Exported to MT4 but get awful results – don’t match SQ results
6 replies
Pierre Basson
2 years ago #276051
I have exported several great SQ strategies to MT4, I have tick data from Dukascopy using TickBuilderSuite on MT4 (2012-2021), and also tested on IG data stream. I am getting like 2% instead of 80% results shown in SQ. What is missing? SQ shows >1500 trades, but in MT4 only about 400. My data streams go back to 2005 so it’s not missing data. All my own written MT4 EA’s work perfectly with this data.
I tried every tick and open only modes but get same results.
Any ideas?
tomas262
2 years ago #276069
Without more details and specific strategy sample with can only guess …
Do you maintain the same timezone in both MT4 and SQX? You can email us the strategy, we could check …
Pierre Basson
2 years ago #276129
HI
So to provide specifics on my problem. Attached is strategy file and MT4 EA export file.
I have TickDataSuite which has 99.99% accuracy using Dukascopy tick data for EURUSD
At first it makes no trades ( see Journal below) and it appears to not size lots correctly. When I manually set lot sizes
in MT4 Journal shows
2022.01.10 12:44:45.098 EURUSD,H1: 40967931 tick events (10676 bars, 163672271 bar states) processed in 0:01:06.156 (total time 0:01:56.984)
2022.01.10 12:44:45.096 2016.09.19 10:24:34 Tester: stop button pressed
2022.01.10 12:44:45.041 2016.09.16 23:00:00 Strategy 3-76-147 EURUSD,H1: – SQ LOG 2016.09.16 23:00 LotSize is too big. LotSize set to maximal allowed market value: 0.00000000
if I switch off MM and set lot size at 0.1 it now trades but
shows error about min size – not sure if important or how to fix that.
2022.01.10 12:48:29.935 2015.01.06 15:00:01 Strategy 3-76-147 EURUSD.FX,H1: – SQ LOG 2015.01.06 15:00 Computing Money Management for order – Risk fixed % of account balance
2022.01.10 12:48:29.672 2015.01.02 00:00:01 Strategy 3-76-147 EURUSD.FX,H1: – SQ LOG 2015.01.02 00:00 Based on its logic, the strategy tried to place stop/limit order at incorrect price. Market price: 1.21048000, min. price allowed: 1.21068000, stop/limit order price: 1.14872000 (this is NOT an error)
2022.01.10 12:48:29.672 2015.01.02 00:00:01 Strategy 3-76-147 EURUSD.FX,H1: – SQ LOG 2015.01.02 00:00 Opening order type BUY STOP with price 1.14872000. Current market prices: 1.21048000 / 1.21038000
2022.01.10 12:48:29.672 2015.01.02 00:00:01 Strategy 3-76-147 EURUSD.FX,H1: – SQ LOG 2015.01.02 00:00 Use Money Management = false, MM not used
2022.01.10 12:48:29.672 2015.01.02 00:00:01 Strategy 3-76-147 EURUSD.FX,H1: – SQ LOG 2015.01.02 00:00 Computing Money Management for order – Risk fixed % of account balance
2022.01.10 12:48:26.564 2015.01.02 00:00:01 Strategy 3-76-147 EURUSD.FX,H1: – SQ LOG 2015.01.02 00:00 ——————————————————–
2022.01.10 12:48:26.564 2015.01.02 00:00:01 Strategy 3-76-147 EURUSD.FX,H1: 2015.01.02 00:00 ——————————————————–
2022.01.10 12:48:26.564 2015.01.02 00:00:01 Strategy 3-76-147 EURUSD.FX,H1: – SQ LOG 2015.01.02 00:00 MT min distance: 0.00020000, SQ min distance: 0.00000000
2022.01.10 12:48:26.564 2015.01.02 00:00:01 Strategy 3-76-147 EURUSD.FX,H1: 2015.01.02 00:00 MT min distance: 0.00020000, SQ min distance: 0.00000000
2022.01.10 12:48:26.564 2015.01.02 00:00:01 Strategy 3-76-147 EURUSD.FX,H1: – SQ LOG 2015.01.02 00:00 Warning! Min distance of this symbol is greater than min distance set in SQ! The backtest results may differ
2022.01.10 12:48:26.564 2015.01.02 00:00:01 Strategy 3-76-147 EURUSD.FX,H1: 2015.01.02 00:00 Warning! Min distance of this symbol is greater than min distance set in SQ! The backtest results may differ
pic of backtest in MT4
What else would you like to see. This makes $569 profit vs. $40k shown in SQ?
Appreciate your help on this so much…
Pierre
Pierre Basson
2 years ago #276224
So I sent more details and no response?
Was I just scammed into buying SQ?
in that it does not actually work in MT4 at all?
tomas262
2 years ago #276259
tomas262
2 years ago #276282
I have check on 3 years for tick data for EURUSD. I get pretty close match but still it could be lot better I believe. I think developers will need to debug this and try to find the reason for some of those differences. I have already reported this.
The differences can occur. We cannot eliminate that completely but we put our best efforts to do so. We cannot have 100% for all strategies and rules ever generated. That’s why we always do verification tests. If the match is not strategy could be kept working on demo preferably and for live operations I would prefer those which have lot better match on historical data
Pierre Basson
2 years ago #276322
So I am new to SQ. I am not committed to any particular strategy. I just pulled the first 3 optimized solutions from SQ to test the MT4 program building feature. Can you provide a sample of another strategy from SQ that does translate ok into MT4.
Viewing 6 replies - 1 through 6 (of 6 total)