Hi guys,
Mark from Helios Automated Trading here. I would like to share some of the things I learned while working as head systems developer at Helios. During my career, first as a trader and then later a developer I have developed a large number of profitable systems and hopefully I can pass on something useful to others here.
I thought I would do a series of posts here, exploring a number of popular trading ideas and testing them to see if they actually work, and if they do, how well they perform. There is so much ‘trading wisdom’ available in books and courses, but unfortunately very few of those actually show what works and what doesn’t by testing the trading concept in an objective way.
After describing and testing a trading system, in the next post I am then going to see if the system can be ‘tweaked’ in any way to make it even more profitable.
The classical dual moving average crossover system.
There are few traders who haven’t come across this simple and classic trend following method. A trade is initiated when a shorter term moving average (the ‘fast’ MA) crosses above/below a slower moving average (the ‘slow’ MA). A long trade is initiated on the next open after the fast MA crosses the slow MA to the upside, and a short trade is initiated on the next open when the fast MA crosses the slow MA to the downside.
We are going to test this system on daily bars of the EUR/USD currency pair. Our data set is going to run from 1998 through to 2010. Our fast MA is going to be a 13 period exponential moving average, while our long MA is going to be a 21 period exponential moving average. In this variation of the system, we are going to be always ‘in the market’. In other words, when an opposite signal occurs we are going to exit our present trade and reverse it (for example, we have a long position and then we get a MA crossover to the downside- we then exit our long position and we go short). At this stage, we are not going to use a protective stop loss or a profit target.
System Results:
The system generated 98 trades, of which 33 were winners and 65 were losers. The win/loss ratio therefore was 0.34 but because the winners were substantially larger than the losers, the system generated $65,109 in profits. While this is less than half as profitable as the Helios trading systems, this is not bad and we can see that the MA crossover could be used as a solid base for a good system. Lastly, the maximum drawdown for this system was $20,472, which is important to know if we want to trade the system going forward.
In the next post, I am going to add some stop losses and/or profit targets to see if we can improve the performance of this basic system. In addition, I am going to play with optimising the moving average lengths to see if this has a major effect on system performance. Until next time,
Regards,
Mark from Helios Automated Trading here. I would like to share some of the things I learned while working as head systems developer at Helios. During my career, first as a trader and then later a developer I have developed a large number of profitable systems and hopefully I can pass on something useful to others here.
I thought I would do a series of posts here, exploring a number of popular trading ideas and testing them to see if they actually work, and if they do, how well they perform. There is so much ‘trading wisdom’ available in books and courses, but unfortunately very few of those actually show what works and what doesn’t by testing the trading concept in an objective way.
After describing and testing a trading system, in the next post I am then going to see if the system can be ‘tweaked’ in any way to make it even more profitable.
The classical dual moving average crossover system.
There are few traders who haven’t come across this simple and classic trend following method. A trade is initiated when a shorter term moving average (the ‘fast’ MA) crosses above/below a slower moving average (the ‘slow’ MA). A long trade is initiated on the next open after the fast MA crosses the slow MA to the upside, and a short trade is initiated on the next open when the fast MA crosses the slow MA to the downside.
We are going to test this system on daily bars of the EUR/USD currency pair. Our data set is going to run from 1998 through to 2010. Our fast MA is going to be a 13 period exponential moving average, while our long MA is going to be a 21 period exponential moving average. In this variation of the system, we are going to be always ‘in the market’. In other words, when an opposite signal occurs we are going to exit our present trade and reverse it (for example, we have a long position and then we get a MA crossover to the downside- we then exit our long position and we go short). At this stage, we are not going to use a protective stop loss or a profit target.
System Results:
The system generated 98 trades, of which 33 were winners and 65 were losers. The win/loss ratio therefore was 0.34 but because the winners were substantially larger than the losers, the system generated $65,109 in profits. While this is less than half as profitable as the Helios trading systems, this is not bad and we can see that the MA crossover could be used as a solid base for a good system. Lastly, the maximum drawdown for this system was $20,472, which is important to know if we want to trade the system going forward.
In the next post, I am going to add some stop losses and/or profit targets to see if we can improve the performance of this basic system. In addition, I am going to play with optimising the moving average lengths to see if this has a major effect on system performance. Until next time,
Regards,
