1 (edited by petko.alexandrov 2016-05-31 20:37:58)

Topic: Different results on Real Account and FSB Pro Journal results

Hello

I have serious trouble with the results themselves. I put 4 strategies on M1 chart(using M15,M30,H1 as filters) and ran them with a small lots on a real account, to see if everything is alright.

I was very surprised to see, that after the first positions were closed, and i Reloaded the History Data in FSB Pro, the results from the account with the results in the FSB Pro Journal were different.

I attach 2 pictures - one from the chart and the other one from the Journal. As you can see the time is different and the price is different. I will attach the strategy also, may be the problem is there. Maybe i miss something..

But when there is such mismatch, that worries me a lot, about the strategy itself...

I will be very happy, if someone has an idea, what could be the reason for that.
Thanks indeed.

Post's attachments

the chart on the real account.PNG 50.97 kb, 1 downloads since 2016-05-31 

You don't have the permssions to download the attachments of this post.

Re: Different results on Real Account and FSB Pro Journal results

here is the picture from the Journal

Post's attachments

FSB journal.PNG 94.34 kb, file has never been downloaded. 

You don't have the permssions to download the attachments of this post.

Re: Different results on Real Account and FSB Pro Journal results

here is the strategy

Post's attachments

EURUSD M1 Str2 Ready.xml 14 kb, 23 downloads since 2016-05-31 

You don't have the permssions to download the attachments of this post.

Re: Different results on Real Account and FSB Pro Journal results

here is the overview of the strategy.

Post's attachments

Strategy Overview_tmp.mht 19.06 kb, 27 downloads since 2016-05-31 

You don't have the permssions to download the attachments of this post.

Re: Different results on Real Account and FSB Pro Journal results

Very good portfolio! Good job.

I tested the strategy and I see difference between the LTF indicator values in FSB and in the MT. I'll examine the case more closely.

Re: Different results on Real Account and FSB Pro Journal results

Actually, there is no difference between FSB and MT indicators data. I exported the history from MT and compared the values. (I forgot that I have to compare the MT value with the previous value in the FSB chart)

http://s33.postimg.org/3quoay0jf/screenshot_1679.jpg


I made some adjustments to the strategy in order to ensure the same values.
- changed all Smoothing methods to Simple;
- changed all Base Prices to Close.
- changed Pivot Point indicator to use D1 time frame. Why?   Because it needs M1 data for the current and the previous day in order to be calculated and this can reach 2048 bars. When I changed it to D1, it needs 2 (two) daily bars only.

These adjustments didn't change the strategy performance.

As you know Weighted and Exponential MA methods use the whole data history when determine the last value. This rises problems when trading with LTF indicators because the EA loads only limited number of bars.

There is also a trick you can use in order to force the EA to load more data:
- Open the EA for editing in the MT Editor.
- Find "Min_Data_Bars" variable in the code in the Options section.
Set some number higher than 0.

I set 1000 bars in my case.

http://s33.postimg.org/6yz5nzmt7/screenshot_1680.jpg

...

With my recent "discoveries" for reliable MT trading I recommend to everyone to use as simple expert as possible, like that:
- one time frame,
- Simple MA methods
- max 98 required bars.

Re: Different results on Real Account and FSB Pro Journal results

Popov wrote:

- max 98 required bars.

Hi Popov, where can I find this setting?

do or do not there is no try

Re: Different results on Real Account and FSB Pro Journal results

The strategy required bars are the difference between Data Bars and Tested Bars.

http://s33.postimg.org/nvstu5jqz/screenshot_1681.jpg

The example strategy above needs 64 bars.

The exported EAs need one bar more than the strategy, so an expert from the above strategy will need 65 bars.

I recommend less than 100 bars because the MT tester loads only 100 bars at start-up and there is no way to change that.

9 (edited by yonkuro 2016-06-01 07:05:51)

Re: Different results on Real Account and FSB Pro Journal results

Is there any way for FSB to only generate strategies that need less than 100 bars?

do or do not there is no try

Re: Different results on Real Account and FSB Pro Journal results

Thank you very much for the quick response!

I will do the same changes in that portfolio and see if the trades will match

i will be keeping your suggestions in mind, when looking for new strategies

thank you

Re: Different results on Real Account and FSB Pro Journal results

Is there any way for FSB to only generate strategies that need less than 100 bars?

I will make some tests.
I want to add a "Conservative" mode that will follow all the above recommendations.

Re: Different results on Real Account and FSB Pro Journal results

Thank you Popov

do or do not there is no try

Re: Different results on Real Account and FSB Pro Journal results

Hi,
Want to add more question to this topic. Maybe there would not be these differencies if we use use FSB bridge to mt4 and not exported mt4 ea? Just building my portfolio and thinking whats the benefits of these two ways to run live trading and what to choose. Thanks!

