HAC standard errors and the event study methodology: a cautionary note

Abstract

In support of Fomby and Murfin's (2005) article published in this journal, we demonstrate empirically, rather than theoretically, the severe consequences of using Heteroscedasticity and Autocorrelation Consistent (HAC) SEs in regression-based financial event studies. Applying an event study to a recent merger, we show that the use of HAC SEs render misleading conclusions. Critical values for t-tests on the event dummy variables are about 15 times larger than the nominal values using only a year of daily return data. Even with samples of only 100 returns, critical values exceed nominal critical values by a factor of 10.

Credit Rating Changes and Stock Market Reaction in the Kingdom of Bahrain

“…According to Ford, Jackson and Skinner (2010) and Freitas and Minardi (2013) the choice of the window should neither be too long nor too small such that it does not encompass other events and fails to capture abnormality in prices. Also, the literature does not seem to have a consensus in defining the event window.…”