Finding Out What Works, And What Doesn’t Work

About the Author Kevin Davey

Kevin Davey is a professional trader and a top performing systems developer. Kevin is the author of “Building Algorithmic Trading Systems: A Trader's Journey From Data Mining to Monte Carlo Simulation to Live Trading” (Wiley Trading, 2014.) . He generated triple digit annual returns 148 percent, 107 percent, and 112 percent in three consecutive World Cup of Futures Trading Championships® using algorithmic trading systems.His web site,, provides trading mentoring, trading signals, and free trading videos and articles. He writes extensively in industry publications such as Futures Magazine and Active Trader and was featured as a “Market Master” in the book The Universal Principles of Successful Trading by Brent Penfold (Wiley, 2010). Active in social media, Kevin has over 15,000 Twitter followers. An aerospace engineer and MBA by background, he has been an independent trader for over 20 years. Kevin continues to trade full time and develop algorithmic trading strategies.

  • rob says:

    Thanks Kevin for the great article! I am looking for 5 to 10 non correlated futures instruments. Would the ones you used in your testing be a good basket for me to use or would you suggest some others?

    • rob says:

      Looking for futures with the same type of margin requirements as the S&P emini.

  • Thanks for the kind comments, Rob. You can approach the non-correlation issue a few ways. You mention trading different instruments, which is very valid. You could also trade different timeframes, and different strategies altogether. Many times, just having a strategy “different” in 1 or 2 ways from the original strategy will give good diversification.

    No matter how you do it, you can measure diversification via a correlation analysis, or via a combining of equity curves. A smoother equity curve, for example, usually results from uncorrelated strategies.

    OK, but you asked about just uncorrelated markets. How would I do that? I’d look to grab 1 or 2 instruments from each of the major groups:

    stock indices
    interest rates
    agriculturals and meats

    If you do that, you should have a nice diversified basket to trade. Of course, there are times when everything becomes correlated (think 2008 Financial Crisis), but baskets are generally pretty good.


  • rob says:

    Thank you sir I really appreciate your response. By the way great book!

  • Thanks, Rob! I do appreciate it!!

  • Alex says:

    Any conclusions made from this article are plagued by data mining bias. In addition, the chosen time period suggests data snooping bias, meaning that a different time period could produce different results. Choice of instruments constitutes selection bias. Right there you have an article plagued by data mining, data snooping and selection bias. I wonder if the author is aware of these problems.

  • Hi Alex – Thanks for the comment.

    At the end of the article, I do state: “Of course, I made these conclusions based on one study. What if the strategy was different? What if the timeframes or markets were different? What if different years were used for the test period? Will the conclusions reached here still hold?” So, it is entirely possible that the conclusions reached will be different for different instruments and different time periods. I believe I was pretty upfront about that.

    Of course, one key thing to keep in mind is that I present results of maximum net profit quite a bit, and those kinds of results are very, very unlikely to be had in the future. They are optimized, after all. I should have made that clearer in the article. I think the more interesting point is that many times even the best case profit is downright terrible.

    Thanks again for taking time to respond.


  • Alex says:

    “So, it is entirely possible that the conclusions reached will be different for different instruments and different time periods. I believe I was pretty upfront about that.”

    It is certain that they will be different and since you admitted that up front, I wonder, what was the whole point of this exercise? To show what to whom exactly? I don’t get it. I spend time to read the article just to find out that no conclusion is certain from it. If that is the way to treat readers, well, you have accomplished your goal.

    I have read very good articles in this website by other authors from which I gained something. I don’t think this article deserved this place because its assumptions question its conclusions. We as traders have no time for that. This is noise, not a useful article. I am sorry but this is the truth.

    • Mark says:

      I totally disagree, Alex. The article shows one method of analysis you could take to reach the answers you seek. This could be replicated with other markets and other strategies. In case you are here to learn about trading system development, that could be very educational. If you already know all this stuff, Alex, then kick yourself for me because you’re here in the first place.

  • Marco says:

    Thanks for the great article! I think the questions that arise with it are many as you stated above, it can be used as a strong base to develop further the simple concept described.

  • I’m glad you found it useful, Marco!

  • Rich S. says:


    Being a little tough on Kevin aren’t ya? Granted, it was only over four years of data. However, it did demonstrate varying results with different time periods and holding times. It is up to you to explore further.


  • >