Last Updated on March 6, 2021 by Oddmund Groette
How do you evaluate a trading strategy? In hindsight, it’s easy to judge a strategy by the result – the CAGR or the annual return. However, this is what Annie Duke calls resulting in her brilliant book Thinking In Bets. A good decision can lead to a bad outcome, and a bad decision can lead to a good result. When looking at just one decision, the quality of the decision is, of course, not necessarily linked to the outcome. However, the correlation between these two is high in the long run.
Likewise, a good strategy can’t be judged solely on the return. This is why we like to quantify strategies to evaluate their performance, and the one we use the most is the profit factor. This article explains what the profit factor is and how you can use it.
How do you evaluate a trading strategy?
The more smoothly the equity chart rises from the left to the right, the better the strategy. By looking visually at the equity chart, you get a pretty good estimation if the strategy is worthwhile on its own or in need of some additional variable(s). If the equity line goes like a straight line, obviously, the strategy must be pretty good. In these circumstances, you don’t need much of a mathematical confirmation if the strategy is good (or not). However, some traders focus on the maximum drawdown, while others focus only on the outcome (the profits).
But in most cases, it’s not as obvious as it looks: You need some quantitative test on the significance of the strategy. That’s why all trading software has in-built metrics to evaluate strategies. Below is an excerpt from Amibroker, which shows the standard metrics on every strategy-test:
How to calculate the profit factor:
The profit factor is easy to calculate:
The ratio between gross profits and gross losses is the profit factor. If you have a strategy that has accumulated 500 in profits and 250 in losses, the profit factor is two. In short, the total profits and losses during the test period are summarized. If a strategy has 156 trades with losses and 199 with profits, the 199 profits are summarized and divided by the 156 losses.
What is a good profit factor?
As a very crude rule of thumb, we prefer to see the profit factor higher than 1.75, but we are not necessarily happy to see values above 4, either. We want to see higher values than 1.75 because a strategy often yields real-time results worse than the test. It’s rare for a strategy to perform better in real life than on your screen. There are many reasons for that, something we will not discuss further in this article, and you want to have a “margin of safety” by selecting strategies with a reasonably high profit factor. If the profit ratio is 1.25, you have a minimal safety margin, which is obviously not ideal. Even small negative changes might make the strategy losing money. We consider 1.75 a low threshold, but we prefer to trade strategies with a number above 2.
You might think that a high reading is good, but that’s frequently not the case. How could that be? First of all, it might be a sign you have curve-fitted the strategy. Too many variables, too few signals, and too short a test period might lead to exceptional results that are unlikely to work on future and unknown data. In such cases, your real trading might disappoint you. Thus, any numbers above 4 should make you wary.
You have to find out yourself what kind of thresholds you want to use.
Which strategies have high profit factors?
Mean-reverting strategies tend to have higher profit factors than, for example, trend-following strategies.
The profit factor is all about risk
Let’s assume you have a strategy with a profit factor of 3, which is pretty high. It has generated 300 trades over the last decade, and it involves only two variables. Thus, the chances of curve fitting are reduced but not eliminated (it never will be). Then you start fiddling with the two variables by changing the values. When you limit the threshold to generate more trades, you notice the profit factor, and the max drawdown worsens.
However, the strategy makes more money overall because of the increased number of trades. This is always the trade-off between assumed risk and payoff. No pain, no gain. This trade-off is something you’ll face daily. You want as much profit as possible, but on the other hand, you want to get the profit with the least amount of stress and headache.
One solution is to embrace automatic and mechanical trading. This allows you to trade many strategies that can smooth your returns. It would be best if you had a portfolio of quantified strategies. By combing many strategies that on their own perhaps have only a profit factor of 1.75 (and might not be so attractive), the sum of the strategies might give a higher profit factor because of its diversity. You need to trade different markets, different time frames, and different types of strategies. The only limit is your own capacity and imagination. The software is unlikely to be the restraint.
The profit factor is a mathematical metric that divides the gross profits by the gross losses. We like to use values between 1.75 and 4 and become skeptical about values under and over this range. A low number indicates a less robust strategy, while a high reading might be too good to be true in real life. We aim for the averages in between and reckon our diversity makes for a smooth total return.
Disclosure: We are not financial advisors. Please do your own due diligence and investment research or consult a financial professional. All articles are our opinions – they are not suggestions to buy or sell any securities.