after generating strategies with about 6/7 years of eurusd 30min data I did the various tests oos,slippage ,montecarlo,other market other time frames.. then I wanted to do more in-depth tests without going to uncomfortable analysis of particular data of pure statistics... then I did Monte Carlo resapling with 80% confidence and 60% main data
after or 100% filtered > 1RET/dd... after I took the chart eur usd atr (9) W1
I chose 3 periods ... 1with strong atr up 1 with strong atr down 1 with atr flat and' I filtered everything with PF>1.1 and RT/DD>1
the surviving strategies I filtered them with last OOS from 01012019 with RT>1
atr period of choice 3 OOS from 2003 to today.... I repeat I have chosen extreme periods... for the worst-case scenario….
little small example of a portfolio m30 5 strategies derived from the first 16
is an little example…. of portfolio….only 1 symbol 1 time frame…..I tried to simulate other portfolio with h1 and gbpusd... looks interesting
Nice results...there are plenty of interesting strategies in our packages..but I think the goal is to diversify over multiple symbols...that's why we need people to wake up and start generating strategies...as you see post filtering/selection and portfolio building is a 'walk in a park' if you have good strategies you can choose from