Backtest and timeseries behaviour change in new release 3621? - page 2
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
What server are you using ?
Hi,
Hi,
Today is DST Summer time switch in America. It affects your trading time and so your backtests on F40.
You keep continuing to NOT provide useful informations : logs, reports, code or screenshots. What you can see or read on your side is of no help to us to identify the issue.
Indeed. In my defense I tried in my initial post and I wasn't allowed to paste images.
As for what specifically the issues is that I should give technical detail about - I have no idea. The backtest results look completely different when running EA's over 3 years now, than what they did before release 3621. I can only take a weak stab at guessing why that might be. Symptoms include lots of Market Closed errors and incorrect dates on the backtest report.
You mentioned that DST kicked in but surely that should not affect historic testing.
I'm busy re-running a backtest and will post a comparison to what it looked like before so at least you'll see what I see.
I'll include what seems like weird dates from the report also.
As far as I known, the Strategy Tester has always correctly simulated the trade session for symbols and correctly reported "Market closed" times with an OrderSend error.
Are you stating that it now incorrectly implements the session times or are you saying that it now more accurately implements it?
Thanks Fernando.
I'll try and get a screenshot of a specific datetime where it gives a market closed error that wasn't there before. Not sure if it seems more or less accurate now - just seems new :)
I don't think the release notes mentioned any change in this regard.
Indeed. In my defense I tried in my initial post and I wasn't allowed to paste images.
As for what specifically the issues is that I should give technical detail about - I have no idea. The backtest results look completely different when running EA's over 3 years now, than what they did before release 3621. I can only take a weak stab at guessing why that might be. Symptoms include lots of Market Closed errors and incorrect dates on the backtest report.
You mentioned that DST kicked in but surely that should not affect historic testing.
I'm busy re-running a backtest and will post a comparison to what it looked like before so at least you'll see what I see.
I'll include what seems like weird dates from the report also.
Today is DST Summer time switch in America. It affects your trading time and so your backtests on F40.
DST in America makes it all very confusing and difficult to test. So before the update EA worked for 9:00 and after the update it does not ? So tomorrow (Monday 13th March ) it will be possible to trade F40 (and other symbols) from 09:00 but it will not be possible to test for trades from 09:00 up to 10:00 ? Does not make sense.
Unfortunately, it does affect it, because no historical data is kept on DST changes. So the strategy tester always applies the current session times to all past historical data, causing the results to be different.
DST in America makes it all very confusing and difficult to test. So before the update EA worked for 9:00 and after the update it does not ? So tomorrow (Monday 13th March ) it will be possible to trade F40 (and other symbols) from 09:00 but it will not be possible to test for trades from 09:00 up to 10:00 ? Does not make sense.
Unfortunately, it does affect it, because no historical data is kept on DST changes. So the strategy tester always applies the current session times to all past historical data, causing the results to be different.
Aah. Lightbulb moment. I did not know that - thanks!
I'm a little surprised by this as I thought the backtest tool on real tick data was very accurate. Applying the current DST to history however skews the results significantly. I'll post a comparison still just to illustrate.