Since using Builder to identify market bias is outside the normal use of the program, you’d have to get a bit creative. If you (1) created a reversal type of strategy, which you can do in Builder by un-checking the option “wait for exit before entering new trade” and (2) removed all protective stops from consideration, then you’d have a strategy that was always either long or short every bar. You would want to focus on accuracy as one of the main criteria and not worry so much about profitability. I’d also set trading costs to zero because you wouldn’t be developing a strategy per se but a direction predictor. I’m not sure if you could reasonably expect good results if you used daily bars and insisted on one trade per day – either your long that day or short – but it would be worth a try. You would also probably want to use market entry orders so that you would be entering on the open of the bar. That would give you the closest thing to a prediction of the bar’s direction.
Mike Bryant
Subject: Using Builder to identify daily bias in markets ...
Probably based on %wins, since that would be “prediction accuracy”.
Mike Bryant
Subject: Re: Using Builder to identify daily bias in markets ...
Mike, so how would you recommend we identify which strategy is the "best" predictor? ... the strategy with the highest "Total Net Profit", "Final Equity", "No of Trades" equal to "Number of Trading Days in date range", etc?
This method sounds like it would work nicely with your "perfect profit" system.
--