Hi,I am trying to test the impact of an event on stock returns. I wrote:To test for the abnormal return, we could use the event study methodology: AR=r-E(r); where AR , r, and E(r) are the abnormal, actual, and expected normal returns respectively for time period t. The expected normal return could be estimated as the average return on the stock i during the estimation period. By averaging these residuals across firms in common event time, we obtain the average abnormal return and we test for significance of abnormal returns using the t test : (average abnormal return/standard deviation of abnormal return) .I received the following comment from a professor:<<THIS COULD BE ONE TEST BUT GENERALLY ONE WANTS TO CONTROL FOR OTHER SYSTEMATIC IMPACTS ON STOCK RETURNS.>>Does anyone know what is he referring to ?

I'm guessing dividends, stock splits, and spin-offs would be examples of systematic impacts that the prof is referring to.

Sytematic impacts refer to standard factors that drive stock returns. What you are trying to do is calculate the impact of the event holding all other factors that drive the returns of the stock constant. Just to give you an example, if the whole market goes up (for whatever reason) during your event, then so will your stock, and not because something special happened during the event period. So at a minimum you should control for market moves. Usually a one factor market model is used, some also use a FF three factor model when calculating abnormal return.

In general, a properly constructed event study uses some sort of regression during the estimation period to control for movements that are caused by other factors aside from the event. A popular model is the "market model" (which was mentioned by gardener3). An alternative is using a two-factor model, that has a market proxy and an industry proxy. You can also use other determinants of stock returns, depending on the type of firm that is involved and the event being studied.