don't you think 85 positions in 5 year what makes about 17 positions a year is a bit less to make a robust analyze abouth your system ? i guess the system is eod ? 5 years is ok for intraday but i would like to see at least 200-300 trades on my backtest; eod I would prefer 10-15 years of data
Well, I figured it's best to have a daily trend strategy accurately depict the trending nature of the market first. Then it serves as input to other strategies that trade in shorter time frames. As often said, it's unwise to trade against the trend. It seems far more profitable, lower risk, and higher odds to make short term trades in the same direction of the prevailing trend. If the market is going sideways, then it's safe to trade in either direction during the choppy periods. Isn't that all accurate? I found a slight improvement to the trend strategies and got the profit factor up to 2.00 with fewer trades and higher profit on average. Most importantly, if you look at the chart, it accurately picks the begin and end of trends with hardly any whipsaws. That's the key to trend strategies. I'm experimenting with different simple strategies to drive the short term trades. I will probably toss out the indicators like RSI, Stochastics, CCI etc and simply use price action. Hey, price never lags. Maybe I'll use PercentR since it never lags. Sincerely, Wayne
Price may not lag, but think of your definition of "choppy period" as just another indicator. Why? It lags as any other useless indicator, you will only see choppy periods after the fact. I also doubt it's "safe" to trade in either direction during the choppy periods, because the choppy period may be just random walk, and you cannot profit from random walk. If you assume mean-reversion, then you simply assume trending tendency to the mean, and this is opposite to random walk. Test price action against random walk and you will see if you're just fooling yourself.