GARCH-based approach for measuring abnormal stock returns

[James82]

Dear Tom,

I am having problem with creating the right code. I tried to reason out the logic with various introduced examples, but failed to solve how I am supposed to do my research exactly. My goal is to measure the effects of events in companies stock returns.

I am trying to use GARCH-based approach, which is introduced in the following working paper by Savickas (2003) in pages 167-168.
URL: ftp://economia.unica.it/mattana/Modelli ... tudies.pdf

Any help is appreciated.

Best regards,
James

[TomDoan]

Do you have the dummy variable? That's a separate univariate GARCH model for each i with a mean model that has CONSTANT, the market return and the event dummy, and an XREG in the GARCH of the event dummy as well.