How does the Metatrader Info MT4Config impact MT4 backtest

[shawky]

Hi
I'd like more information about the MT4Config file and how it impacts the history data created for MT4.

In particular I am trying to track down why my back test results differ significantly depending on the MT4Config file I use. I suspect the key difference relates to the reported spread but have no idea how it is affecting the back test, so I do not know where to start in order to compensate for it.
File A has spread 0.9 pips
File B has spread 4.3 pips
My export uses Bid and Ask price and not Spread() though all 3 are displayed on the chart for indication purposes demonstrating that the back test is getting a variable spread via the bid/ask price.
I use Birt's TDS to start MT4 and all slippage features are currently disabled.

Below is a back test chart showing the differences caused by file A and B.

Thanks
Chris

Background: I'm an telemetry systems Engineer, I have traded part time and full time (shares and CFDs) for 15 years, been floundering with Forex for 2 years, still trying to recover initial losses but there is light at the end of the tunnel.

P.S. I would be very happy if someone identifies my mistake, and I do hope it is just a mistake not an unsolvable problem.

[shawky]

Don't worry, Birt from EarView helped me out.
For now, to get live spread in MT4 history I am using Birt's CSV2FXT.
I would have thought that given the number of requests on this forum over a long period of time that TickStory Lite would already support exporting of live spread, given the hard work of creating FXT and HST files is done.

Regardless, keep up the good work.

[tickstory]

Hi Shawky,

Sorry for the delay getting back to you. Supporting variable spreads is still on the cards, however our first priority is to revamp Tickstory so it can support this and many more features and platforms that our users are asking for.
Appreciate your patience and support.

Kind regards.

[shawky]

Thanks
In creating the variable spread files I noticed that spreads for AUDNZD (my favourite) are significantly higher the further one goes back in history.

My current challenge (and everyone else is my guess) is creating back test data that simulates live trading for the current time frame. The spread differences are huge and I bet this impacts all of the indicator calculations that work on Bid price alone. How to solve this I can't say at present but it is causing my expert no end of grief on two live accounts, one has wide spread, the other very narrow.

Any comments on this topic welcome.

All the best

[tickstory]

Hi Shawky,

I'm guessing that you have an EA that's particularly sensitive to spread. What you could do is work with a fixed spread and break back-test down into yearly time-periods. This would at least give you some indication of spread level and year(s) that your EA starts to degrade and perform further analysis from there.

Hope this helps.