Market Cycles and the Performance of Relative-Strength Strategies

Abstract:

In a two-state market with state-contingent mean stock returns, we study trading strategies based on past relative return strength. As compared to the profits implied solely from the cross-sectional variation in unconditional mean returns, we show analytically how market cycles can naturally contribute to generating both higher medium-run and lower long-run relative-strength profits. Empirically, we study both firm-level and industry-level symmetric relative-strength strategies at the medium-run 6 and 12 month horizons and the longer-run 18, 24, and 36 month horizons. We present evidence that supports our analytical framework's empirical relevance: (1) the average profits decline with a strategy's horizon; (2) there is an appreciable commonality in the time-series behavior of the firm-level and industry-level strategies; with all 10 strategies exhibiting average payoffs that are appreciably higher when the ranking and holding period are from the same market state, as compared to when the ranking and holding period are from different market states; (3) the observed market-cycle durations, the cross-sectional variation in state-contingent mean returns, and the cyclical variation in autoregressive return behavior all supports our analytical framework's empirical relevance; and (4) for all ten strategies, the market's lagged cross-sectional dispersion in stock returns, suggested as a leading state variable in past literature, is negatively related to the subsequent relative-strength payoffs.

Source: Stivers, Christopher T. and Sun, Licheng, Market Cycles and the Performance of Relative-Strength Strategies (October 6, 2011). Available at SSRN: http://ssrn.com/abstract=1940148