| Abstract: | In this paper the meaning of the January effect is clarified. The early literature indicates that differing views on the existence of such an effect were based on different data sets, various time periods, and different measurement tools. A model is formulated to test the January effect hypothesis with seven assets from 1871 to 1986, including tests performed for subperiods corresponding to different monetary regimes. The hypothesis tests incorporate an unbiased estimator to avoid the biases caused by heteroscedasticity and autocorrelation. Findings suggest that asset return behavior differs among assets and regimes. |
