Hi All, Fairly new to these forums... new to posting anyway. I've spent quite a lot of time trawling through the many posts on here reagrding the variations & pros/cons of different backtesting/automated trading system software. Not wanting to reinvent the wheel here... but I'm still not quite clear in my mind as to what some of the major differences are between various offerings. I should state that I am intending to trade both intraday and swing trade over longer time periods. I guess at this stage it will be roughly a 50/50 split between intraday vs longer term (up to 2 month absolute max). So far, I seem to have narrowed my own search down to four possible alternatives. 1. Ninja Trader - I've spent the last two weeks, on and off, combing through their on-line manual... and test driving the software in various simulation modes. From what I've seen so far (on 6.5), Zen Fire feed (which is lightning quick) I'm pretty impressed with the product. It looks to be a better solution for intraday traders than EOD traders, but that suits me fine. I very much doubt I will be executing at the open based off last nights close. Other possible solutions that I have been eyeing, but not yet test driven... 2. OpenQuant 3. Traders Studio 4. Trading Blox - While this looks like a very well thought out and solid backtesting system, it looks aimed at EOD traders, and since it doesn't handle automated trading (i.e. connections to broker API's) I'm inclined to just assume that this product is not going to be right for me, no matter how solid it's backtesting. For the record I should state that I have a background in writing, designing & trading statistical arbitrage strategies for a couple of big investment banks. I'm an experienced programmer in VBA & C#. Now looking to trade my own money in a more structured way, putting some of my experience to use. I am well versed with the in's and out's of backtesting, optimization and alpha generation. I've done a lot of portfolio level optimizations of risk adjusted alpha conditional on various risk parameters, within both individual position and portfolio level VAR contraints etc. However, I'm not sure I see a lot of benefit in this unless you are trading a big account... From my own perspective I'm initially hoping to concentrate on 5 to 10 liquid futures contracts/major Forex pairs. So I'm not clear what sophisticated portfolio level risk control will add. I think it would be incremental to what you'd get from some hard coded money management and risk control parameters. Anyway, in a nutshell, I defer to those of you who have more experience of this software within the context of trading your own accounts. What are your experiences? Given my needs & goals (short term goals at least), what do you think would fulfill those best - while giving me a good amount of future proofing for expanding my portfolios/asset universe...?? Any specific limitations of the various systems that I should be aware of...? Any particular features specific to one of the products that I should know about...?? etc...etc... Thanks for your time. I look forward to your views.
I have just read a post that makes clear that Traders Studio is only an EOD product... not a real time solution. On this basis, I think for my needs, that discounts Traders Studio... So it becomes a comparision of Ninja Trader vs OpenQuant. Many Thanks
I'm just curious that if, in your opinion, more sophisticated strategies like statistical arbitrage, provides superior returns to a buy and hold strategy or even a simple MA strategy? I'm just wondering if all the programming time, effort and mathematical gymnastics is worth it?
Neoticker never seems to get listed in these what platform should I use posts. Take a look at their help file and give the demo a shot if you have not already.
Panzerman... I think it depends how much money you have to invest and how big your asset universe is. I was running various European portoflio's of around $500m long/short... When you're investing that kind of money then every basis point counts, market impact is crucial, and portfolio optimization techniques add something to the mix because of the size of your asset universe/porfolio. With regards to buy & hold... well, clearly a lot depends on when you buy. If you'd bought before the great depression or the US stagflation era then you're returns would have been absolutely pitiful (inflation adjusted) over a 15 year period. I think there a danger that those of us who have grown up in the bull markets of the 90's think that 'by & hold' is a great strategy and always will be. I don't think it is... we will eventually, inevitably in my view, have another prelonged period of sub-historical growth, and likely elevated inflation. In those circumstances I think buy & hold will look very vulnerable over any time period that a normal human being could withstand. With regards to versus MA crossover... well, if I comment more generally on strategies based on measurable technicals vs statistical arbitrage... No, I'm not convinced that there is a great benefit. I've looked at both in some detail and I actually find many measurable techincals more intuitively in0line with market action than some of the more esoteric mathematical techniques. I think this is particularly true of traders in my current situation. i.e. looking to focus on a relatively small asset universe, and trade something like a $500,000 account. At that level of action I do not think statistical arbitrage will give you antyhing that measurable technicals will not. However, you have to rememeber in a competitive corporate environment there are rewards for appearing to be smarter than the next guy... and I think there is a bias towards complex mathemtical techniques in that environment. I give one example... the Index Arb guys made good money for the bank year on year... often they had flat books at the end of the year. It wa pretty vanilla stuff... but they did it well and made hard cash, no risk left on the book. However, the exotic options guys, mainly doing global exotics, also made a lot of money... but they were locked into long term risk that was OTC and hard to get out of or hedge often. Anyway, because of the complexities of the products they traded (and frankly because management didn't fully understand it) they were usually paid signifigcantly more than the Index Arb guys.... Anyway, of course, most of the exoitc options books completely blew up about 4 years later... and all the traders who'd been running them had either moved on or semi-retired on fat bonuses... the banks, and management being left with tons of unhedgable risk... Like I say, in terms of bonuses, sometimes it pays to look smarter than the next guy...
Thanks olwisepieeye... I'll take a look. p.s. sorry to post another 'what system' thread... I was just stuggling to find the info I was hunting for after hours of reading threads.
Check out this thread, it might help you: http://www.elitetrader.com/vb/showthread.php?s=&threadid=108534
I think once you do enough research it will basically come down to that there is little reason to not use OpenQuant other than the documentation isn't up to snuff for those of us who are not professional programmers. Since you are, the choice should be pretty easy.
Thanks all for your responses. What was I using at the banks...? Well, proprietary C++ software... usually programmed by myself & a quant I had working for me. But to be honest, it was all written around their in-house systems... also it had no charting functionality... The time I would spend hacking around with all that old code, I might just as well purchase something if it's available and has the right fuctionality... Thanks all.