A question about exporting data to the MT4...

[thomas2004ch]

Hi TickStory,

Beofre I export data to the MT4, I will disconnect the MT4 to my broker and delete all the historical datas fro mmy broker. Then I begin to export the TickStory data to my MT4. I thought the data will be export/download to the "Store location" and only when I start the MT4 from TickStory, the historical data will be read into the MT4.

But I notice, after finishing exporting data, I start the MT4 not from TickStory but manually, I find the historical datas cold be shown in MT4. This means during exporting datas the TickStory will not only save the datas to the "Store location" but also into the MT4. Is it right?

The backgroud to my question is: I want to export one-year datas and two-year datas and save them separately. The reason I do so is because I notice, if I use the one-year data to do the backtesting, assumed from 01.06.2017 to 10.06.2018, the result is quite different from that when I use the two-year data but select the same time period, i.e. from 01.06.2017 to 10.06.2018. This sounds unbelievable, right?

If I can do so, how can I configure TickStory or my MT4 to use the one-year or the two year datas?

[tickstory]

Hi Thomas2004ch,

In response to your questions:

This means during exporting datas the TickStory will not only save the datas to the "Store location" but also into the MT4. Is it right?

The "Download" function only downloads the specified data range from the internet and stores it locally. The "Export" functions export to either File or MT4.

the result is quite different from that when I use the two-year data but select the same time period, i.e. from 01.06.2017 to 10.06.2018. This sounds unbelievable, right?

It's possible depending on what your EA does. For example, if you happen to start trading during a period where there is heavy draw-down, you could break the bank earlier than if you started during a good period for your EA. You'll need to analyze why there are differences as it might be highlighting a problem with your strategy.

If I can do so, how can I configure TickStory or my MT4 to use the one-year or the two year datas?

You can try export the entire 2 year period and then select the date range you wish to test in Metatrader's Strategy Tester (be aware that the date ranges are sometimes problematic if your FXT files are larger than 4GB). Alternatively, you can try using 2 separate MT4 installations and separating your export ranges & testing in that way.

Hope this helps.

[thomas2004ch]

Hi TickStory,

Now I do as follow:

1.
I install the same MT4 two times separately in different location as serv1 and server2.

2.
I delete all the historical datas from the broker.

3.
I export the EURUSA data from TickStory from 2017.0101 to today to the server1. And I export from 2016.01.01 to today to server2.

4.
I start the backtesting one after the other on the same EA and use the same period: from 2017.06.23 to 2018.06.11. But the result are quite different. See attached PDF files.

[tickstory]

Hi Thomasch2004ch,

At a glance, it looks like your first report is reporting "90%" modelling quality which would seem to indicate that it is not Tickstory data you are testing with. Perhaps this is the cause? It certainly looks like that first trade you take in either back-test are different so you will need to identify the cause of this. If you are having difficulty, the best course of action might be to go back to basics and just print the first x prices prior to a trade occurring and seeing if you can identify any differences that would affect your strategy.

Regards.

[thomas2004ch]

Hi TickStory,

I re-do the backtesting on server1 again. The modelling quality is still 90%. I compare the datas from MT4_1 and MT4_2. They are identical (see screenshots). What could be the reason?

[thomas2004ch]

Here is the last screenshot.

[thomas2004ch]

I deinstall the server1 and reinstall it again. This time I get the 99.90%. The results are the same.

[tickstory]

Hi Thomasch2004ch,

Thanks for the screenshots. Now that you've established the price data-sets are the same, the only variations can be:

1) Something is different in your "MT4 Export settings" (such as lot size, etc). You can make sure this is not the case by loading the FXT files in the FXT Editor and ensuring all settings are the same). If you have exported with default settings, they are likely to be the same.

2) Something is different in your Strategy Tester setup such as spread, initial capital, EA parameters, etc, causing different trading conditions and/or;

3) There is something in your EA that is causing it to trade on a different price, time or condition. This is something we unfortunately can't help you trouble-shoot, so you will need to look further into why your EA is trading at different times. As previously mentioned, this could be due to many factors such as an differing MA at the beginning of the trade sequence. This can obviously have differing affects on your results depending on how your EA is designed.

Hope this helps.

[thomas2004ch]

Hi TickStory,

Sorry for my last reply. What I mean "The results are the same" is the result from server1 and server2 are the same. This means the problem is solved.

[tickstory]

Thanks for clarifying, Thomas2004ch! And good to hear