Using daily stock returns: the case of event studies

Abstract

journal of financial economics 14 (1985) 3-31. north-holland using daily stock returns the case of event studies* stephen j. brown yale universiry. new haven, ct 06520, usa jerold b. warner universrty of rochester, rochester, ny 1462 7, usa received november 1983, fmal version received august 1984 properties of daily stock this paper examines these data affect event study methodologies. daily data generally present studies. standard tics are ignored. however, their variance cross-sectional for potential are typically well-specified of autocorrelation on an event can sometimes be advantageous. conditional dependence and have higher power can be well-specified dependence. recognition procedures and how returns in daily excess the particular characteristics few difficulties of for event even when special daily data characteris- and changes in returns in addition, tests ignoring tests which account than 1. introduction this paper examines properties of daily stock returns and how the particular of these data affect event study methodologies events. the paper extends of firm-specific (1980)] in which we investigate returns. based on the market model in our previous work, we conclude is both well-specified for assessing the earlier work event study methodologies that a simple and relatively in special cases even simpler to of these conclusions [e.g., brown and warner and rice under a wide variety of conditions, also perform well. however, the applicability and studies using daily …

Source: pdf_first_chars