Testing Whether Relative Volatility Predicts Futures Returns
Summary
This article tests whether recent volatility levels relate to next-month risk-adjusted returns across futures markets. It builds a relative-volatility measure by dividing estimated volatility by a long-run exponential average, then compares next-month returns scaled by beginning-of-month volatility across ranked volatility groups. The author applies the analysis to an S&P 500 contract and broader groups of equity, bond, and foreign exchange futures, contrasting these results with a momentum forecast. The reported patterns vary by market; the stock-level result motivating the exercise does not appear consistently across futures.
The author then proposes a contrarian volatility forecast: higher relative volatility implies a stronger long signal, while lower volatility implies a stronger short signal. The signal uses a historical percentile and smoothing to reduce noise and turnover. The article warns that reversing an initially expected relationship after seeing the data creates data-mining bias, so the backtest may overstate performance. The excerpt describes the method and interaction with trend following, but ends before presenting system-test results. Its evidence is exploratory and does not establish a robust standalone strategy.
Key ideas
- Relative volatility is defined by comparing current estimated volatility with a long-run average.
- The study scales next-month futures returns by volatility known at the start of that month.
- Results differ across equity, bond, and foreign exchange futures, so the motivating stock-level pattern does not generalise cleanly.
- A proposed contrarian rule maps higher volatility percentiles to stronger long forecasts and lower percentiles to short forecasts.
- The rule’s backtest is vulnerable to data-mining bias because its direction was chosen after observing the results.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.