World Finance 100
LiteForex community on Facebook LiteForex community on Twitter LiteForex community on Google+ LiteForex community on Youtube RSS Feed LiteForex community on MQL5


Testing your own algorithm: tips for better testing results

The rules described above can be exposed in any programming language including MQL4. We will not take into consideration the refinements of programming in this language. Those who desire can find on the Internet not only ready guides, that help to realize Turtle algorithm, but also step-by-step manuals for the writing of such strategies. Let us discuss some other things.

In case if that or this algorithm is exposed in the form of programming language, the possibility appears to define its strong and weak sides using historic data. MT4 also gives such a possibility. It has to be noted that modeling of trading system behavior may have some mistakes and should not be considered as absolutely accurate.

To make the testing to give important statistic data we should take into consideration the following recommendations:

  • Within testing available time the system should make big enough quantity of transactions. If n accordance to the system the average transaction has 10-hours-period, that the strategy has to be tested at the 1000-hours interval (to receive about 100 transactions). In other words it can be exposed as a demand of representation of statistic output.
  • As the results of testing may depend on the moment of beginning of testing period, it is reasonable to test the system, changing the point of start of the test.

As the result of tests of Adviser a trader gets statement and graphic of changes of account features like balance and equity. The data analysis - it is a separate task, that, unfortunately, cannot be described in this course. It should be noted that it is necessary to analyze attentively and estimate such important test results as correlation of profitable and unprofitable transactions (in accordance to the quantity and total volume of profit and losses), profit-factor, maximal drawdowndown in one transaction, maximal drawdowndown in continual series of unprofitable transactions, and also maximal profit in one transaction and maximal profit in continuous series of profitable transactions.

You should also pay attention to the even interchanging of unprofitable and profitable transactions. The results of analyses allow to make a conclusion about the most suitable strategy of money management for given trading strategy.