How to get 99.90% Modelling Quality Backtesting in MetaTrader4

The modelling quality of a backtest in MetaTrader4 is an indicator of how accurate the backtest is. In this article you will see how to obtain the maximum accuracy, in short you will see how to get the 99.90% of modelling quality backtesting using MetaTrader4.

 

 

 

Introduction

In the previous articles you learned about backtesting, what it is and how to do it in MetaTrader4. If you haven’t read the article I suggest you to do it, in any case, Backtesting is that process of testing your Expert Advisor on historical data, so that you can understand the potential of your strategy and possible bugs in the code.

MetaTrader4 includes a strategy tester and some options to download free historical data, however, with these methods it is very hard to achieve highly accurate backtesting results.

Browsing the web you can probably find some resource that will allow you to download free historical data and import it in MetaTrader4, but this will require a lot of time and testing. The solution I found is an affordable commercial software that will give you the option to download accurate and reliable data, import it in MetaTrader4 and run backtests with the maximum accuracy.

The software I am talking about is Tickstory.

 

 

Problems with the standard backtest in MetaTrader4

You are probably asking yourself “why should I buy a software when MetaTrader already has a backtesting feature?”. You are right, MetaTrader4 has a backtesting feature, the problem is that this feature is not really accurate unless you have accurate and reliable Tick Data. If you don’t know what is a tick you can read the article What is a Tick in MetaTrader 4? you will find the answer there.

To obtain an accurate backtest you need to use tick data, which is all the minimal changes the exchange rate went through in the period of the backtest. When you backtest you need to choose as model “Every Tick”. This will be the most realistic simulation.

 

MetaTrader4-99ModellingQuality-1
Use Model “Every Tick”

 

The result obtained with a backtest of this type will usually be a report with a 90% modelling quality and a bar not completely green. What does this mean? It means that the strategy tester did its best with the data available and there is an accuracy of the 90%. Is 90% enough? In some cases it could be enough but many Expert Advisors need more accuracy and since you are investing money I recommend to backtest with a 99% accuracy.

 

MetaTrader4-99ModellingQuality-2
90% Modelling Quality

 

How to get Accurate Tick Data with Tickstory

In the article How to Download Free Historical Data on MetaTrader 4 you learned how to download historical data for free, that data though is not Tick Data, the data you download with that method is usually only Close Price.

As I mentioned before to obtain a 99% modelling quality you need to use a “Every Tick” model, and for this you need Tick Data. You can find some free tricky way to do download free tick data but I found that usually the data is not accurate or it takes too much time (and time is money!). Tickstory allows you to download historical data for many currency pairs. The data is reliable as it uses the Tick Data database from Dukascopy and with a few clicks you can download months, even years of historical data.

 

MetaTrader4-99ModellingQuality-4
Instruments available

 

MetaTrader4-99ModellingQuality-5
Selection of the historical period to download

 

The download time depends on how much data has to be downloaded. Once completed you can export the data directly to MetaTrader4.

 

How to Import Tick Data in MetaTrader4 using Tickstory

Once the historical data has been downloaded you can Export it to MetaTrader4 directly from Tickstory. You can specify for which timeframes you want to export the data and the working folders for your MetaTrader4 software, click and done!

 

MetaTrader4-99ModellingQuality-6
Export Tick Data to MetaTrader4

 

Obtain the 99.90% Modelling Quality

At this point your MetaTrader4 has accurate Tick historical data. From Tickstory we run MetaTrader and execute the same backtesting performed previously, the result is the following.

 

MetaTrader4-99ModellingQuality-7
99% Modelling Quality

 

As you can see there are many differences:

  • The modelling quality is 99.90%
  • The modelled ticks are many more than in the previous test
  • The modelling quality green bar is all green
  • The statistics have slightly changed

This is to confirm that a backtest executed with accurate historical tick data is more reliable than one executed with simple MetaTrader4 historical data.

 

How to Achieve 99.90% Modelling Quality Backtesting in MT4

Tickstory is a very very useful tool when it comes to download tick data and backtest Expert Advisors in MetaTrader 4, I can guarantee it will save you a lot of time and help you improve the accuracy of your backtests.

Watch the video and download the free version of the software. If you want you can also consider the paid version, it is worth the money!

 

 

Conclusion

If you are going to buy or create an Expert Advisor you absolutely need to backtest it before running it in a Demo or Live account, and you want the backtest to be reliable. Unfortunately I don’t believe that the free historical data downloaded by MetaTrader4 is enough to obtain reliable results. The affordable solution I found is to use Tickstory.

With Tickstory you can download accurate tick data and export it in MetaTrader so that your backtests will be more realistic. You can find more information about the product or buy it clicking the link below.

 

 

If you need more information regarding backtesting with high modelling quality or if you have any feedback please comment below.

Please also remember to Like and Follow Us on Social Media!

 

 

 

2 thoughts on “How to get 99.90% Modelling Quality Backtesting in MetaTrader4”

  1. Hi
    I have downloaded free Tickstory data to JFD. I show large numbers in the Data History(300 000+). When I test my EA on the strategy tester I get a full green bar on the modeling quality. Except that where it is supposed to show the %, it says n/a.
    What does that mean?
    I would really appreciate your input.
    Thank you very much
    Karl

    Reply
    • Hi Karl,
      I believe the only problem is that you are using the free version.
      The free version allows you to download free historical data and you can run the back test with that, but it will show accuracy N/A. The paid version will run the backtest to achiveve 99.90% modelling quality.

      Luca

      Reply

Leave a Comment

We use cookies to ensure that we give you the best experience on our website.