Hi How many automated trading systems are tradeble at the same time without technical problems? The equipment is: quadcore, 2 GB, Ninjatrader, realtime data and broker is IB. The Ninja Trader Support had not enough experiance with multiple systems running at the same time with Ninja Trade. So I am hoping that someone has practical experience. Thx T
My system is a tad better than yours memory wise. I got problems loading a second system into a separate chart. Ninja is terribly in quality - until NT7 comes out, it looks more like some beginner programmers work. They run the complete thing in one thread, fromt end wise. THis means that when the second strategy loads, it interferes with the frist one (the back "testing" part to arrive at the current position overloads the UI thread and results in blocks). NT7 supposedly is multi threaded, though the implications of that sound similarly bad due to the programming model. I personally am just giving up on NT and looking for something else.
There are a lot of things that can help NT run better. Running your systems without charts running will help for example. But how many systems you can run depend on how complex the systems are and how well they are programmed.
A portfolio of 30 stocks (EoD signals but on a basis of realtime data) and about 10 intraday futures systems. Ninja Trader is connected with IB for datamanagement and orderrouting. I want to avoid a software crash caused by overload.
In another thread on ET a poster points out that IB's API limits the number of messages per second: "50 per sec for the API & 150 per sec for FIX http://www.interactivebrokers.com/d.../10-PDF0505.pdf "
With XP 64 (for being able to use more than 3gb of ram), and the hack described here: herehttp://forum.bigmiketrading.com/programmers-paradise/76-ninja-6-5-largeaddressaware-hack.html (because Ninja crash happens when you are using more than 1.5gb of memory), it's not too bad...
Note that this means flying totally blind. One can not see orders within the strategy, it is impossible to jsut get real dime updated chart on demand. Most limiting for me: one can not close out one strtegy AND ITS POSITION with one click - one has to manually remove the position from the account. Whatever you do, things get an unprofessional maintenance nightmare the moment more than one strategy runs - pretty obvious noone at Ninja ever USED ninja for automated trading. I personally need something that gives me full overview, reporting and the possibility to jump in and correct things with the smallest amount of work possible when necessary. Oh, and it helps to have something SMART ENOUGH TO ASK FOR HELP WHEN IT NEEDS IT Ninja is particularly nice in NOT asking, just stopping the strategy and - let the open positions ruin your accoung. Helpfully without stop loss, as the stop got cancelled with the position (sadly it does not bother to alos remove the associated market position). Sadly. Someone said it is the best out there (Ninja) - i find that disgusting, personally. How can the best have a standard that low? I started prototyping my own now. Too fed up. Anyone interested can inquire by PM, but do not expect anything visible until for some months. That said, there is always TickZoom right now. While a lot bad can be said about it's basic interface.... it is an open architecture, and the UI actually is open source, so can easily be extended. THis means it may be lesss feature wise, but it CAN BE EXTENDED.
Trading software advances at a snail's pace and as soon as you venture out of the little area of one system on one instrument you are screwed... it took them years to get it that we needed to have our account info and positions status available to a strategy... I don't think Tradestation has that yet... If I ramp up the income I'm going with Tickzoom and hiring a programmer, waiting for this s$%t to evolve is nuts... I don't really want that much, speed, a grid app that can handle a lot of instruments, some truly intelligent design leading to just a few basic capabilities... hmmm, an interface to DTN Satellite.. everything I really want is ENTIRELY DOABLE with current technology but I can't quite get it all without doing it myself... I think this software situation is what drove me out of automation and into screen trading.. I always thought it was my inability to escape randomness but really, the biggest struggle was with the software limitations... Edit: I remembered that NeoTicker has a grid application, I looked at the recent reviews, they say it's more stable, so I checked the data providers, DTN satellite is on the list.. maybe I need to migrate there... It seems to be most intelligently designed software, users say it can do anything but one guy mentions a two year learning curve.. Two years?? Nothing has a two year learning curve, by the time you learn it all you would have forgotten the first years worth...