Testing Market Regime Indicators

About the Author Jeff Swanson

Jeff is the founder of System Trader Success - a website and mission to empowering the retail trader with the proper knowledge and tools to become a profitable trader the world of quantitative/automated trading.

follow me on:
  • Mike S. says:

    Thanks for the great article. Can you go into more detail about how you use the RS Ranking Method as a trend indicator? I think I understand how it works as a means of comparing different markets, but not in terms of determining whether a market is in a bull trend or bear trend.

    Thanks again!

  • Yes, I’ll work on something to help clarify. Glad you found the article helpful.

  • […] classify Bull and Bear Regimes I used the RSRank Indicator described by Jeff Swanson in this post at System Trader […]

  • […] classify Bull and Bear Regimes I used the RSRank Indicator described by Jeff Swanson in this post at System Trader […]

  • Mark says:

    Why are the Sharpe ratios so low?

  • Francisco says:

    Jeff, I cannot load the ELD onto Tradestation. ¿Is there anything wrong? ¿Maybe is an old version?

    • Looks like there was a problem with the download file. I just uploaded it again. Please try the new file and let me know if you still have problems.

  • AnTZ_TK says:

    Hi Jeff,

    have you ever considered the importance of sample size? You draw conclusion based on samples as small as 17 (for ROC).

    Imagine what a difference just one more sample may have on your results! Now imagine it’s an outlier (i.e., 2 standard deviations away from the mean)!!!

    • Of course sample size is important, but the topic for this article was not sample size. It was ideas on creating and testing a simple regime filter to divide the market into a bullish and bearish mode. The ROC has a sample size of 25. The “since-2000” graph is a zoomed in view of the last 12.5 years. In short, I’m explaining a concept in this article and I encourage everyone to do their own testing. Thanks for the comment.

      • AnTZ_TK says:

        Hi Jeff,

        I understand and appreciate what you were trying to demonstrate.

        What I wanted to point out to you was that it might be better to do it in a “scientifically sound” way. Otherwise beginners may believe it’s OK to work with such small samples and knowledgeable people may think you don’t know better and ignore you.

        Cheers! 🙂

  • […] Testing Market Regime Indicators […]

  • vimal says:

    Hi Jeff. Good article. i am loading the indicators and eld but am getting a message that the ROC , je and rs smoothed indicators have been deleted. why would this be? also have you tested the ultimate oscillator? it seems to do a reasonable job overall in terms of determining the trend.

  • […] your trading systems to the changing market conditions. In a previous article entitled, “Trend Testing Indicators“, I tested several indicators that could be used to divide the market into two modes: bullish […]

  • Peter B says:

    Thanks Jeff, for another interesting article. Would you please explain why you use a non-uniform weighting in your smoothing. I assumed you would have used a simple average. Were you intending to build in some lag to weight the previous values more – to act as hysteresis?

    • You’re welcome Peter. The formula used is from John Ehlers and is a finite impulse response (FIR) low-pass filter. That’s a lot of words to smooth data. Lots of his material is from signal processing. You can read a bit more here.

  • […] in the conventional sense as using something such as the SMA200, ROC200 > 0, or RS Rank (see this post from […]

  • Jesse says:

    Hi Jeff,

    Do you know where I can find a simple strategy based on momentum (Price ROC). similar to the ROC calculated above that says go long/short when the momentum is greater than X over X amount of calculated periods?

    • Hey Jesse. Sorry, I’m not familiar with an available strategy that fits your description.

  • >