EFF/KRF: There are academic papers that address this question. Some papers say that the differential exposures of stocks to a liquidity risk factor account for differences in average returns. We are far from convinced. The problem is that exposures to this risk factor are estimated quite imprecisely. They look like statistical (random) noise, and noise can't be the source of real differences in expected returns. (You can't explain something with nothing.) From a more practical perspective, if estimated exposures to a risk factor are just statistical noise, they cannot be used to improve investment decisions.
Eugene Fama and Ken French are members of the Board of Directors of the general partner of, and provide consulting services to Dimensional Fund Advisors LP.