There is considerable investor interest in allocating to commodity futures amid the growing awareness of commodities as an asset class in their own right. So, investors need to know how commodities are expected to perform under different economic circumstances and, consequently, what value they bring to a diversified portfolio.

But how well are the drivers of commodity futures returns known? Plenty of studies have been carried out, but they have used data from only the last few decades. The authors reason that if they could go back a lot further and analyse data since the 1870s, they could gain more useful insight. Going back that far includes many more periods of recession, inflation, and backwardation or contango, which have all been previously identified as key drivers of commodity futures prices.

A significant challenge facing the authors was how to obtain commodity futures data in a useful format. To create usable data, they had to manually transcribe data from 1877 to 1951 from the Annual Report of the Trade and Commerce of the Chicago Board of Trade (CBOT).

The authors calculate annual performance over the period using month-end returns. The performance is measured both by arithmetic and geometric mean returns because volatility can lead to a large difference between the two measures. They then check the liquidity of commodity futures in this period against liquidity after 1951 to make sure the returns are reliable and comparable.

Armed with these data, the authors then create two portfolios of commodity futures. The first is an equal-weighted portfolio, and the second is a long–short portfolio that essentially overweights commodities in backwardation and shorts commodities in contango. A futures market is said to be in contango when futures prices are higher than the spot price and in backwardation in the rarer cases when futures prices are lower than the spot price. One of the best-documented findings for commodity futures returns is that commodities in backwardation outperform those in contango.

The authors seek to discover the effect on portfolio returns when the commodity futures market as a whole is in backwardation or contango during recessions and during unexpected inflation.

The portfolios’ returns are compared with long-term US government bonds and the US stock market.

In terms of absolute performance, commodity returns are substantial at 4.6% a year for the equal-weighted commodity index across the 140-year period of the study. This compares with returns of 6.7% for the stock index and 1.1% for bonds.

The source of this performance is fundamentally commodity price changes, not simply carry—where carry is the return, or dividend yield, when the price of the commodity future does not change. This finding contrasts with previous studies that argue that carry plays a significant role in commodity performance. The authors contend that the monthly spot price change is the more significant driver.

Over most periods of the study, the Sharpe ratio of portfolios is seen to increase when commodities are included. The exceptions are when inflation is falling and during times of recession for the equal-weighted portfolio. The rise in the Sharpe ratio is mainly because of reduced volatility rather than higher returns.

Extracting performance data under different conditions confirms that backwardation, unexpected inflation, and the business cycle stage all have a significant effect on returns.

The authors find that returns average 7.7% in periods of backwardation, compared with 1.8% in periods of contango. Meanwhile, in periods of unexpected inflation, returns are 10.1%, compared with –0.1% in periods of unexpected negative inflation.

Commodities perform very poorly during severe recessions. In the Great Depression (1929–1932), the equal-weighted commodity portfolio underperformed a portfolio of stocks and bonds by –72.3%, and in the global financial crisis of 2007–2009, it underperformed by 38.7%.

But the performance of commodities is not always negative during downturns; the equal-weighted portfolio returned a sizeable 9.5% a year during the longest downturn in the study (1902–1921).

In comparison, the long–short portfolio produced average returns of 5.0% and is found to be positive in all economic environments. 

The authors find that liquidity in the early period of the study was little different from in the later period, so returns from the full sample are not skewed by liquidity issues.

With the conclusion that the movement in the spot price of commodity futures, rather than carry, drives returns, this study clearly shows the diversification advantages of adding commodities to portfolios of stocks and bonds. Managers of multi-asset portfolios may want to consider adding commodity futures to provide returns when other assets are underperforming and to lower overall volatility in the portfolio. While stocks may well be the best investment for long-term investors, stocks can suffer prolonged periods of sub-trend performance. The authors demonstrate that commodity futures prices move to a different cycle.

In addition, the widely held view that commodities are the best hedge for inflation is not borne out by this research. The authors argue that the reason commodities provide lower returns than stocks in high-inflation periods is, perversely, because investors attach considerable value to the inflation hedge that commodities provide. During those high-inflation periods, investors do not require such a high premium for commodities as they do for stocks.