Is the aggregate sentiment of futures traders predictive for asset returns? In the June 2008 update of their paper entitled “How to Time the Commodity Market”, Devraj Basu, Roel Oomen and Alexander Stremme investigate whether information in the weekly Commodity Futures Trading Commission’s Commitments of Traders (COT) reports enable successful timing of U.S. equities and commodities markets. These reports aggregate the size and direction of the positions taken by different categories of futures traders in different assets. “Commercial” traders use futures contracts for hedging, “non-commercial” traders use them for other types of speculation and “non-reportable” traders operate below the reporting threshold. The study seeks to exploit “hedging pressure” (the fraction of positions that are long) for each of six liquid commodities (crude oil, gold, silver, copper, soybeans and sugar) and for the S&P 500 Index. Each Friday, the six trading strategies studied: (1) take a long position in a commodity if hedging pressure for both the commodity and the S&P 500 Index are below their 52-week averages; or, (2) take a long position in the S&P 500 Index if hedging pressure for both the commodity and the S&P 500 Index are above their 52-week averages; or, (3) hold 3-month U.S. Treasury bills. Using COT reports and associated weekly futures prices for October 1992 through December 2006, they conclude that:
- Timing strategies based on commercial hedging pressure work best for gold and silver during the 2000-2006 test period, generating annual alphas of 11% and 19%, respectively, relative to the S&P 500 Index. An equally-weighted portfolio of all six commodity strategies has low volatility (because of low correlations between individual commodities) and an alpha of 11% relative to the S&P 500 Index.
- Results based on non-commercial hedging pressure during 2000-2006 are similar. Non-commercial hedging pressure appears to be more (less) informative than commercial hedging pressure for copper, crude oil and soybeans (silver and sugar).
- The study also constructs benchmark trend-following strategies optimized via data mining that: (1) take a long position in a commodity if its prior-week return is below its 52-week average and the prior-week return for the S&P 500 Index is above its 52-week average; or, (2) take a long position in the S&P 500 Index if its prior-week return is below its 52-week average and the prior week return for the commodity is above its 52-week average; or, (3) hold 3-month U.S. Treasury bills. An equally-weighted portfolio of the six benchmark strategies has low volatility but an alpha of only 6.7% relative to the S&P 500 Index during 2000-2006. Over the test period, hedging pressure timing works considerably better than trend following (see chart below).
The following chart, taken from the paper, shows the cumulative returns during the 2000-2006 test period of equally weighted portfolios of the six individual commodity-S&P 500 Index timing strategies based on commercial hedging pressure (blue line), non-commercial hedging pressure (green line) and benchmark trend following (red line). Hedging pressure timing beats trend following over this period. Commercial and and non-commercial hedging pressure are in aggregate about equally informative.
The study apparently ignores any trading frictions. Note that the overall sample period (about 14 years) and especially the test period (five years) are not long for strategies with trading rules involving a 52-week moving average.
In summary, evidence from a limited sample period indicates that information in Commitments of Traders reports regarding aggregate positions (hedging pressure) of commercial and non-commercial futures traders may support successful timing of commodities and equities markets.
