Testing Whether Futures Skew Predicts Returns

Summary

This analysis asks whether futures with more negative return skew earn higher returns, both across assets and when skew changes over time. It estimates skew from percentage returns after filtering extreme volatility-normalized observations, then uses repeated resampling to show how uncertain skew and mean-return estimates can be. The document reports that many assets appear negatively skewed, but confidence intervals are wide, especially for assets with outliers or extreme skew.

For forecasting, it compares subsequent returns and standard-deviation-adjusted returns after sorting observations by measured skew. It also considers skew relative to an instrument’s history, the current cross section, and the asset-class average. The reported association with future risk-adjusted performance is smaller than the association with raw returns and is heavily influenced by volatility markets. Predictive effects persist under several comparisons but mostly disappear against the asset-class average. These are historical findings with sampling uncertainty; the excerpt does not establish a robust live strategy or account fully for costs and implementation.

Key ideas

• Resampling illustrates substantial uncertainty in estimated skew, especially for outlier-prone assets.
• The document reports that assets with lower average skew tended to have higher long-run returns.
• The relationship is weaker for Sharpe ratios and is driven largely by volatility markets.
• Lower recent skew is reported to precede better returns over several measurement horizons.
• The predictive pattern mostly vanishes when skew is measured relative to the asset-class average.

Tags

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