Removing Time-Series Persistence from a Stock Volatility Factor

Summary

This Chinese equity research summary proposes refining a traditional idiosyncratic volatility factor by removing its cross-period cross-sectional dependence. The starting factor is derived from the Fama–French three-factor model. The report argues that volatility clustering causes a stock’s volatility signal to persist, so monthly selection can reuse stale information and weaken the value of the current observation. Its adjustment adds a regression step intended to isolate fresher selection information.

For A-shares over January 2005 through April 2020, the summary reports that the adjusted factor had a monthly mean IC of -0.055 and a long-short quintile information ratio of 2.17, with an 18.89% annualized return and 8.29% maximum drawdown. It also says the adjusted factor was less correlated with turnover and retained some selection ability after turnover was removed. These are historical single-factor backtest results; the document cautions that future markets may differ and that practical use requires risk management.

Key ideas

• The baseline idiosyncratic volatility factor is estimated from the Fama–French three-factor model.
• Volatility clustering can create persistence in a stock’s volatility signal across periods.
• A regression step is proposed to remove cross-period dependence and emphasize fresher factor information.
• The reported historical A-share backtest shows improved long-short information ratio for the adjusted factor.
• The factor’s relationship with turnover reportedly falls, though the evidence is historical and single-factor.

Tags

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