I am doing an event study on the transfer market in football. However during this research, I have encountered two problems I need to correct for: thin trading and event clustering. Football clubs may not trade every day and the transfer window is opened only a couple of months each year. During those few months, a lot of acquisitions and sales are conducted.
For the thin trading problem, I have used Dimson Beta’s with 3 lags and a lead. However, for event clustering I am unsure how to proceed. I have looked at the crude dependent adjustment (Brown & Warner) however have been unable to put this method into practice(Stata). And so, my question is: what method to deal with event clustering would be most suitable to implement in this case (in combination with the Dimson beta)? And should the crude dependent adjustment be most suitable, how do you perform such a test in stata?
(We already have the Abnormal returns, CARs and CAARs)
0 Response to Event study with thin trading and many events(clustering)
Post a Comment