just to make sure i get it: you ran backtest on TEN days? and you chose the best market out of FOUR indices?
man You can make a simple backtest if you have a minimum of 30 incidents. I think for an intraday view 10 days are sufficient. (I am also looking at 1y and 4y timeframes in the indexes). What are others thinking?
well, thirty observations are way too few in my understanding. a few hundred is the minimum IMHO. plus they should happen in different environments. high vola, low vola, bullish bearish market, you name it.
man I agree, but taking the recent 30 incidents and assuming that the market conditions don't change, this could be a quick way to go. The other thing is experience which signal to take and which signal to skip.