index versus a capitalization-weighted stock index (S&P 500 index) is a fair comparison?