3.8.1 producing mega losses compared to 3.8.0??

Previously tested/produced highly profitable strategies created on 3.8.0 now produce only mega losses on 3.8.1 when I try to Improve Strategies.

What is going on???

Wrong settings maybe? Here all strategies (30 of them that I trade live) are looking exactly the same in 3.8.1. Also, without you posting the strategy that has this behaviour, no one will be able to help you.

I have same problem after downloading new data. Different calculation of cagr dd etc. Everything is x10 and instruments are set correctly. Will try reinstall java then win 10 and let you know in two
Days.

there was no change in 3.8.1 vs 3.8.0 that could have caused this big difference in backtest. The difference must be caused by something else – different data or settings.

All good here as well, all strategies from 3.8.0 (even the scalpers) look exactly the same in 3.8.1 for me. Please attach the strategies that now produce “mega losses”. Then we could have a look.

All of my strategies have different results. This happens after i have downloaded new data in august. Actually im downloading theese data and ill put them to the 3.8.0. If it will be ok then problem would be in data or something else. 3.8.1 version ( on another computer ) works fine for me till i have downloaded data via TDD in august. Let you know in 2 hours

Update:

I just reinstalled my pc turn on win 10 , fresh instance of SQ381 and problem is there . Wrong calculation CAGR is on inbuild data too. In attachment you can find a strategy. Look at average yearly return. data si gmt+3  but even if you have gmt the difference is unrealistic  big.

When i open this in QA4 it gives me correct results …

i don’t understand, there is no CAGR in SQ 3.

Average Yearly % return is not CAGR, CAGR is Compound average yearly return. But I see only small difference of Average Yearly return between SQ (0.72%) and QA4 (0.75%).

It is caused by the improved algorithm of computing the total number of trading days in the new version 4, which affects all averages, but only slightly.

But I don’t see any extreme differences, the values are same or very simular, including % Drawdown.

i don’t understand, there is no CAGR in SQ 3.

Average Yearly % return is not CAGR, CAGR is Compound average yearly return. But I see only small difference of Average Yearly return between SQ (0.72%) and QA4 (0.75%).
It is caused by the improved algorithm of computing the total number of trading days in the new version 4, which affects all averages, but only slightly.

But I don’t see any extreme differences, the values are same or very simular, including % Drawdown.

Mark but i have on IDENTICAL settings including intstruments different results. not 3.5 but 62 on IDENTICAL SETTINGS even i try inbuild data or i try TDD downloader.

My instruments settings are same as settings inbuild fnbh instruments from the beginning using SQ3. OK so lest go to test it on inbuild data

1.Settings:

2. DATA

3: MM

4. RESULT IN SQ

5.RESULT IN QA

LOOK AT THE DIFFERENCE IN SQ AND QA: IN SQ are statistics like %AVG Return computed wrong … in QA are computed good. In my case this happens even i made fresh new install of my computer.

Update: i have installed 3.8.0 and results on same strategy  are ok. So problem should be somewhere in 3.8.1

now I see what you mean. You are right, there seems to be a bug how Yearly Avg % return is computed in version 3.8.1

I’m sorry for this. But this is the only difference between these two versions, right?

Version 3.8.1 is mostly a bugfix version related to Tradestation backtesting, MetaTrader backtesting is as same as in 3.8.0. 

So if you are using MetaTrader you can keep on using 3.8.0. The MQL generation (for MT4 EAs) will be updated to the most recent version using automatic updates also in version 3.8.0

no problem, happy to find a solution. I have installed now two 3.8.0 instances and continue work.  What i found else is that reloading strategy settings from retest to builder does not work. But with this i have problem in 3.8.0 too.

Clonex