Amplifying Momentum Returns with Idiosyncratic Volatility
Posted in Momentum Investing, Volatility Effects
March 5, 2010
Does positive feedback trading, indicated by an adjusted measure of return autocorrelation, enhance momentum profitability? In the February 2010 version of their paper entitled “Positive Feedback Trading Activities and Momentum Profits” [apparently removed from SSRN], Thomas Chiang, Xiaoli Liang and Jian Shi examine the relationship between positive feedback trading and profitability of momentum strategies. The momentum parameters for their investigation are a six-month ranking interval followed by a six-month holding interval. Measurement of positive feedback trading is for a six-month window coinciding with the momentum ranking interval. Using daily stock return data for a broad sample of U.S. stocks spanning 1985-2005, they conclude that:
- Stocks that exhibit positive feedback trading generate much higher average momentum profits than stocks that do not (although profits are still substantial for stocks without positive feedback trading).
- Stocks exhibiting past positive feedback trading are much more likely to be extreme losers than extreme winners, perhaps due to stop losses and portfolio insurance policies. Amplification of momentum returns from positive feedback trading therefore derives largely from shorting losers.
- However, stocks in extreme winner momentum portfolios exhibit the highest concentration of positive feedback trading six months later.
- Both positive feedback trading and momentum profitability increase with market volatility and idiosyncratic volatility, with idiosyncratic volatility dominating. Controlling for idiosyncratic volatility, the difference in momentum profitability between stocks that do and do not exhibit positive feedback trading nearly disappears. In other words, idiosyncratic volatility is a good proxy for the level of positive feedback trading (see the chart below).
- Positive feedback trading relates positively to firm credit risk and negatively to firm size.
The following chart, constructed from data in the paper, compares average monthly momentum returns over six-month holding periods for different levels of idiosyncratic volatility. An initial sort ranks stocks into three groups by idiosyncratic volatility (stock volatility minus beta-adjusted market volatility) over the past six months. A second sort ranks each idiosyncratic volatility group by momentum (cumulative return over the same six months). Average monthly returns are for a six-month holding period that immediately follows the six-month ranking period.
Results show that a hedge portfolio that is long (short) past momentum winners (losers) performs best within the third of stocks exhibiting the highest idiosyncratic volatility. Results suggest that there may be both fundamental (earnings growth persistence) and behavioral (reaction to returns) components to this amplified momentum return.
In summary, evidence suggests that investors may be able to enhance exploitation of downside momentum by focusing on stocks with high idiosyncratic (non-beta) volatility.
