After programming a promising FX strategy (tested over 70k+ trades on multiple instruments and time frames ) I wanted to forward test it on my real money account.
But the backtest result in the strategy analyzer, over a long or short period, is substantially different than the historical strategy performance over the same period.
Please have a look at the attached results (Overview and trades). As you can see the entry end exit prices are different (trade 7/8 up to 20 Pips?!) See trade 7/8 for example
It is the same strategy over the same price data (IB), and the exact same strategy setup.
The strategy works with limit orders to enter a trade and take profits/stop loss to exit.
I also deleted and reloaded all historical data.
After many tests, the historical strategy performance shows always worse results than the strategy analyser, why is that?
So, did someone of you experienced such an behaviour?
Which one is right, or close to reality? Are NT8 numbers statistically reliable ?