- This topic is empty.
-
Posts
-
Hello Jay-R,
very good job! This is an important test that everyone should do.
First, you are doing the test on M1, which is the most active time frame and normally there come to the most differences between the Data.
The first strategies match perfectly. Strategy number 3 I see it matches just the recent period is different…
Actually the difference with you very very small….and there are 2 reasons for me from what I see:
1. The number of trades that are in your Demo and Live data is different. Small difference but it can cause small differences in the backtest.
2. I think there is a gap in your data. Meaning that there are days missing.
You can check your data if you go to Data – Data Statistics – Statistics.
Look at Maximum Days Off. If there is no missing data, it should be no more than 2-3 days (weekends and holidays).
If this is the case, there might be different data missing in your Live and Demo data, which might coauses the difference.
BUT your data is just fine! The tests are fine. If couple of trades are different, it is alright.
Difference normally we consider when the backtest is totally different.
But in your case, it is just because of missing bars, which is normal for most brokers and it does not affect the future trading.
Regarding the data, we expect new data in EA Studio that will be 100% matching and without any missing bars.
I hope it will be ready next month.
Cheers,
I appreciate the very detailed answer, Petko. I’ll do what you suggested.
Regarding what you mentioned that “new data in EA Studio that will be 100% matching and without any missing bars.” Can you explain that further?
You are welcome Jay-R.
Yes, the thing is everyone has issues with the Historical data, just because the brokers do not really care about that. They do not provide enough data, or there are gaps, missing days etc.
The best data so far is from Ducascopy, but the problem is that they provide tick data which is not applicable in EA Studio and FSB Pro.
So now from Forex Software they are developing a program that will convert this data from ticks to bars and we expect that the result will be fanstastic….no gaps no missing days and 100% match between Demo and Live.
I will let you know when it is ready.
Cheers,
Very interesting topic here, Jay-R,
I will do such tests for my data as well.
the feedback from Petko is always useful.
I would be looking for this historical data from Ducascopy.
I was planning to open an account with them but I will wait for the data, because if I just download it and it is tick data, it will not work for me,
