Dukascopy vs Darwinex Historical tick data. Back-test Results Differ
you are write, drawinex data quality are better, i have noticed that also
Forex and CFD are not centralised exchanges. Quotes data between brokers will often be different.
It is not necessarily a question of quality. They are simply different.
Don't over-fit your optimisations to just one set of data.
Make your EA's logic capable of adapting and adjusting to the these possible differences.
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Hello MQL4 community,
I am having some discrepancy mismatches when comparing tick data generated from Dukascopy & Darwinex during MT4 back-testing.
I'm not sure if any one has some insight on how tick data is collected from these perspective broker providers, and how precise the tick data is for the broker that you use....
I am currently using Oanda as my demo account broker and plan on using them when I start live trading.
When i ran my back-test on one of my EA's. I ran a back-test using Dukascopy tick data, and then ran it once more on the same terminal but using Darwinex tick data. Same time frame and same input parameters were used. The spread was also the same which is seen in the reports attached.
Tick data was imported into MT4 terminal using Data Manger from StrategyQuant X. The MT4 Terminal Properties are attached for reference.
When observed there was a 54% Bias in "ticks modulated" yet the bars in test/Bars mismatches and modeling quality were identical. This is am enormous difference and when analyzing data I would prefer to stay within the 95% CI at most 90% CI. Is there a minim acceptance target or range when it comes to Modulating Tick Bias? Is there such a parameter assigned ?
(Darwinex and Dukascopy Back-test reports attached, as well as other supporting file info).
Another key note difference observed is that the Volume tick data from Historical History Center is much different from each other. Attached are screenshots for visual reference. Darwinex Tick data specifically has tick data volume reading a lengthy digit then changes to "1". I'm not sure why this is.
Overall the back-test reports generated are similar to each other except for the Profit, & Tick Modulating where it seems Darwinex may have higher tick precision based on the larger amount of ticks reported. But I can't confirm this.
I would appreciate any insight into the back-testing discrepancies seen.
Thank you very much