I'm an experienced professional software developer, looking to develop trading strategies for a client who's a pro trader. After struggling around with TradeStation and EasyLanguage, we've decided we need something a little less...bad. Ideally I'd like to work in a C#-based trading platform. We've zoomed in on NinjaTrader, Multi-charts and Wealth-Lab. What I'm looking for: 1) Most relevantly: the platform should get out of the way and let me program. I don't need any drag-dropping or other cute things, and more importantly, I don't want framework/API design that make things easy for beginner developers but get in the way when you try to implement something more complex. I basically would like a platform that will automatically do common things like charting and back-testing but leave me a blank slate to implement things in whichever way I need; 2) Integration with Visual Studio would be a big plus because I seriously doubt that anyone else has developed debugging capabilities up to par with VS; 3) Transforming a backtested strategy into an automated trading system should be seamless; 3) Please point out weird quirks and/or limitations (hey, did you know that in EasyLanguage, you can't backtest GTC orders?). Thank you in advance!
Very few probably have experience in 2 or 3 of these. I only know NT. 1) NT lets you do anything, but within the scope where the code is called. If you need custom beyond Indicator / Strategy-level, you may need to roll your own. I'd recommend using an established platform to prototype on, and refactor into more custom later when there's a real need for it. Be open to change opinions later. 2) Dunno about VS and why you'd want to use that. If you're an experienced programmer, the feedback from charts is good enough and you can debug using extra plots for that. Too much IDE features quickly gets in the way of the creative process. 3) I believe NT supports this using the Strategy class, which you can build from indicators. 4) EasyLanguage is a joke. Using C# you can work around alot of cruft. Beware though that a framework may degrade performance, but often you can find ways around the slow standards, ie. by rolling your own arrays instead of using BarData etc. I often do workarounds for stuff I know can be slow or use alot of memory, but usually avoid premature optimization.
WLP - does not do automated trading. TS/EL is light years above WLP. Why not use Python? It's the future and what all those quant nerds are using.
You can use a general-purpose roll-your-own solution when you know exactly what you're going to be doing and needing. Otherwise, it's wise to use something to make rapid prototyping and testing possible. Roll-your-own may be very wasteful if you get bogged down in technical details instead of creating value. Saying this with 30+ years programming experience. Good programmers code fewer flaws. Great programmers code as few lines as possible. Elite programmers strive to avoid programming entirely.
4th option - Consider making your own platform. Depending on what you're trying to do, it might not be that much work. If you're ultimately aiming for a completely automated solution, you can cut out all the unnecessary parts (charting / visualization) and optimize the whole thing to be super efficient.
Multicharts.net is lovely software, but I would not recommend it for general-purpose automated trading. It is very optimized to the use case of "write a straightforward indicator that provides some signal, which the user can then decide whether or not to act on". In theory that's automatable, but moving beyond writing simple indicators is very painful. Furthermore, the .NET version is relatively new compared to the EasyLanguage version and it is very obvious which one was made first. The .NET version feels more like a reimplementation of all the EasyLanguage stuff, with all of the same limitations. For example, you can't actually create a trailing stop order at the same time you enter a new position programmatically. Don't just take my word for it, go try it -- it doesn't work. You have to have already made a profit on the position or it will just exit immediately. The combination of mutable objects (you have to derive your class from their set of objects) being the way to work with their system, combined with their multithreading implementation, will cause maddening bugs. The moment you try to do anything more complicated than a basic indicator, you will run into this. Solid software for trading if you've never touched options before (it does not handle options at all), not at all recommended for as a framework.
HI , i am also looking for the software coder for my strategy to develop and implement on the trading chart and backtest the results .. Plz can anyone help me out . MULTICHARTS and METATRADER5 are the .NET based .. this software are proffesional and compatabile .
Hi, I can code and backtest trading strategies for MultiCharts, NinjaTrader, Wealth-Lab and other trading platforms.
Wealth-Lab is the best platform for a large portfolio backtesting. For example, all NASDAQ or NYSE stocks. NinjaTrader: good for indicators and simple strategies. There are some limitations and bugs in strategies backtesting and autotrading. MultiCharts.NET: I still prefer PowerLanguage for a more compact and clear code.
You may want to look at using some of the math/statistics/scientific programming software that is being tweaked for securities use. Such as matlab, r, mathematica, maple, etc. If you do, please also contribute to those communities and share your work. Post back and let us know how it works out.