I have been trying to move my intraday Strategy from NinjaTrader to TradeStation. The trade executions times are in sync about 95% of the time, which is acceptable. However, the back testing on both systems is a night and day difference. In backtesting on TradeStation, 70% of the time, the fills are unfavorable and off by 2-3 points NOT ticks. On TS, the Net Profit is always negative and the Equity Line is downward sloping, but on NT it is profitable with a positive equity line.

I am hoping this observation can ring a bell about a solution:

When running forward testing with real time data on both platforms at the same time, the PnL either matches to the tick in MOST cases, but sometimes is off by 1-2 ticks. This is acceptable and realistic.* *That tells me that both TS and NT are getting close fills during real time.*But the question is - if real time has close fills, why are the historical back test fills so drastically different?* *

I am hoping this is a configuration issue or something small that I am missing.

Thanks in advance.

szayed,

are you using intrabar order generation for your strategy? Including screenshots of the "General" and "Backtesting" tab of the "Strategy Properties" with your reply might help in getting some ideas.

Regards,

ABCTG

szayed 

I have been trying to move my intraday Strategy from NinjaTrader to TradeStation. The trade executions times are in sync about 95% of the time, which is acceptable. However, the back testing on both systems is a night and day difference. In backtesting on TradeStation, 70% of the time, the fills are unfavorable and off by 2-3 points NOT ticks. On TS, the Net Profit is always negative and the Equity Line is downward sloping, but on NT it is profitable with a positive equity line. I am hoping this observation can ring a bell about a solution: When running forward testing with real time data on both platforms at the same time, the PnL either matches to the tick in MOST cases, but sometimes is off by 1-2 ticks. This is acceptable and realistic.* *That tells me that both TS and NT are getting close fills during real time.*But the question is - if real time has close fills, why are the historical back test fills so drastically different?* * I am hoping this is a configuration issue or something small that I am missing. Thanks in advance.

This problem has been resolved thanks to a contributer at the TradeStation forum. In the Backtesting tab, there should be no additional slippage. That was set to to $0 and the checkbox below should be disabled. Slippage and Commissions should only be in the General tab.