Estudo empírico sobre anomalias ligadas a eventos de calendário na Bolsa de Valores de São Paulo no período de 1987 a 2002

Abstract

This research aims to analyze the hypothesis of manifestation of a set of capital market imperfections linked to calendar events in the Brazilian stock market during the period from 01/01/1987 to 12/31/2002. Only the four most important effects linked to calendar events were studied; the "January" or "Month-of-Year" effect, which attributes exaggerated returns to some specific month of the trading year, most often January; the "monthly" effect, which manifests greater gains during the first fortnight of trading; the "Turn-of-the-Month" effect, which attributes a higher average return to the first four trading days of the month and; the "Weekend" effect, which recognizes unexplained losses on the first day of the trading week. The analyzed data are intended to reveal not only the manifestation or otherwise of these anomalies, but also to detect changes in investor behavior with regard to them. The study compares two arbitrary eight-year subperiods: the pre-Real plan period, containing data referring to the eight years prior to the beginning of the Real plan (01/01/1987 to 06/30/1994) and the post-Real period, which contains data referring to to the first eight years of management at Real (01/07/1994 to 31/12/2002). The results reveal an inexplicable pattern of behavior by investors in the Brazilian stock market, with the manifestation of three of the four anomalies studied for the period in question. Such "imperfections" or "anomalies" were treated statistically in order to suggest a probable pattern, although not certain, in the behavior of investors in the Brazilian capital market during certain calendar events.