Re: Different results on Real Account and FSB Pro Journal results

Hello Mr Popov

i have had done what you  recommended me:
- changed all Smoothing methods to Simple;
- changed all Base Prices to Close.
- changed Pivot Point indicator to use D1

and i see better results and around 50 % of the trades are exactly the same in the real account and in FSB Pro Journal

but the others, still mismatching..

as far as i understood from you replay, the 100 bars are only a problem when we use MT back tester, and i don't use it

but it is it a point where the mismatches are coming from?

I also use higher time frames as M15, M30 , do you think that that could be the problem? As i saw you recoment to use only 1 time frame for a strategy..

Thanks indeed

Re: Different results on Real Account and FSB Pro Journal results

I'm not sure so far.

I noticed a wrong close of a position.

http://s33.postimg.org/odv1t30jv/screenshot_1686.jpg

A long position must close at Pivot Point R3, but it closes at around the Pivot Point.

There is no such closing in FSB Pro.

http://s33.postimg.org/97p025aiz/screenshot_1688.jpg

The FSB Shows that the closing must be at 1.12138, which is also the value shown in the expert.

I checked the PP code, but I cannot see a bug.

I'll check further.

Thank you for the reports. We must eliminate every possible issue in the program as fast as we can.

Re: Different results on Real Account and FSB Pro Journal results

I think I found the problem. It looks like there is some bug when the EA transfers the LTF indicators signals to the current time frame. It looks like it shifts the signals with 1 LTF bar.

I'll try to fix this problem.

Re: Different results on Real Account and FSB Pro Journal results

The bug is fixed. It was introduced in the last updates of FSB. I was used an outdated function in the MQL code.

Now we have a perfect match between the EA and the Trader for that strategy for all LTF indicators.

http://s33.postimg.org/xyjqd2w7f/screenshot_1689.jpg

I'll publish an Early Access Preview in the Premium Club very soon.

Re: Different results on Real Account and FSB Pro Journal results

The MT backtest matches completely the FSB backtest.

http://s33.postimg.org/exevxbcej/screenshot_1690.jpg

http://s33.postimg.org/r1u5eapaj/screenshot_1691.jpg

Re: Different results on Real Account and FSB Pro Journal results

Great!

do or do not there is no try

Re: Different results on Real Account and FSB Pro Journal results

Thank you very much for the response.

Now i will have an idea what to look also in the other strategies.

Re: Different results on Real Account and FSB Pro Journal results

Every time, before start trading a new expert, we have to compare the indicator values of the expert with the FSB chart.
We also have to compare several entry and exit signals in MT and FSB tester.

22 (edited by yonkuro 2016-06-03 23:35:48)

Re: Different results on Real Account and FSB Pro Journal results

Popov wrote:

- changed all Smoothing methods to Simple;
- changed all Base Prices to Close.
- changed Pivot Point indicator to use D1 time frame. Why?   Because it needs M1 data for the current and the previous day in order to be calculated and this can reach 2048 bars. When I changed it to D1, it needs 2 (two) daily bars only.

These adjustments didn't change the strategy performance.

Hi Popov, by changing smoothing methods to simple and base price to close, you mean like this right?
http://s33.postimg.org/vccvgj06j/SS_03062016.jpghttp://s33.postimg.org/83hu5r4bv/SS_03062016_2.jpg
as you can see the performance become worse, is there any trick to prevent this?
This strategy uses Bollinger Bands only, the result between MT4 and FSB is different. The curve is similar, and gives more profit (almost twice), but it also have larger drawdown and less trades.
http://s33.postimg.org/l0mqin4bv/Strategy_Screenshot_03062016.jpg
Could you please check it?
Regards.

Post's attachments

EURUSD H1 02062016.mq4 218.02 kb, 1 downloads since 2016-06-03 

You don't have the permssions to download the attachments of this post.
do or do not there is no try

Re: Different results on Real Account and FSB Pro Journal results

I think you may want to get a higher system quality number before working with this much more.

My 'secret' goal is to push EA Studio until I can net 3000 pips per day....

24 (edited by yonkuro 2016-06-04 02:02:15)

Re: Different results on Real Account and FSB Pro Journal results

Yes you're right, I don't think this is an ideal strategy, I also don't want to use it as a part of my portfolio. Instead I want to test various indicator in both FSB Pro and MT tester.

As Popov said that his goal with the export of MT Expert Advisors is to achieve very near results in FSB and in MT, and it's important to report any issue that is found in the program

do or do not there is no try

Re: Different results on Real Account and FSB Pro Journal results

Try to reduce Bar Close Advance to 5 seconds.
Are the spreads and swaps same?

I see you use "FSB Demo Data" data source. Are the data files same in FSB and in MT?