Backtesting: A Critical Study

Filed under: Learn Forex Trading |

I am currently conducting my own little study on backtesting other people’s systems. What I have done is to open an account with Zulu Trade, and I am following the top five ranked signals providers for a period of one month. This is in essence the live test (although it is being done on a demo account). The five signals providers are:

 

  1. SYSTEM HOUR SIGNAL
  2. FOR333
  3. F8
  4. DUANXIAN (GBP/USD)
  5. ARTEM1584

 

For the test, I have set my risk level to 50% (so if all of the systems experience that of equal to their worst drawdown all at the same time, then my bank will fall by 50%), and I have an equal weighting.

 

A great new feature of Zulu Trade is that you can now conduct backtests of combinations of signals providers. I did exactly this with the five signals providers combined with the aforementioned settings over the past year, and I got the following results (see Figure 1).

 

Figure 1.

As you can see, by following these signals providers over the past year, we would have achieved a very nice forty-five degree equity curve. This is an extremely good-looking graph, and we would all have loved to be following these over the past year. Based upon this, over the past year, these five signals providers combined have not had a negative month. Therefore, during our test in the next month, there should be a very high probably of a profitable month.

 

However, there are tens of thousands of signals providers on Zulu Trade, and there may be an element of something called the ‘survivorship bias’. This refers to the fact that we are excluding the thousands of providers who have failed over the past year, and only include those that have succeeded. Out of thousands of providers, it is inevitable that some have been successful over the past year due to chance alone. Therefore, this could be a misleading strategy to take. We will see over the coming month…

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Your email address will not be published.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>