Search

Following the Brown-Warner simulation approach and using Chilean daily security returns data, we examine the specification and power of three parametric t-tests commonly used in event studies: the standardized the cross-sectional and the portfolio t-test. Our findings show that even though symptoms of nonnormality in security returns and security abnormal returns persist even at the portfolio level, methods based on the use of parametric tests for samples of ten or more securities are well specified, at least at a significance level of 5%. In terms of power, our simulation results show the standardized t-test is always more effective in detecting the presence of an abnormal return than its two competitors: the cross-sectional and the portfolio t-test. We also find however, that the power of the three t-tests is very sensitive to both the sample size and the length of the event period. In particular, conclusions from event studies conducted in the Latin American equity market involving multiday event periods have to be taken with caution. [ABSTRACT FROM AUTHOR]