The Pitfalls of Forex System Creation

Filed under: Learn Forex Trading |

 

The human mind naturally seeks out patterns wherever it looks. If you look at the moon you see a man’s face. If you look at some clouds, you see a rabbit or a sports car. And it is the same when you look at a forex chart in that you tend to see patterns that are sometimes not really there.

 

Screen shot 2013-03-05 at 11.15.16What happens when you try to create a forex trading system is that you tend to spot all of the good cases, while overlooking and disregarding the plethora of bad cases. This is why mechanical backtesting is so important. Another thing that you really need to focus on considering is the spread. I once created a system that was unbelievably profitable and stable with a minimal drawdown when not considering the spread, but once the spread was added to the equation, it became a loser. This is the power of the spread. It can take away your market edge in an instant.

 

There is also a tendency in system creation to over-optimise. Optimisation refers to fine-tuning your system and removing any variables that do not add to the profit figures. However, it is possible to over-optimise. For example, you might find that trades made on a Thursday after 2:30pm are not profitable. However, this is just a random occurrence, and there is no logical reason behind it.  The idea then, is to have a logical hypothesis, and then to test that hypothesis  – not the other way around. Thus if you over-optimise you will end up with a system that is just not feasible, and that is unlikely to repeat itself in the future. Ultimately, what you really need to test a system is some really good trading software and a working knowledge of mathematics (and the jury is still out on MT4’s strategy tester).

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>