At times, commodity futures can turn out a strong performance. When they do, they seem to make a very appealing addition to an investment portfolio. But is the addition of an investment in a commodity futures index really such a great idea?

Proponents of commodity futures assert that the investment vehicles offer a positive average excess return. Additionally, it has been asserted that the addition of commodity futures to a conventional equity and fixed income portfolio can significantly reduce portfolio risk with added diversification, and deliver a great hedge against inflation.

If, through an investment in commodity futures, an investor can increase returns, lower risk and hedge against inflation, why wouldn’t we all buy them? And to further that point, why would anyone ever sell them? Both of these questions can be answered in a study conducted by Truman Clark, former professor of finance at University of Southern California. Clark’s analysis of commodities futures investments is titled “Commodity Futures in Portfolios” and was published for limited distribution to institutional investors and financial advisors in 2004.

Clark’s in-depth study of commodity futures includes experiments using the Goldman Sachs Commodity Index (GSCI), the industry benchmark. He stated that not one of the claims regarding excess returns, reduced risk or hedge against inflation was strongly supported by the empirical evidence.

GSCI Analysis

The GSCI contains 19 allocations to commodities with everything from metals, harvest food products, livestock and a hefty dose of petroleum and natural gas. Essentially, if you can mine it, grow it or put it on the BBQ, it’s in the index.

Daily reporting of the GSCI returns began in 1991. Goldman Sachs provides simulated returns histories for the 21-year time period from 1970 through 1990, but those returns carry selection bias. The figure below shows that the cumulative total returns for the GSCI since 1991 lag the returns of the S&P 500 Index, Long-Term and Intermediate Government Bonds, and just barely outperformed the One Month Treasury Bills for the 13-year 6-month period from January 1991 through June 2004. Of course, investors who cannot take on the risk associated with the S&P 500 Index would accept returns that come with lower risk. But, this is not the case with commodities. The standard deviation of returns for the GSCI was 16.62%, higher than the 14.51% for the S&P 500 for that same period.

The risks and returns associated with commodities don’t create a compelling argument for an allocation to them. However, many investors argue that commodities add an extra element of diversification to a portfolio, citing a negative correlation between commodities and equity or fixed income indexes. Clark tests this theory. He discovered that sample correlations between GSCI total returns (GSTR) and the returns of several equity and fixed income indexes shift depending on whether monthly or quarterly data are used. When monthly data are compared, the annualized standard deviation of returns for the GSTR is 17.7%, with the correlations between GSTR and the returns of other assets being slightly positive or near zero. These results suggest that there is no increased diversification benefit to be gained by combining commodities with stocks and bonds. Quarterly data reveals a slightly different picture as the correlations between GSTR and stock and bond returns appear to be negative or near zero and the standard deviation of returns at 16.4%.

Aware of this differential, Clark set out to quantify the diversification benefit of commodities. He constructed three base portfolios and added fully collateralized commodities futures to each. He identified the combination of stocks, bonds and commodities that would minimize standard deviation and hold the average return constant. In his models, Clark applied the quarterly data for commodities correlation in order to give them their best chance of producing economically significant reductions in standard deviation. Clark’s conclusions revealed that “in all cases, the reductions in volatility are trivial. The annualized standard deviations fall by at most seven basis points. The resulting increases in annualized compound returns are one basis point or less.” He summarized the results of his study by stating, “The potential diversification effects of commodity futures are meager.”

Finally, many investors are drawn to commodity investments because they are considered to be an effective hedge against inflation. Quarterly data shows that GSTR are correlated positively with inflation, while equity and fixed income indexes are negatively correlated. These data suggest that commodity futures can provide an effective hedge against inflation. Once again, Clark tested this assertion. Using the same model portfolios identified in the diversification study analysis, Clark repeated the same application using real returns. For each base portfolio, Clark’s objective was to minimize the standard deviation of real returns while maintaining the average real return. Again, these samples used quarterly statistics because they offered a best case scenario for the inflation hedge theory.  The results of Clark’s experiments showed that the reductions in the standard deviation of real returns mirrored the results of the diversification study which revealed a reduction of seven basis points in standard deviation and an annualized compound rate of return of only one basis point, or less.  Clark concluded, “the potential of commodity futures to serve as effective inflation hedges is trivial.”

In summary, as pertains to commodities allocations, Clark’s study concluded the following:

First, the average excess return of the GSCI over T-Bills was indistinguishable from zero. The average expected return of fully collateralized commodity futures may not be any greater than the Treasury bill return.

Second, an investor who added the GSCI to his/her equity and fixed income portfolio had very limited potential to reduce standard deviation to achieve a constant level of average return and dampen volatility. “Commodity futures do not appear to be “good diversifiers” for stock and bond portfolios,” Clark concluded.

Third, despite the GSCI’s positive correlation with inflation, adding the GSCI futures to a portfolio of conventional assets produces negligible reductions in the standard deviation of real returns and no appreciable hedge against inflation was identified.

Clark’s final assessment of the promise of commodity futures states, “The evidence indicates that the purported benefits of commodity futures are exaggerated... Investors acquiring commodity futures in expectations of higher returns, lower risk, and improved inflation protection are making bets. Current evidence indicates that the odds are against them.”