Last Updated on 17 November, 2020 by Samuelsson
It is common for new traders to spend all their time looking for the perfect trading strategy. They assume that once they find it, they are set for the rest of their trading career.
However, reality is different. Trading strategies stop working and need to be replaced.
So why do trading strategies stop working?
Trading strategies stop working because the market changes, leading to the trading strategy falling out of sync with the market. However, quite often trading strategies stop working because they were curve fit!
Let’s go a little deeper into why trading strategies work in the first place, and how you should adjust your approach so that the failure of a trading strategy has the smallest impact possible!
Why Do Trading Strategies Work?
Trading strategies work because they manage to identify recurrent market behavior. Recurrent patterns could be almost anything, but are most times based on technical analysis. For example, a really simple trading strategy could be to buy once the market has performed three consecutive lower closes, and sell one day later. A trading strategy does not need to be much more complicated than that!
For example, look at this strategy in the Sp&500 futures market. This strategy consists of only 2 conditions, and despite its simplicity, it has managed to perform incredibly well for a long time!
In trading, we often use the word “edge” to describe a pattern that we can use to our advantage.
Trading strategies are edges, but depending on the definition, you could say that a trading strategy is more polished. A robust trading strategy has been vetted through rigorous robustness testing, and has a good chance of continuing to make profits also after it has been developed.
As we went into before, trading strategies define market behavior that we make use of in our trading. However, when we analyze market data, we do not only look at market behavior and tendencies but also inevitably random noise!
Therefore it is crucial that we do not build our trading strategies around these random market tendencies. If we do, we can expect to get a trading strategy that is no better than picking entries and exits randomly!
How Do You Know If a Strategy Has Stopped Working?
So, let us say that you have created a strategy that has passed your robustness testing criteria, and that you have started trading it! What would you do if the strategy dipped right after you went live with it?
Well, many traders would panic and turn it off immediately.
However, those with some more experience know that drawdowns are a natural part of trading. They occur every now and then, sometimes so small that they go unnoticed, and other times of a size that might make you question your strategy.
So how do you know when the strategy only is in drawdown, and when it has stopped working completely?
Unfortunately, there is no way you can tell with 100% certainty. Once a strategy goes into drawdown, that could very well be the beginning of the end or merely another dip in the curve. You cannot know!
Therefore, you need to set a stop loss on the strategy level. When the strategy hits the stop loss, you just stop trading it! The strategy might start working again, and in that case, you preferably should wait to confirm that the upswing is not just temporary!
Making Sure That a Strategy Does Not Stop Working
Since the robustness of trading strategies is of major importance for trading performance, there exist many ways of ensuring that a strategy does not stop working. Still, no method is idiotproof, and many of the methods can be fooled if not employed correctly!
Here are some of the more common methods to gauge the robustness of a strategy:
- In Sample/Out of sample testing
- Walk Forward Analysis
- Forward Testing
In sample and out of sample testing is the most widely used method to ensure the robustness of a trading strategy. It is done by dividing the market data into a validation set and a training set. All the testing is then carried out on the training set, to then be validated on the validation set.
The premise of this method is that true market behavior will persist throughout both data sets, while random market noise will not!
Walk forward analysis is a technique that builds on in-sample and out of sample testing. In short, you could say that Walk Forward Analysis is a technique where you perform several in and out of sample tests.
For example, if you had data for the period of 2010-2018, you might begin with optimizing the strategy on data between 2010-2012, and then apply those settings on the coming out of sample period, which could be the following year of 2013. Then you take the period of 2011-2013, find the optimal values, and once again apply those to the following year of 2014.
Foward testing simply is a method where you wait and see how the strategy performs going forward!
Ways of Protecting Yourself from Strategies that Stop Working
To minimize the impact of strategies that stop working, you need to have two things in mind:
- Position sizing
Diversification is very important for many reasons. A diversified portfolio will earn you more money at lower overall risk, and create a smoother equity curve. Last but not least, it will effectively reduce the impact of failing trading strategies.
The more markets, timeframes, and diverse logics you trade, the less impacted you will be by the failure of a single strategy. Even if one market experiences disruptions, there is bound to be at least one market where your strategies work well. Here at the Robust Trader we sometimes trade more than 100 trading strategies at the same time, all possible through algorithmic trading and automated trade execution.
By the way, if you are interested in learning more about algorithmic trading, why not have a look at our massive article on the topic!
Position sizing is also very important. When designing your portfolio of strategies, you need to make sure that one strategy does not risk too much of the account! You need to consider the maximum loss in case of the strategy failing, and make sure that your account can withstand such an event.
As in all types of trading, striking a balance between risk and return is one of the trickiest decisions you stand before.
Strategies stop working either because the market changes, or because it was curve-fit when developed. Still, no trading strategy lasts forever, and therefore it is important to take preemptive actions to ensure that you can withstand the failure of one or even several trading strategies, without getting wiped out.