What the program does is find an "optimal signal", which basically means it looks at historical data for an instrument and identifies where the optimal trades would have taken place, it then uses neural networks to model that signal, creating a strategy which in theory should be profitable in the future.
What I would like to be able to do in NT is load a set of indicators on a chart, manually enter on the historical data where I would have liked to have made the trades, and then have NT output the values of the indicators at the time of the order so that I could look for correlations...
Optimally, I'd prefer to have a multi-core computer identify those correlations for me, but even if I had to do it manually its better than nothing.
Basically if the contents of the data box could be written to a spread sheet format and sorted by trade type, that could be sufficient. That way I could easily look at all the trades I wish I had made and look for patterns in the data.
i.e. the ADX was rising, the MACD difference was above 'x' and the 50 period EMA had crossed above the 100 period EMA in 90% of the long trades...
I could then use that information for developing strategies.
This brief YouTube clip best explains what I'm talking about.
http://www.youtube.com/watch?v=HeVn0riFoy4
Comment