Optimisation

Viewing 4 posts - 1 through 4 (of 4 total)
  • #73563

    I have developed a few strategies on back test and provide annual profits. I remember reading here that similar indicator settings should be consistent across similar markets. This is not the case with my strategies for a few fx pairs. For example each will have different ema settings. They are not similar. I was told that this may because it’s to optimized. Is this ok, or should they be the same? My thoughts are if each system is profitable in it’s own tailored settings. Then all having the same settings doesn’t really matter?

    #73565

    As long as your EMA periods are potentially adapted to reflect the market behaviors, why not .. You can test their periods and if they are suitable to the specific market movement at that times, by optimizing with Walk Forward.

    I remember reading here that similar indicator settings should be consistent across similar markets.

    Also true, because traders react all the same to the empirical moving average periods, such as 7,20,50,100,200.

    #73566

    So Nicolas are you saying that if  the ema settings are not constant on most fx pairs it is ok? You mention adapting emas to market behavior so do I assume it is ok then to customize?

    #73574

    Yes, adapting periods is a part of optimization, just like any other parameters. Keep in mind that moving averages slopes only reflect how the market is moving though. Depending on your strategy, and if it depends mainly on MA, optimizing their periods sounds ok to me. Of course, don’t be fooled by the overfitting and survivor bias phenomena while optimizing.

Viewing 4 posts - 1 through 4 (of 4 total)

Create your free account now and post your request to benefit from the help of the community
Register or Login