Building, Improvement and Optimization

I have understood is best to use the complete data-set of 30 years to build strategies, because of the statistical relevance of at least 2000 trades. But have i use the 30 years data for improvement and optimization also? WalkForwardMatrix and WalkForwardOptimization working many hours for one strategy. Are the strategies really better then? I am very unsure with this…

In M1 timeframe the 2000 trades are there in 2-3 years, will they be better with 30 years data, too?

Every year the market works in slightly different regime (also depending on what markets we are talking about). That’s why it is recommended to use as much data as possible for building and optimization. For me personally when you go down to M1, M3 it’s complete noise – very difficult to find something which would survive for a longer period especially while using the classic technical analysis approach (indicators, price action)

i can only find one answer to this. If you on M1 timeframe find that the strategy in WFO is working on 2 years of data which it might even if not perfect you should then extend the period to find out when it failed and why it did . I usually find that the settings change to much between WFO periods to ever be able to keep it profitable in realtime . My latest one worked for 4 months but got wiped out in 5 days.

Thank you for your fast answers!