Does momentum trading (and its contrarian counterpart) work better for certain kinds of stocks? In their August 2009 paper entitled “Systematic Risk and the Performance of Mutual Funds Pursuing Momentum and Contrarian Trades”, Grant Cullen, Dominic Gasbarro, Gary Monroe and Kenton Zumwalt examine mutual fund trading activity and performance to measure the prevalence of and results for momentum and contrarian equity investing strategies. Using the quarterly stock holdings of 2,829 U.S. equity mutual funds and associated stock price data for the period 1991-2006, they conclude that:
- About 15% (14%) of equity mutual funds substantially follow a momentum (contrarian) trading strategy.
- These contrarian and momentum trading behaviors tend to persist over time.
- Based on simple excess returns, the contrarian strategy beats the momentum strategy on average by 2.2% and 2.5% over post-trade horizons of three and six months, respectively. Outperformance is robust to previous fund return, fund portfolio turnover, fund size and fund portfolio liquidity. However, outperformance disappears on a four-factor (market, size, value and momentum) adjusted basis.
- Mutual funds following a momentum strategy improve their performance by buying (selling) high–beta stocks that have been recent winners (losers). Similar funds that trade low-beta stocks hurt their performance.
- Conversely, funds following a contrarian strategy improve their performance by buying (selling) low-beta stocks that have been recent losers (winners). Similar funds that trade high-beta stocks hurt their performance.
- A possible explanation for these beta-related effects is that pricing for low-risk stocks is more efficient than that for high-risk stocks. Reversion to intrinsic value is therefore faster for low-beta than high-beta stocks, which tend to be small, thinly traded and meagerly covered by analysts.
In summary, evidence from mutual funds suggests that momentum (contrarian) investors/traders can enhance returns by focusing on stocks with high (low) beta.
