The Fundamentals of Commodity Futures Returns (Digest Summary)

Investors with portfolios of commodity futures expect to earn a positive return. The theory of normal backwardation implies that producers need to hedge their commodities and are willing to sell at a discount to the long-term value of the commodity to induce speculators to take long positions. The expected return to long positions in commodity futures increases as inventories decrease and price momentum increases. Ultimately, all returns and the volatility of commodity futures can be traced back to the level of inventories and the related convenience yield. Because consumers of commodities require these raw materials for their production processes, lower levels of inventories increase the probability of disruption to their businesses. Therefore, the price and volatility of commodity futures increase as inventory levels decline.

Commodity market investment products are moving beyond index products that offer beta exposure and into the realm of alpha prediction. Segmenting the commodity portfolio into commodities with higher and lower expected returns can increase the expected alpha of the portfolio but reduce the diversification within the portfolio. Investors who desire to hedge inflation risk while adding alpha to their portfolios should examine this research closely.

Inventory data are correlated with a number of price signals. Commodities with higher-than-average inventory levels tend to have a negative basis, a lower price return, and a higher price volatility over the trailing 12 months. Inventory levels, as well as trailing returns and volatility, can be used to select commodities with higher expected returns. When compared with an equally weighted portfolio, commodities with lower inventories outperform, high-basis commodities outperform, commodities with high trailing price momentum outperform, and in 56% of months, commodities with high trailing price volatility outperform. Commodities with the highest return in the spot market over the trailing 12 months outperform in 58% of all months.

Investors can add substantial returns to their commodity portfolios through these strategies because the commodities expected to outperform earned excess returns between 8% and 12% per year compared with those expected to underperform, which earned an average excess return of less than 1% per year.

The authors measure the risk and return of 31 separate commodity futures markets from 1971 to 2010 and relate the risk and return of each commodity to price and inventory signals. Over this time period, an equal-weighted basket of commodity futures earned an excess return of 5.75%. Unlike equity returns, commodity futures returns are skewed to the right with fat tails; the largest returns occur during times of low inventories.

This research extends that of Gorton and Rouwenhorst (Financial Analysts Journal 2006), but the authors use a slightly reduced set of commodities because the prices of such financial commodities as gold and silver do not have price sensitivity to inventory, sugar and rice do not have monthly inventory figures, and electricity cannot be stored.

Two key hypotheses are tested. First, prices of commodities in the physical spot market increase as the levels of inventory decline. This relationship is explained by convenience yield, which rises sharply as inventories decline to the level at which commodity consumers become concerned about short supplies. Second, risk premiums increase with volatility and decline as inventories rise.

The basis, which is the difference between the futures prices and the spot market prices, is also equal to the difference between carrying costs (interest rates plus storage costs) and the convenience yield. Unless the convenience yield is high, the spot price is typically lower than the futures price, which implies a negative basis. There is a strong relationship between excess returns and the basis, with the highest expected returns occurring in times of a positive basis and a high convenience yield. The basis of more difficult-to-store commodities, such as oil and gas, can be more volatile than that of easier-to-store commodities, such as industrial metals. Energy and food products have substantially greater seasonality, as well as a higher probability of a stockout, than industrial metals.

The inventory variable compares the most recent monthly inventory estimate with the trailing 12-month average of inventories. For all commodities, lower (higher) inventories are associated with an above-average (below-average) basis. Monthly excess returns for a particular commodity are higher when inventory levels are low compared with returns of that commodity at other times.

Data from the U.S. Commodity Futures Trading Commission Commitments of Traders reports are used in tests. Although commercial hedgers typically have net short positions, the size of these net positions does not predict future returns in a meaningful way.

As greater levels of assets move into commodity markets, the demand for both passive and active management strategies will increase. Historically, commodities have appeared to be less efficiently priced than equities. The question, however, is how these predictable differences in risk premiums will hold up in light of the increasing asset level and investor attention in the commodity market.