Systematic trading | Page 35 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Trade another day through Systematic Trading 203 replies

Nonparametric Association Measures in Systematic Trading 57 replies

Anyone use Systematic from Alpari? 1 reply

A Systematic Approach to Markets 68 replies

Systematic Trading vs. Discretionary Trading 14 replies

Attachments: Systematic trading
Exit Attachments

Systematic trading

  • #681
  • Apr 3, 2010 3:30pm Apr 3, 2010 3:30pm
Hi Guys,

Is there a definitive resource for GP and learning it, I dont want to end up learning from a crap source? Ive got a few links from google, but as you know it can take time to filter the garbage.
Economists have forecast 9 out of the last 5 recessions
  • #682
  • Apr 3, 2010 6:09pm Apr 3, 2010 6:09pm
It don't think there are any great secrets in the field, I would just fish around on Amazon, any book which gets a few good reviews will probably cover the basics well enough. A quick search will turn up many research papers on the subject, most of which seem quite readable.

As with anything, it's not really learning the tool which is important, but applying it in an original way, that's the hard part.
The breaking of a wave cannot explain the whole sea.
  • #683
  • Apr 4, 2010 3:55am Apr 4, 2010 3:55am
Quoting markmm
Hi Guys,

Is there a definitive resource for GP and learning it, I dont want to end up learning from a crap source? Ive got a few links from google, but as you know it can take time to filter the garbage.
Kozas books and his sites http://www.genetic-programming.org/ and http://www.genetic-programming.com/

This is a good, simple tutorial http://www.genetic-programming.com/g...dtutorial.html
  • #684
  • Apr 4, 2010 6:28am Apr 4, 2010 6:28am
Excellent guys thans,

It seems the idea is not hard to understand but the implementation is the hard part, as Craig mentioned. Any you guys looked at immunlogic response modelling, I read something about it on Acrary's posts at the other forum.

I have gone back to C# to build my system as use it every day at work and know it inside out, it will cut my dev time down (compared to using C++). I can then look for areas where I need more speed and use C++ at those points.

Mark
Economists have forecast 9 out of the last 5 recessions
  • #685
  • Apr 4, 2010 1:40pm Apr 4, 2010 1:40pm
Here are a couple of links on the subject

http://intelligenttradingtech.blogsp...financial.html.

http://www.nuclearphynance.com/Show%...stIDKey=138225.

What Acrary is talking about is also known as 'Market Regime' identification, if you Google on that term, there is lot's of stuff about.

E.g.
http://tr8dr.wordpress.com/2010/03/24/market-regime/
The breaking of a wave cannot explain the whole sea.
  • #686
  • Apr 4, 2010 5:44pm Apr 4, 2010 5:44pm
Quoting Craig
Here are a couple of links on the subject

http://intelligenttradingtech.blogsp...financial.html.

http://www.nuclearphynance.com/Show%...stIDKey=138225.

What Acrary is talking about is also known as 'Market Regime' identification, if you Google on that term, there is lot's of stuff about.

E.g.
http://tr8dr.wordpress.com/2010/03/24/market-regime/
Thats excellent Craig, thanks more reading added to my growing list. Decided to go with Java rather than C# as its so similar, but seems to be better supported on Linux. Think the latest version is pretty much closing the gap in terms of speed with C++, and IB's Gateway API has a Java version.

Thanks

Mark
Economists have forecast 9 out of the last 5 recessions
  • #687
  • Apr 4, 2010 6:07pm Apr 4, 2010 6:07pm
Mono (C# for *nix) still has aways to go, I have ported my C# stuff to Mono but it ran like a tranquilized moose. As for C++ vs. Java in terms for speed in combination with TWS, it doesn't really matter as TWS is the bottleneck. The Java API to TWS is certainly more pleasant that the C++ version (which is why I created my own socket client).
The breaking of a wave cannot explain the whole sea.
  • #688
  • Apr 4, 2010 6:23pm Apr 4, 2010 6:23pm
I just updated my TWS on my windows machine, I look what the cat dragged in...it's the new IB API gateway.
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot.png Size: 27 KB
The breaking of a wave cannot explain the whole sea.
  • #689
  • Edited Apr 5, 2010 11:56am Apr 4, 2010 6:23pm | Edited Apr 5, 2010 11:56am
Quoting Craig
Mono (C# for *nix) still has aways to go, I have ported my C# stuff to Mono but it ran like a tranquilized moose. As for C++ vs. Java in terms for speed in combination with TWS, it doesn't really matter as TWS is the bottleneck. The Java API to TWS is certainly more pleasant that the C++ version (which is why I created my own socket client).
I think ill skip TWS and use their FIX CTCI , looks like less of a resource hog. Also I am thinking about speed in terms of cpu intensive stuff like GA etc, still a long way off from there though.

Rough plan in to get Quote Engine and Order Engine Interfaces up then implementations of them for IB. Then move to strategy engine and so on.

Regards
Mark
Economists have forecast 9 out of the last 5 recessions
  • #690
  • Apr 5, 2010 2:14pm Apr 5, 2010 2:14pm
I want to open an account with IB, it seems the min balance is 10k dollars. Is it possible to open the account then withdraw say 8-9k just after the account is open while I get to know the platform and test their API? i.e. Is there a minimum working balance?

