Forex Mechanical System Backtest Results: Triple SMA Crossover

Greetings, earthlings! I’ve just finished crunching the numbers on the Triple SMA Crossover system and I gotta say, this forex mechanical system could use a bit of improvement. But first, here are the yearly backtest stats:

forex triple sma crossoverAt first glance, the numbers don’t look too bad, as the system had close to a 50% win rate and an average win greater than the average loss. However, a review of the trades taken reveals that the system ended up positive just because of one big win:

forex triple sma crossover tradesPutting this in context, the last winning trade was scored during the months when markets had a very strong pro-dollar and anti-euro bias, right around the time when the ECB cut interest rates for the second instance and the Fed expressed an inclination to start tightening next year. Well, that was one lucky break!

The rest of the trades indicate that there’s a lot of room for improvement though, as the SMA crossover signal seems to be taking place late in the trend. Of course I’m still running the forward tests for this month but I’m not very hopeful that this can be able to score another huge win, given how we’re seeing a range-bound environment in the forex market these days. Got any tips on what kind of adjustments need to be made?

  • Filip

    Tips? Yeah drop the system because mechanical systems don’t work long run.
    Your moving average system only worked because as you said the past few months where a strongly trending environment and moving average systems are only profitable in trending markets. All moving average systems are profitable in strongly trending markets!!

    Likewise oscillator based systems like Stochastic , MACD and RSI work terribly in trending markets but are much better/proffitable in range bound markets.

    • I tried adding the ADX to filter out the signals that are occurring in a ranging environment.

  • Rick S

    You should try running this system over a 10+ year time frame and across as many markets as possible. You may find that results might be much different year to year than the results you have above. I trade an EoD breakout system and when I was building it I found that it worked on some markets better than others and even though the CAGR was very good (over the 20+ years tested) some years were good and some were not as good.

    • Yikes, not sure if I could backtrack enough data for 10 years but I’ll look into it. Thanks for your suggestions!

  • Just came up with updated SMA parameters in the new version of the system. Check it out!

  • Thanks for sharing your thoughts!