Building a Stock Factor from Volume Spikes and Price Responses

Summary

This study proposes a monthly stock-selection factor based on how prices respond when individual stocks experience unusually large intraday increases in trading volume. It identifies spikes by comparing each minute’s volume increase with that stock’s daily average and dispersion. It then measures returns and return volatility around spike minutes, comparing each stock’s response with the market cross-section. Rolling averages and variability measures are combined into two components, which are then equally weighted into a single factor intended to capture unusually weak or excessive market reactions.

The report describes historical tests on mainland Chinese equities, including industry and size adjustments, style-factor residualization, different index universes, and alternative windows around spikes. It reports strong historical ranking and long-short results, with reduced but still positive results after style controls. These are backtest findings, not evidence of future performance; the authors warn that market conditions and factor drivers can change, and the construction depends on intraday data and specific choices for defining spikes and response windows.

Key ideas

• The method flags intraday volume increases that exceed a stock’s daily average increase by one standard deviation.
• It measures returns and return volatility around flagged minutes, then compares each stock’s response with the cross-sectional average.
• Rolling summaries of the response measures are combined into a monthly stock-selection factor.
• The report presents historical tests across several equity universes and after controls for common style factors.
• The authors caution that historical relationships may weaken or fail as market conditions change.

Tags

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