Mark
Economists have forecast 9 out of the last 5 recessions
  • #691
  • Apr 5, 2010 4:14pm Apr 5, 2010 4:14pm
I don't think there is anyway around the 10k minimum, in the larger picture you're going to need a lot more than that to trade forex at IB within tolerable risk limits and reasonable commissions. Once you have the account, you can open a 'paper trade' account against which one can develop, but you still need the 10k in the 'real' account.
The breaking of a wave cannot explain the whole sea.
  • #692
  • Apr 5, 2010 4:22pm Apr 5, 2010 4:22pm
Quoting Craig
I don't think there is anyway around the 10k minimum, in the larger picture you're going to need a lot more than that to trade forex at IB within tolerable risk limits and reasonable commissions. Once you have the account, you can open a 'paper trade' account against which one can develop, but you still need the 10k in the 'real' account.
Hmm, ill need to look elsewhere for now then, until I get some kind of system built and then some consistent results im not going to be able to risk that much. Its a bit crap that they don't have a demo account so you can at least see if its something you want to use.
Maybe MB trading, hopefully they can sort their fix gateway out, in the next few months.

Mark
Economists have forecast 9 out of the last 5 recessions
  • #693
  • Apr 5, 2010 5:04pm Apr 5, 2010 5:04pm
Quoting markmm
Hmm, ill need to look elsewhere for now then, until I get some kind of system built and then some consistent results im not going to be able to risk that much. Its a bit crap that they don't have a demo account so you can at least see if its something you want to use.
Maybe MB trading, hopefully they can sort their fix gateway out, in the next few months.

Mark
Are you intending to do high frequency?

My system is currently based on H4 data, and I will trade it manually for start. The computer will SMS me when it wants to enter/close a position and I'll do it from my mobile. This will allow me to double check and ensure that it doesn't act on a bad tick for example or when some extraordinary event happens.

I will need some time to implement bullet proof execution and monitoring (2 dedicated servers in different data centers with kill switches). At the same time, I don't want to wait until that part is done. I will need automated execution only for H1 and lower.

I'm starting with a much easier to use COM based interface (bonus: free demo). I'll invest in FIX only when required (better broker which only has FIX).
  • #694
  • Apr 5, 2010 5:08pm Apr 5, 2010 5:08pm
Quoting Adal
Are you intending to do high frequency?

My system is currently based on H4 data, and I will trade it manually for start. The computer will SMS me when it wants to enter/close a position and I'll do it from my mobile. This will allow me to double check and ensure that it doesn't act on a bad tick for example or when some extraordinary event happens.

I will need some time to implement bullet proof execution and monitoring (2 dedicated servers in different data centers with kill switches). At the same time, I don't want to wait until that part...
Im only interested in Automated Trading to be honest. Ill develop against MBs FIX gateway and if it remains buggy then at least I still have a fix implementation.
Economists have forecast 9 out of the last 5 recessions
  • #695
  • Apr 5, 2010 5:14pm Apr 5, 2010 5:14pm
Quoting markmm
Im only interested in Automated Trading to be honest. Ill develop against MBs FIX gateway and if it remains buggy then at least I still have a fix implementation.
I wanted to ask this in the original reply, but I'll ask here.

Are you interested in Automated Trading or in making money with Automated Trading. Me, I was quite confused about this some time ago. But I eventually sorted it out
  • #696
  • Apr 6, 2010 3:17am Apr 6, 2010 3:17am
Quoting Adal
I wanted to ask this in the original reply, but I'll ask here.

Are you interested in Automated Trading or in making money with Automated Trading. Me, I was quite confused about this some time ago. But I eventually sorted it out
Well as much as I enjoy this stuff my main reason is to get some consistent profits, my aim is to be profitable most months of the year, and aim for around 4-8%.
Economists have forecast 9 out of the last 5 recessions
  • #697
  • Apr 6, 2010 4:44am Apr 6, 2010 4:44am
I think this thread (suprisingly here on FF) is really good from Merlin.

http://www.forexfactory.com/showthread.php?t=4446
Economists have forecast 9 out of the last 5 recessions
  • #698
  • Apr 6, 2010 3:10pm Apr 6, 2010 3:10pm
Quoting markmm
Well as much as I enjoy this stuff my main reason is to get some consistent profits, my aim is to be profitable most months of the year, and aim for around 4-8%.
I am sure that is most peoples aim (-:
No point aiming for certain figures, the market decides.

The only aim initially is not to blow your account.
  • #699
  • Apr 7, 2010 2:36am Apr 7, 2010 2:36am
Quoting spekkiefx
I am sure that is most peoples aim (-:
No point aiming for certain figures, the market decides.

The only aim initially is not to blow your account.
Very true Spekkie, my figures are ballpark and of course anything > 0 is a win. I have a lots of coding and testing before I get there though.
Economists have forecast 9 out of the last 5 recessions
  • #700
  • Apr 7, 2010 8:56am Apr 7, 2010 8:56am
Quoting spekkiefx
The only aim initially is not to blow your account.
That's my year 1 goal
Hindsight is 20/20
3 traders viewing now