What are the most common strategies for trading commodity futures? In their brief January 2017 article entitled “Commodity Futures Trading Strategies: Trend-Following and Calendar Spreads”, Hilary Till and Joseph Eagleeye describe the two most common strategies among commodity futures traders: (1) trend-following, wherein non-discretionary traders automatically screen markets based on technical factors to detect beginnings and ends of trends across different timeframes; and, (2) calendar-spread trading, wherein traders exploit commercial/institutional supply and demand mismatches that affect price spreads between commodity futures contract delivery months. Examples of the latter are seasonal inventory build and draw cycles (as for natural gas) and precise roll cycles for expiring contracts included in commodity futures indexes. Based on the body of research and examples, they conclude that:
- Successful trend-followers quickly cut losses (false trends) and lever into wins (real trends). Applying the strategy across multiple, diverse markets is key to dampen overall portfolio volatility.
- Calendar-spread speculators pursue:
- “Insurance” premiums by trading against commercial (build and draw) hedgers. This approach tends to have consistent returns, but is of limited scalability.
- Liquidity premiums by frontrunning schedule roll dates of each contract series within commodity futures indexes. Index roll trades create selling pressure for the maturing contract and buying pressure for the next contract. Speculators unwind positions on roll dates.
In summary, the body of research indicates that technical trading to follow price trends and calendar-spread trading to exploit cyclic supply and demand are the most common commodity futures trading strategies.
Cautions regarding conclusions include:
- The paper is a broad overview only. The authors do not quantify magnitudes or consistencies of expected returns of the two strategies.
- Nor do the authors detail how successful trend-followers explore and estimate trend durations or determine which trends are false (to be exited) versus real (to be levered). Backtesting to set trend, stop-loss and leverage parameters is readily subject to data snooping bias.
