How about using slope of long ma as trend, and slope of fast ma as countertrend for you to enter in direction of long ma trend.
GarrettKimmel A couple clarification questions: 1) your reference to "previous high"... Do you mean prior BAR high, or prior high Pivot? 2) do you have interest in applying to equities? The reason I ask, is it seems this should be fairly easily scripted to backtest thru different equities/indexes over longer time periods, to better understand how/when it is profitable. (I currently have access to equity historic prices, not currencies.) Regards,
Your strategy actually looks pretty good. Simple can be good. It will surely have losing periods though. How many markets are you trading it in, as maybe some diversification will help to avoid massive drawdowns? I'm doing something like this myself and use currency ETFs. My account's not big enough to get into trading lots of futures contracts alongside one another and I didn't want to lose the diversification, so I opted for the ETFs (although I'm now looking at some of the e-Micro futures - anyone got any experience with these contracts?).
What many don't talk about is individual instruments, what might work well in Crude Oil usually losses in ES, you have to know your market. I have never liked systems where you have to risk so much and this is the method where I believe that to happen cause when you get in chop, too tight of a stop, you will numerous losses in a row. I think you have half of a good system, keep 25sma but use it as main trend ID, using the slope. Entry would always be on the other side of the 3sma and must be close or touching the 25sma, in downtrend look for highest low to sell 2-3 ticks below and stops above that area highs. Use a Time stop and/or distance to go to breakeven plus one tick stops. You will get in at better prices and if in chop-randomness will allow you to often get plus one tick profit. Yea, I tested sma delayed back in mid 90s, I even paid $8k for one minute system that had delayed EMAs and based on fibs to boot, it actually worked well on big S&P500 cause risk was tight, but didn't work well on ES, I try to check it out every year but still doesn't work, but I did well with it for 3 years.
Whether or not it works doesn't depend on how "simple" it is. Like anything based on moving averages, it will appear to work well in trending markets, and chew up your account in ranging markets. Whether it "works" depends on whether the market's trending or ranging, and if you know that with confidence, you don't need a moving-average-based system anyway.