Comparison of Tickdatasuite backtests and SQ tests
9 replies
lsburrows
8 years ago #113994
I have been using Tickdatasuite for backtesting using Dukascopy data for several years with excellent results. I am now running an evaluation between Tickdatasuite results using Dukascopy data and SQ results using the same data. Has anyone else performed the same exercise?
felipebr
8 years ago #131606
Hi, I’m getting very mixed results… even after read the SQ user help, finding answers in this forum and trying some different ways to import. I’m not getting reliable results from SQ to MT4 backtests. I already send e-mail to support and I will post here when I figure it out.
kzfx
8 years ago #131653
I have been using TDS (Tick Data Suite) for four years and SQ for a couple of months. Well, results from TDS and SQ using Dukascopy data can differ in a few ways. Results are approximately the same very often, but sometimes quite different. For example, a strategy with an excellent result in SQ turns to be incapable in TDS. I don’t figure out why this kind of thing sometimes happens, but it may be due to the logic – combination of indicators, rules, etc.
felipebr
8 years ago #131655
Hi kzfx, if we can’t generate strategies and compare with Tick data how to validate the strategy ? How will we know if even the optimizations are right ? Tomas, can you cleary it please ?
kzfx
8 years ago #131656
Hi felipebr,
From my experience, most of the generated strategies have similar results between SQ and TDS. In this case, I will conduct a further test by using different data. If a strategy has a quite different result between SQ and TDS, I will abandon it immediately due to lack of reliability. Anyway, I think it’s very important to see test results from several different data with SQ and a trading platform (e.g. MT4 with TDS).
felipebr
8 years ago #131657
I’m using SQ for about 2 weeks and I have been using TDS for years also. TDS allways acurate with live trading.
My results until now are very different, even in H4 time-frame.
I got the free portfolio with strategies from the last promotion and I’m live testing this one I can’t say nothing yet because it only have 1 week on demo test.
I will try to backtest this portfolio on TDS and report back.
And I’m still waiting a answer about it from support… I sent Tomas a e-mail message.
Regards
kzfx
8 years ago #131658
Which precision do you use in Settings – Data? I use M1 data imported from Dukascopy and choose “Tick simulation (slowest)” in SQ. Sure, results between SQ and TDS are not identical, but -10% ~ +10% difference in Net Profit and Drawdown in many cases. I agree that TDS is one of the most accurate backtest tools, so my final decision whether a strategy should be used or not is always made by TDS.
felipebr
8 years ago #131659
kzfx, I just sent you a private message, please check it. I use the “current timeframe, fastest”… since I’m finding strategies for H4 I was thinking that I would not need tick simulation.
Even for H4 I need to run tick simulation on SQ ?
tomas262
8 years ago #131684
felipebr,
it depends on strategies being created. When your strategy will enter and exit within the same bar you definitely need to use the tick simulation
cyberparadise
10 months ago #282521
Viewing 9 replies - 1 through 9 (of 9 total)