This is the task:
"For each industry, using a reasonable level of significance, test the null hypothesis that the parameters a and ß are identical for all companies in the industry and are equal over both the January 1978 - December 1982 and January 1983 - December 1987 time intervals. (Be particularly careful in calculating degrees of freedom.)"
Before I had to test that Alfa and Beta are equal for one company over the two periods (halves), and then test that Alfa and Beta are identical for all the companies in the same industry over the whole period, I did the first part easily using the Chow test, and the second part after restructuring the data into panel - stacked time series I think I solved too. But I'm not sure what to do with the third part (quoted above)
Alfa and Beta come from simple OLS - Company return = alfa + beta*market returns. It should be relatively simple since we haven't gotten too deep in this course, but I guess I'm missing something.
Thanks for any help!