It is not at all clear what you are doing. The first portfolio you linked to had 60 stocks, and we were discussing a stock bought 5/5/10. Now the discussion is involving a portfolio of 210 stocks theoretically bought 5/15/10?
It seems to me your method will face headwinds, because for it to be successful (and I don't see enough evidence to say it is) - it seems to me to require a large number of picks -- as you have done (210). The issue is that no one will take that seriously. I mean actually put down money on such a large number of positions.
Apparently you've realized this, and you've now selected 21 on your site for consideration, and as a representative sample or your best guess. If you have 210 and claim great results for 34 and fails for 22, if that were normative, then of the 21 on your site you might get 3 success and 2 failures. Or maybe just as easily 2 successes and 3 failures. So 21 picks is not enough, because the variation imay be too large for the sample size. In some samples of 21, you'd be a winner, and in some samples of 21 you'd wind up a loser.
You apparently get around this by creating a weighting factor -- more money is committed to some stocks than others. That presupposes then, imo, that you can determine which stocks in a list are going to have a better chance of being a positive black swan than the others you've chosen to also be potential black swans. The thing about black swans, as I understand it, is it's very very difficult to predict which are going to be such. And the reader has to rely on a weighting system -- on your site, it would apparently be a reliance on your ability to predict the occurrence of the black swan. |
