Last Updated on 17 November, 2020 by Samuelsson
Finding trading strategies, like the crude oil strategy below, indeed takes a lot of time. You have to consistently turn up new concepts and logics, only to see nearly all of them fall apart, either in the initial backtest, or during your robustness testing!
Yesterday I was testing some very simple concepts in crude oil futures, when I stumbled upon an idea that I had tested around 6 months earlier. I had left it right where it was, probably since something else happened to catch my attention at that time.
Below is the equity curve for this very edge in the Crude Oil market. Important to note is that slippage and transactional costs are included here!
This trading strategy in crude oil goes both long and short, and uses only 1 condition for the entry, which is symmetrical for long and short entries.
You might wonder how it’s possible to let a good edge like this one fall into oblivion?
Well, when testing a strategy and you finally find something, that has a tendency to branch uncontrollably. In this case, I probably went on some other market or timeframe, and simply forgot about this very promising edge.
Fortunately, when searching one of my old chat logs with traders I cooperate with, I found that I had shared my findings in the crude oil market, but never taken a closer look at it. Given that the out of sample looked alright too, this indeed seemed promising!
Creating a Crude Oil Strategy
Since the logic was so easy, there definitely was room for at least one more filter, without putting me too much at risk of curve fitting.
I started by looking at some volatility filters that I have found work well on many other strategies. The results were quite alright, but I soon realized that while the entry conditions are symmetrical, the strategy might be better off with different filters for the long and short sides.
Subsequently, in order to make the crude oil strategy better, I decided to apply another filter. This time it was based on a time based tendency in the crude oil market that I already knew of. As you can see in the results below, applying this condition, together with the volatility filter, drastically improved the results!
From looking at this chart, an experienced trader might realize that this is a trading strategy that uses quite a tight stop loss, which is true. This Crude Oil trading strategy uses a tight stop and lets the winners run.
And for those who are familiar with common backtesting pitfalls, I gladly confess that the Look-Inside-Bar backtesting feature is enabled!
Am I Confident in This Crude Oil Strategy?
Well, I certainly believe that this will prove to be a great trading strategy. The logic is incredibly simple, and it uses very few filters.
However, I cannot know for sure until I’m done with my robustness testing procedures. More than once I’ve seen promising trading strategies fall apart, against my perception of it being valid.
In this case, however, I already have some true out of sample, which is a strong indicator of the validity of the trading strategy!
Want to Know More About Building Trading Strategies Like This One?
And if you’re interested in getting trading edges, our unique trading membership might be exactly what you’re looking for!