How to Optimize an Expert Advisor Using MetaTrader 4 Strategy Tester How do you correctly optimize expert advisors using the MetaTrader 4 strategy tester? Why do you need optimization at all and which expert advisor parameters should be optimized to achieve profitable currency trading on Forex? You will find all these answers in this article. MetaTrader 4 Strategy Tester has two important functions: the testing and optimization of expert advisors. Optimization is the process of selecting the optimal parameters for expert advisors. As a rule, a trader optimizes (improves) the input parameters of robots to make them as profitable as possible for minimum risk. The expert advisor optimization process is one of multiple testing using different input parameters in automatic mode by MetaTrader 4. That said simple optimization is nothing more than adjusting the expert advisor parameters to reflect history, i.e. to previous market conditions and price movements. If you optimized an expert advisor on historical data and set it to trade, pleased that the optimization results had been great, you wouldn’t receive the same results during real-time trade. The reason is you only performed an ADJUSTMENT of parameters, but not real optimization. Correct and qualitative expert advisor optimization, or forward testing, includes both stages. The first stage is optimization (or, to be more precise, adjustment) of an expert advisor on a certain segment of history, which is called a test (historical) period. The results of the adjustment are presented in a table in the form of expert advisor input parameters. Using the best parameters, a robot is set to “trade” on a new time segment where it hasn’t been adjusted and doesn’t know how to behave. This segment is called a forward period. If an expert advisor shows good results (it is no longer optimization!) on this segment, it means you can use this expert advisor for trades on a real account. Test and forward periods differ for expert advisors working on different timeframes. Here are the recommended historical and forward periods for expert advisors working on the following timeframes: H1 – the recommended historical period is 2 years, forward period – 6 months; M30 – 1.5 years and 4 months; M15 – 1 year and 3 months. It is not recommended to test expert advisors on smaller timeframes. In order to better understand historical and forward periods, look at the picture below: For example, we can assume today’s date as November 30th 2011 and we have decided to optimize an expert advisor on the period M15 with a historical period of 1 year and the forward period of 3 months. We would therefore have the end of the historical period coming in at August 30th 2011 and the
10
Embed
How to Optimize an Expert Advisor Using MetaTrader 4 Strategy Tester
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
How to Optimize an Expert Advisor Using MetaTrader 4
Strategy Tester
How do you correctly optimize expert advisors using the MetaTrader 4 strategy tester?
Why do you need optimization at all and which expert advisor parameters should be optimized to
achieve profitable currency trading on Forex? You will find all these answers in this article.
MetaTrader 4 Strategy Tester has two important functions: the testing and optimization of expert
advisors. Optimization is the process of selecting the optimal parameters for expert advisors. As a rule,
a trader optimizes (improves) the input parameters of robots to make them as profitable as possible for
minimum risk. The expert advisor optimization process is one of multiple testing using different input
parameters in automatic mode by MetaTrader 4. That said simple optimization is nothing more than
adjusting the expert advisor parameters to reflect history, i.e. to previous market conditions and price
movements. If you optimized an expert advisor on historical data and set it to trade, pleased that the
optimization results had been great, you wouldn’t receive the same results during real-time trade. The
reason is you only performed an ADJUSTMENT of parameters, but not real optimization.
Correct and qualitative expert advisor optimization, or forward testing, includes both stages. The first
stage is optimization (or, to be more precise, adjustment) of an expert advisor on a certain segment of
history, which is called a test (historical) period. The results of the adjustment are presented in a table
in the form of expert advisor input parameters. Using the best parameters, a robot is set to “trade” on a
new time segment where it hasn’t been adjusted and doesn’t know how to behave. This segment is
called a forward period. If an expert advisor shows good results (it is no longer optimization!) on this
segment, it means you can use this expert advisor for trades on a real account.
Test and forward periods differ for expert advisors working on different timeframes. Here are the
recommended historical and forward periods for expert advisors working on the following timeframes:
H1 – the recommended historical period is 2 years, forward period – 6 months;
M30 – 1.5 years and 4 months;
M15 – 1 year and 3 months.
It is not recommended to test expert advisors on smaller timeframes. In order to better understand
historical and forward periods, look at the picture below:
For example, we can assume today’s date as November 30th 2011 and we have decided to optimize an
expert advisor on the period M15 with a historical period of 1 year and the forward period of 3 months.
We would therefore have the end of the historical period coming in at August 30th 2011 and the