Serving the Quantitative Finance Community
Pooling Stocks in high frequency.
I have two questions:1) I have around 140 stocks and not all stocks are poolable meaning the distribution is quite different. I am using the standard F test to test for poolability. Just wanted to know if there is any better way to pool stocks together for analysis rather than a hit and trial method of checking the F test. 2) My second question is that I could run stock by stock regression but then report the distribution of the betas and Rsquared. But this is not granular enough. Any other thoughts?