Man,
I am going to try and do some programming tonight to test the peers trading. Your summation looks generally correct to me, but I don't understand a couple things you might be able to clear up.
Quote from man:
basket of n stocks within the same sector on the lowest level of SP Sector Gics (or lowest level on Bloomberg, which is even more precise).
Lowest level? SP Sector Gics?
Just to be sure, in my own words:
1. Stocks with close above $10, avg 5-day volume greater than 250,000
2&3. Okay
4. set all stocks in the two baskets at 100 at t-x.
5. calculate the current normalised average price in basket a
6. calculate the current normalised average price in basket b
-Let's say A1=14.50, A2=25.6; B1=50.2, B2=12.4
What would steps 4-6 do? I realize in general that you are trying to standardize the groups for comparison.
8. if this difference exceeds p %...
-Any suggestions for the level p%? or would one optimize to attain the greatest value?