Hello all,
I have tested some of the strategies that one can find in books and currently not figuring out anything that works.
Eg MACD, Bollinger, Stochastic, ARIMA, OLS, GLS either as mean-reverting or trending, long-short.
Everything looks great during back-testing (as usual).
Still I am not able to generate a strategy with Sortino ratio >= 6 when cross-validating (fees included).
System does 1 future contract on 1 min candles, trade on complete candle only.
Optimization period 9600 bars, cross-validation 9600 bars K-Fold 5 splits, all without Stop Loss which gets added at the end as a final step.
I have tried different time periods to rule out any market regime effects and also different markets (FGBL vs FDAX which is more "whippy") to rule out market dynamics.
Any ideas what I am doing wrong? Or is that THAT hard to find a working strategy?
Best Regards,
Mov