A Stochastic Dominance Approach to Spanning. With an Application to the January Effect

We develop a Stochastic Dominance methodology to analyze whether rational non-satiable and risk averse investors benefit from a particular expansion of the investment possibilities. This methodology avoids the simplifying assumptions underlying the traditional mean variance approach to spanning. The methodology is applied to analyze the stock market behavior of small firms in the month of January. Our findings suggest that the previously observed January effect is remarkably robust with respect to simplifying assumptions regarding the return distribution.