Are momentum profits driven by the cross-sectional dispersion in expected stock returns?

Consistent with the hypothesis that momentum profits are attributable to the cross-sectional dispersion in expected returns, Bulkley and Nawosah (2009) report that momentum is nonexistent in demeaned returns. Motivated by their work, I examine whether absence of momentum in demeaned returns is robust to methodological adjustments that mitigate microstructure biases. I find that with commonly employed techniques including skipping a month between the formation and holding periods and excluding firms priced less than $5 (penny stocks) from the sample, the mean monthly momentum profit in demeaned returns increases from -0.37% to 1.02% over the 1963 to 2006 sample period. The results highlight the critical importance of using microstructure screens in empirical momentum studies.