Skip to content
All library documents

Idiosyncratic Momentum in Commodity Futures

Article BigQuant

Summary

The article reviews empirical research comparing commodity futures momentum based on total returns with momentum based on residual, or idiosyncratic, returns after accounting for selected systematic factors. The study forms ranked portfolios across commodities and varies ranking and holding periods. Its reported findings favor idiosyncratic momentum: it is more persistent and generally stronger than total-return momentum, particularly in long-only portfolios. Across both approaches, long positions in recent winners drive the results; shorting losers does not produce reliable positive returns. The analysis also finds evidence that hedging pressure and term structure act as commodity risk premia, while size and value measures are less convincing.

The reported tests use monthly futures data for 28 commodities over 1997–2017, with factor position data available for a shorter period. The article reports statistical tests and comparisons with a broad commodity index, but its findings are historical and do not guarantee future performance. Results depend on portfolio construction, ranking and holding horizons, contract selection, and treatment of futures rolls. The summary presents research evidence, not a ready-to-trade implementation with current costs or execution assumptions.

Key ideas

  • Idiosyncratic momentum ranks commodity futures using returns after removing selected systematic factor effects.
  • The study reports stronger and more persistent results for idiosyncratic momentum than for total-return momentum.
  • Long portfolios of winners account for the momentum effect, while short portfolios of losers lack reliable positive returns.
  • The analysis finds evidence for hedging pressure and term structure as sources of commodity futures risk premia.
  • Reported results use historical data and depend on portfolio design, horizon choices, and futures contract construction.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.