Anyone tried strategy building software?
Forums › ProRealTime English forum › General trading discussions › Anyone tried strategy building software?
- This topic has 11 replies, 5 voices, and was last updated 3 years ago by jebus89.
-
-
05/13/2019 at 1:07 PM #98355
Hi, has anyone tried strategy builders? Like StrategyQuant (SQ) or adaptrade?
Im fiddling around with trial-version of them.. I have not done big research, just discovered these 2 today, and it looks like its just trying everything and anything you could think of. Including different candlestick patterns and stuff like that. Obviously its expensive software so im (probably) not gonna buy any of them, but wondering if anyone here has used them.
Also very usable with PRT as you can get the code/algo/system written in a “stupid language” so its very easy to understand what its trying to do (at least with SQ) and you can just re-create them in PRT if thats ur prefered software.
I would be lying if i would said it dosnt look interesting.. I think you can customize the building blocks (RSI, candlestick patterns, ADX, open, low, high, close ++++) so u can even build/make systems with ur own personal preffered indicators and stop loss/profits ++..
Would be awsome to hear someones opinion about these kinds of softwares.. Also regarding curvefitting, it uses in sample / out of sample + a whole bunch of different parameters and tests to at least try to counter the curvefitting stuff..
Edit: StrategyQuant is the one ive been testing around with the most, and the most interesting feature i can see is the possibility to create a portfolio of strategies. Seems like you could make 1 strategy and test/shape to fit many different forex pairs, and then be able to run the 1 strategy on all the pairs and give u a backtest of the portfolio equity line as a whole. sounds pretty cool.
05/13/2019 at 2:07 PM #98362I have found these links:
https://www.prorealcode.com/topic/genetic-strategies-creation-for-prorealtime/
https://www.prorealcode.com/topic/neural-networks-programming-with-prorealtime/
in case it helps.
05/13/2019 at 2:23 PM #98364I’m used to SQ and other ones I will not mention. Like I said in the other topics Roberto mentioned, even if that softwares generate millions of strategies, there still a lot of work to be done (did I mention robustness? 🙂 ).
Yes, that kind of software is interesting because they do all the jobs at the same time: data mining + filtering + optimization + robustness tests (Monte Carlo and WF, ..), but you’ll find that at the end, 99% of strategies are rubbish. Be prepared to spend more nights to find the 1% golden pepita through the literally billions of strategies generated.
I’m not saying that this is a bad way to go because we learn a lot from this kind of experiments, but just that it goes back to what algorithmic trading really is and that this kind of software, even if the vendor strategy is to make us think that this makes it simpler, complicates it even more eventually.
Anyway, that is what I think from my own experience.
1 user thanked author for this post.
05/13/2019 at 8:01 PM #9839505/15/2019 at 8:37 AM #98524I bought StrategyQuantX almost a year ago.
After several months of developing strategies and running in demo live, I got to these points:
- Overfitting is the way of genetic building program.
- Forex market is too efficient to discover edges nowadays.
- The software itself is full of bugs, ending in continuous updates with new bugs (especially in retesting and WFA)
Don’t waste money on it.
05/15/2019 at 8:42 AM #9852505/15/2019 at 9:06 AM #98529Don’t waste money on it.
Well… it’s dangerous to write such things, the company might not like this advice and they might claim YOU do not know how to use their software properly!
The first lines are acceptable, since you highlight something that did not satisfy you, but I suggest you refrain from writing such advices!
05/15/2019 at 9:29 AM #98530Don’t waste money on it.
This is your own opinion. I don’t feel the same, but everyone should know what it means to use that kind of software: work, more work, and still more work.
05/15/2019 at 2:32 PM #98555Of course it is my opinion.
Jebus asked for opinions about these kind of softwares, I suggested him not to waste his money(more than 500 euros) on it.
By the way, their forum is full of people complaining about the bugs and noone has been able to create a profitable system for more than six months. Who claims so has no proof of it.
Just my two cents.
05/16/2019 at 9:17 AM #9858901/20/2021 at 8:40 PM #158724I wanted to revisit this topic as ive been exploring strategyquant.
I feel if I discovered that before this forum I would of been led in the wrong path definately.
But now with a few months of experience and gaining knowledge, thanks to this forum. I think this platform can be very powerful if used right.
Very expensive for me to buy. I just wanted to ask first if anyone has recent experience with strategyquant and how its going for them?
01/25/2021 at 12:18 AM #159307I wanted to revisit this topic as ive been exploring strategyquant.
I feel if I discovered that before this forum I would of been led in the wrong path definately.
But now with a few months of experience and gaining knowledge, thanks to this forum. I think this platform can be very powerful if used right.
Very expensive for me to buy. I just wanted to ask first if anyone has recent experience with strategyquant and how its going for them?
I have not used the software since i dabbled with the trial stuff. almost 1 year after making this post im fairly sure that at least for me, this would not be an option to use this kind of software. Its too much “throw stuff at the wall and hope it sticks” (and hope its not curvefit lol)
-
AuthorPosts
Find exclusive trading pro-tools on