There are many metrics that can be used to evaluate a system’s performance. Of course, the best fitness function is the one that produces the best OOS performance. The most important metric might or might not be the best fitness function. On considering the most important metric: we have many reasonable candidates: total return, return/max dd or some variation (Calmar), win percentage, profit factor, and max intra-trade/intraday drawdown. Beyond the simple metrics there are metrics that try to measure the smoothness of the equity curve or how close it is to a straight line.
All of these metrics convey some important information. Total return is obviously important because if the total return is too low then a system might not be worth trading even if it works. The profit factor is a good indication of whether or not a system can overcome trading costs (if those weren’t baked in). The maximum intraday drawdown tells us how painful a system might be too trade and also suggests whether or not it can be leveraged. The percentage of winning trades suggests that a system will be more stable and easier to trade. Some metrics focus more on the end state or where you, ultimately, end up while others weight more heavily the path that it took to get there.
But, what’s the most important metric? Perhaps a metric that is less frequently discussed. The most important metric is what I will call the “minimization of time between new highs”. Think about it, if a system is always making new highs then we won’t really care if it has a low or high win ratio– even though it suggests a high win ratio. A system that is always making new highs is an easy system to trade regardless of any other factors. If a system is always making new highs then even if you take a large drawdown then there’s a better chance you have already built some profits up and you know you should recover quickly. A system that makes a new high every day would be the best system to trade. No matter how you slice or dice the data, if a system is always making new highs then you could have started trading it at any time during its history and would have expected to achieve a similar result. And, knowing when to stop trading would be easy, you don’t wait for a drawdown. You just stop when you fail to make a new high within the allotted time. We might expand this idea to be “minimize distance between new equity highs”. Distance is the product of time and drawdown.
I will add to this metric, one other metric, which is the total return/max dd. This ratio is really nice because it adjusts the returns for the leverage or risk it took to acquire them. This ratio, also, provides a strong suggestion as to whether your results are due to luck or skill. A very high ratio would be very unlikely the result of fortunate chance.
So, there you have it: two metrics that are the most important; percentage time between new equity highs and return/max dd. Of course, the other metrics such as average profit per trade are still important for determining if a system is tradeable. But, those two metrics really capture what a great system should be. Let me know if you agree or disagree and why.
Curtis is passionate about markets. He has developed top ranked futures strategies. His core focus is (1) applying machine learning and developing systematic strategies, and (2) solving the toughest problems of discretionary trading by applying quantitative tools, machine learning, and performance discipline. You can contact him at firstname.lastname@example.org.
Please log in again. The login page will open in a new tab. After logging in you can close it and return to this page.