I noted that if I run the exact same strategy and parameters in the strategy analyzer and live that I see a difference in the trades. I traced the problem down to the fact that the strategy analyzer correctly accounts for a split that occurred, while the live strategy (historical performance) does not. I have deleted the historical data both in the historical data manager and file system and still this persists.
The live strategy does see the split, but for some reason it seems to see it late and therefore makes a bunch of trades based on what it perceives as a temporary significant drop in stock price (reverse split). I should mention that all of these strategies have been running for months and have occassionaly disabled/enabled for backups, maint, etc...
I did start a new live strategy instance of this exact same strategy/parameters and it did not suffer from the same problem with the split when I look at its historical performance.
So I am wondering if NT stores separate historical data for different instances of live strategies. This doesn't make sense to me, but I can't think of why I would be seeing this behavior.
Any help would be appreciated.
Comment