What turns out to be more fun for over 3 beta stocks is to pick them off about an hour and 1/2 before they BO. You can make a short list the night before and then track them in the AM for their BO. i used to post the lists and the timing on the BO's. It did two things: when I traded the SEC cited me for "insider stuff" and the posts got me phone calls from corporate execs who want to know where in their corp I got the info. What you are doing is neat for collecting records etc,; you might want to just move ahead in the cycle a little bit to pick up the period before they go into their cycle where money can be made. What are doing can get frustrating because you know they do it and you have no timing to take the trades properly.
I am missing something, but let me regurgitate and see: 1ReturnHistory = LinearReg of Stock Returns 2ReturnHistory = LinearReg of Market Returns Numerator = Covariance (Correlation??) of (1ReturnHistory,2ReturnHistory) Denominator = 2ReturnHistory Beta = Numerator / Denominator I don't see where "multiplicative coefficient" comes into play... Time to find the old statistical methods textbook. Thanks Bibine, RLB
What might I do to move ahead the cycle? You are correct that I have no timing method. This post was focused only on the selection of a trading universe, one which would be comprised of very fast moving stocks. That being said, I'd be curious to know how one would incite the SEC to such a response... and I'd love to talk with some corporate execs... RLB
Here are the results of testing on Nasdaq stocks over the same period. I bumped up the required closing price to $5, but got significantly more stocks passing through the filter... 114 to be exact. The average absolute percentage change for this group was 14.4. The comparison groups registered average changes of 8.5%, 10.5%, 8.2%, & 8.4%. The difference is not as dramatic but still significant.