FJeffreyMartin
 Member
Posts: 13
Joined: 6/19/2013
Location: Los Alamos, NM
User Profile |
I've tested the effects of "Trade Selection Order". I found nothing that consistently produces better results.
This setting should have no effect for most of the trading. It only comes into play when hitting the margin limit. In the case that the margin is limiting I see no basis for why any choice would be superior. Using the Alphabet is as good as any, even though its obvious that it has no logical tie to expected performance. The same could be said of any other setting. The bigger question is what percent of the recommended trades should we strive to use and, bigger yet, is there a way to rank symbols by prior performance that would add a positive bias?
There's been a lot of sweat and tears expended on this topic including my own. But I no longer see testing as a productive effort unless we develop more informative data to assist symbol selection and/or % of recommended transactions.
For now. I'm moving on.
Adjusting the "% trades taken" by scaling size of the trades to fit the cash available seems to be a trade off that could be optimized if testing would yield a reliable impirical result.
As you go higher the more transactions per trade will be required which has its penalties.
|