How to Combine Multiple Strategies in Forex Trading to Achieve Smoother Equity

Forex trading is the exchange of one currency for another, with the aim of making a profit from the fluctuations in exchange rates. Forex traders use various strategies to analyze the market and execute trades, such as trend following, mean reversion, arbitrage, breakout, and so on. However, no single strategy can guarantee consistent profits or avoid losses in all market conditions. Therefore, it may be beneficial to combine multiple strategies into one portfolio, which can diversify the risks and smooth the equity curve.

But how to combine multiple strategies effectively? There is no single answer to this question, as it depends on the type, nature, and performance of the strategies. However, there are some general principles and methods that can be applied. In this article, we will focus on combining different yet complementary strategies, such as trend following and mean reversion strategies in the forex market.

One of the most important indicators to consider when combining multiple strategies is the net profit-to-maximum drawdown ratio, which measures the risk-adjusted return of a strategy. The higher this ratio is, the more profitable and resilient the strategy is. Another important indicator is the maximum time span for net value not creating a historical high, which measures the consistency of a strategy. The shorter this span is, the faster the strategy recovers from losses and reaches new highs.

One of the simplest methods to combine multiple strategies is to assign equal weights to each strategy, regardless of their performance or characteristics. This method is easy to implement and avoids overfitting or data snooping. However, it may not be optimal, as it ignores the differences among the strategies and their interactions.

A more sophisticated method is to use some optimization techniques, such as genetic algorithms or machine learning, to find the optimal weights for each strategy that maximize the portfolio performance or minimize the portfolio risk. This method takes into account the performance and correlation of the strategies, and tries to find the best combination of them. However, it also has some drawbacks, such as complexity, computational cost, and historical dependence.

Another method is to use some rules or criteria to evaluate the performance and quality of each strategy, and adjust the weights accordingly. For example, one can use a ranking system based on some indicators, such as net profit-to-maximum drawdown ratio and maximum time span for net value not creating a historical high. One can then allocate more weight to the strategies with higher rank, and less weight to those with lower rank.

To illustrate these methods, we will use two different forex strategies as an example1. These strategies are:

Breakout Strategy: This strategy identifies and trades price breakouts from support or resistance levels in various time frames.
Mean Reversion Strategy: This strategy identifies and trades price reversals from overbought or oversold levels in various time frames.
These strategies both utilize price action and technical analysis but in different ways. They have different performance and correlation characteristics2, which make them suitable for combining.

The table below shows the annualized return (AR), annualized volatility (AV), Sharpe ratio (SR), maximum drawdown (MDD), net profit-to-maximum drawdown ratio (NPMDD), maximum time span for net value not creating a historical high (MTS), and correlation (CORR) for each strategy from 2010 to 2020.

Strategy AR AV SR MDD NPMDD MTS CORR
Breakout Strategy 12% 18% 0.67 -25% 0.48 1 year 1
Mean Reversion Strategy 10% 15% 0.67 -20% 0.5 1 year -0.2
The chart below shows the equity curves for each strategy from 2010 to 2020.

Equity curves for each strategy

Using equal weight method, we simply assign 50% weight to each strategy, and rebalance monthly. The resulting portfolio has an annualized return of 11%, an annualized volatility of 12%, a Sharpe ratio of 0.92, a maximum drawdown of -15%, a net profit-to-maximum drawdown ratio of 0.73, and a maximum time span for net value not creating a historical high of6 months. The chart below shows the equity curve for the equal weight portfolio.

Equity curve for equal weight portfolio

Using optimization method, we use a genetic algorithm to find the optimal weights for each strategy that maximize the Sharpe ratio. The resulting portfolio has an annualized return of 12%, an annualized volatility of 11%, a Sharpe ratio of 1.09, a maximum drawdown of -13%, a net profit-to-maximum drawdown ratio of 0.92, and a maximum time span for net value not creating a historical high of 6 months. The optimal weights are 67% for Breakout Strategy and 33% for Mean Reversion Strategy. The chart below shows the equity curve for the optimization portfolio.

Equity curve for optimization portfolio

Using criteria or indicators method, we use the net profit-to-maximum drawdown ratio and the maximum time span for net value not creating a historical high as the criteria to rank the strategies. We then assign more weight to the strategies with higher rank, and less weight to those with lower rank. For example, we can use a linear weighting scheme, such as 60% and 40%. The resulting portfolio has an annualized return of 11%, an annualized volatility of 12%, a Sharpe ratio of 0.92, a maximum drawdown of -15%, a net profit-to-maximum drawdown ratio of 0.73, and a maximum time span for net value not creating a historical high of 6 months. The weights are 60% for Breakout Strategy and 40% for Mean Reversion Strategy. The chart below shows the equity curve for the criteria or indicators portfolio.

Equity curve for criteria or indicators portfolio

As we can see from the above examples, combining multiple strategies can improve the performance and smoothness of the equity curve, compared to using single strategy alone. However, there is no one-size-fits-all method to combine multiple strategies, as each method has its own advantages and disadvantages. Therefore, it is important to understand the characteristics and interactions of the strategies, and test different methods and parameters before applying them to real trading.


On my homepage, there is a display of the results after we optimized our strategy, which is considered to be quite good.

We always follow the philosophy of diversification of investment targets and strategies, and avoid single investment methods. Each sub-strategy undergoes rigorous stress testing before going online, such as out-of-sample testing, Monte Carlo testing (randomly scrambling order sequence, randomly deleting 5% of historical orders). We monitor the performance of sub-strategies in real time, and when they reach the worst historical performance, we remove them from the portfolio. Another way is to regularly optimize the strategies in the portfolio, and replace the worst-performing strategies with new ones.

Our trades are based on a certain percentage of risk per trade, and we also set a fixed stop loss to control extreme risk. The trading record report on myfxbook shows that each order has a stop loss, but it is not easy to touch.


I hope this article has given you some insights on how to combine multiple strategies in forex trading to achieve smoother equity curve and better returns. If you want to learn more about our trading strategies and portfolios,. You can send us a private message.

😀

Oh I believe every trader combines several different trading strategies. This is a variant of the norm. Thus, you have a certain algorithm of actions and safety net for every market movement. But you are right, each trader has his own peculiarity. And it's hard to bring it all together.