Abstract The "January effect" refers to the phenomenon that risk-adjusted January returns are unusually large for small firms. Prior studies of year-end tax-loss selling and the January effect lack a control for institutional trading. This study extends prior research and excludes the effects of institutional trading by using a unique data set, the average daily transaction prices for New York Stock Exchange odd-lot sates. Dividing the average daily transaction price of odd-lot sales over the "last half" of December by the average price over the "first half" of January forms the DEC/ JAN ratio. This study addresses (1) whether the DEC/JAN ratio behaves as a proxy for the intensity of year-end tax-motivated trading by individual investors and (2) whether the DEC/JAN ratio and the January effect are positively related. For the 1962-85 year-ends, the DEC/JAN ratio exhibits a pattern consistent with tax-loss selling by individual investors of increasing intensity as the end of December approaches. At the year-end 1986 just prior to the elimination of the long-term capital gain deduction, the DEC/JAN ratio captures gain realization in December by individual investors. Using the DEC/JAN ratio as a proxy for the intensity of year- end tax-loss selling, this study provides evidence that tax-loss selling and the January effect were positively related during the 1962-85 period. However, the data of this study are not inconsistent with prior research that found evidence that the January effect may also be attributable to events other than tax-loss selling.
Koogler et al. (Tue,) studied this question.