1 Topic by damir321 (edited by damir321 2013-10-31 22:36:06)

  • Registered: 2010-03-01
  • Posts: 12

Topic: Bug report using Spread and Break Even

Using Strategy builder version 3.0.1
On daily time frame with 1 minute 100% intrabar scanning so data isn't a problem

I notice something strange when using spread and break even, I didn't get in depth analyzing of that but it is noticeable when you have system who has lots of trades and on longer time period so when you just look at it logic says that something is not right without looking deeper so for this purpose I created example strategy just so you can look and see what i see and  i was using time period on eurusd from 1.1.2005 till 12.9.2011 and 4 digit data.

The problem is that when changing spread it seems that there are two different calculations going for two different group of spread data and what i mean by that is that for spread increment of 0,01 to 0,49 is one calculation and for 0,50 to 1,00 second,  meaning one group spread is from :

0,01 - 0,49
1,01 - 1,49
2,01 - 2,49
3,01 - 3,49 etc

and second group

0,50 - 1,00
1,50 - 2,00
2,50 - 3,00
3,50 - 4,00 etc

to elaborate further i will write approximate range results in points (in brackets range result for that spread range) for above groups for selected strategy


group one

0,01 - 0,49  (10431 - 9644)
1,01 - 1,49  (8620 - 7863)
2,01 - 2,49  (7056 - 6299)
3,01 - 3,49  (5400 - 4535)

second group

0,50 - 1,00 (6942 - 6220)
1,50 - 2,00  (5181 - 4446)
2,50 - 3,00  (4064 - 2923)
3,50 - 4,00 (2430 - 1165)

Sum result without spread is 7851 points (Ambiguous 0 bars )

group one had 0 Ambiguous bars
group second had 1 Ambiguous bars

So you see a pattern difference from group one and second group  thats cca 2000-3000 points difference from two groups and 1 Ambiguous bar cant get that kind point difference.Apparently Break Even is making this kind of problems.Cheers

Post's attachments

example.xml 13.13 kb, 3 downloads since 2013-10-31 

You don't have the permssions to download the attachments of this post.
  • Registered: 2010-03-01
  • Posts: 12

Re: Bug report using Spread and Break Even

So nobody sees here some sort of issue ???

  • footon
  • footon
  • ◄≡≡≡►
  • Offline
  • Registered: 2009-08-17
  • Posts: 2,969

Re: Bug report using Spread and Break Even

I can't speak about the rest, but I'm not understanding it well (I'm not a native English speaker after all).

If you could explain, what you think is wrong, in a different manner - that would be brilliant!

A few things, which caught my eye: if you think something calculates wrongly, then it doesn't depend on timeframe, they all would show similar behaviour. Your example strategy has too many variables - why TP/SL settings? Price Move is difficult to track also in my opinion, to prove a point or to start bug hunting it should be the most primitive strat FSB allows to make. Only then you get a good base to dig deeper.

  • Blaiserboy
  • Blaiserboy
  • Junior Part-Time Aspiring Space Cadet
  • Offline
  • From: Toronto
  • Registered: 2010-09-06
  • Posts: 2,849

Re: Bug report using Spread and Break Even

I think we have to be assured that you have good data.... you seem to have checked it carefully.

I think you are say that this situation occurs when you change spread, ie broker spread.

Can I ask how you are gathering your data which demonstrates this problem.....

To work on this I would like to reproduce everything that you are using.

The developer will be quick to make a change if we can isolate the cause clearly.

Blaiserboy's Website

5 Reply by damir321 (edited by damir321 2013-12-18 23:40:30)

  • Registered: 2010-03-01
  • Posts: 12

Re: Bug report using Spread and Break Even

Hi thanks for reply ...English in not mine native language so i was trying to explain it best I can.

ok i tried to eliminate logic that isn't needed so you can see it easier in example2.xml, so the problem that i see is happening when you use spread in combination with Break Even you will notice it if you start playing with changing spread.So you can start from zero to 5  and look closely when you change spread from numbers
0,49 to 0,51
0,99 to 1,01
1,49 to 1,51
1,99 to 2,01
2,49 to 2,51
etc in that order

In prior post that I posted are the results "posted in brackets" of the prior example strategy when i changed the spread from zero to 4 so what i was trying to explain that there are two different calculations for using spread and that I could specify spread numbers that calculates on one way and the second way.I divided that spread numbers by group one and group two in previous post.Hope this makes it little clearer.But the best is that you try to change it your self.

The data that is used is obtained from asirikuy community who obtained the data directly from Alpari.Regards

Post's attachments

example2.xml 13.47 kb, 3 downloads since 2013-12-18 

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