Hello all, I'm a newbie to trading and currently evaluating NT and MC.NET(30 days trial, counting down!).
My goal is to become an algorithmic trader, so it's important to be able to backtest over 100(or more), if not all stocks of a market, and also be able to watch for signals in pre-selected N stocks from overnight full list processing. (Maybe this approach is wrong, maybe there is a standard better way?)
When I input a wrong symbol in NT, it simply hang forever....that I have to kill the program. What if I'm trading for real? I should go away from keyboard for safety.
The backtesting and optimization looks simple. And according to a post of Big Mike, there is no intrabar and bar magnifier but only MTF option left to get an accurate backtesting in an old post. Is it still true for the latest v7 or v8?
The performance report is no way comparable to MC.NET
The charts are quite easy to work with, adding indicator, and it has Volume Profile indicator though limited to 20 bars. But I feel it is not hard to understand how the C# work and I've managed to remove this limit.
After playing with the Volume Profile of NT and some simple strategy in MC.NET, I found that I prefer the object model of NT more...maybe love at first sight already as I tried NT first?
There is basket testing which is good that I can backtest a group of instruments.
It's all good that it's free as long as I'm not doing real trade. So that I can try out everything and even do paper trade on it.
I found that it is easier to work with the chart in NT in terms of bar interval. It's easy to switch from 1 minute to 1 day to 10 days chart and also I can move back in time. But in MC.NET I've selected Free Quotes as data feed and then 1 day charts are constrained with certain number of months. To show a 10 days chart I can't simply choose from the dropdown but I need to choose it from the command window instead. Then once the interval is used once I can choose from the chart dropdown. I wonder is that the limitation of the trial version of MC.NET or the Free Quotes data feed?
For chart UI beautifulness, MC.NET is better. The quick access bar at the top of the chart allows quickly switching between some options and toggling on and off the indicators.
However, I cannot move easily and intuitively using the arrow keys as in NT.
There seems to be no hanging problem in case I've input a non-existing symbol.
The backtesting, optimization and performance report are better than NT.
It comes with a few commission rules and I can even define my own. (maybe NT can too but I have not see it yet, I didn't try to find it)
However, when it comes to stock backtesting, it seems that I can only use a single "minimum" volume trading unit (or those can be input for each instrument or come from data feed automatically?) if I'm to test multiple instruments.
Biggest problem, 30 days trial is quite limited as the price of MC+MC.NET is pretty expensive for me. And I also wonder if I'm to buy MC, I can only use Easy Language, but if I'm to buy MC.NET, I can only use C#...Easy Language is quick and easy while maybe limiting in more complex or uncommon situation, where .NET will fit in.
About data feed:
It seems that if I want to evaluate the deeper aspects of both platform I should get some intraday data feed to work with. It'd be a waste if I subscribe two data feeds just because I'm evaluating two platforms. What data feed is recommended that has the greatest compatibility between NT, MC.NET and AmiBroker?
I've also checked up both the NinjaTrader and MultiCharts threads in this forum that ask why do you use them and I feel that more ppl use NT for discretionary trading while more ppl use MC for algorithmic trading.
As my goal is becoming an algorithmic trader, the backtesting, optimization and performance report(MC's term, which looks pretty) features seem to be the most important for me.
Furthermore, I've read a little bit about AmiBroker and it looks a fantastic platform in terms of flexibility, disk/CPU/memory loading and scalability(it claims that it can backtest all(around 8000) US stocks). But as the trial period is counting down in MC.NET, I don't want to try AB atm.
Maybe someday I will buy them all, but as a beginner, I hope to control the expense as much as possible until necessary. That is, maybe MC or AB is an overkill for now, that I just need to keep studying the market and do more paper trade and strategies analysis.
So, to answer my own question, perhaps I would go for mastering NT first(since it is free), then consider whether it is necessary to buy MC/MC.NET(which is a hard decision too) or try out or buy AB.
Finally, is the R language or Python considered to be the next stage(can co-exist with instead of replace) of NT/MC/AB if I found restricted by those tools?
Please feel free to correct my false illusion or incompetence.
So I've paid 100 USD to enter the elite circle to get the data.
Once I'm in, I can feel the deep pressure of Big Mike since he is facing a whole lot of user demands which in fact could be better off resolved by the forum itself.
Thus I think I shouldn't wait for an answer but instead explore on my own to relief the pressure of everyone here.
My current problem is having no idea what a proper strategy should be since I hope to find a strategy that have frequent trade(the default strategies often have 0 trade on my testing eod data) yet a little bit profitable. I found Big Mike's backtesting on ES is such one good example for my starting point.
I think using which platform isn't as much an important question than to using what strategy, I will abandon MC.NET atm and concentrate on doing what I can do with NT, and find out why NT gives such an inaccurate backtesting result to most of ppl here and whether there is simple way to workaround it.
Please don't do that (switching platform in the future). Of course, you might outgrown your platform in the future, but there's no point in learning C# and the NinjaTrader platform, and then switching to learning and using Python. There's no holy grail in platforms, and each can do what you want.
You mentioned your trading goal, but did not share any personal about your life. While that might not seem relevant, I think it's the most important factor to base your decision on.
For example, if you can work on this full-time (and want to make this your job), then go with C# and either NinjaTrader or MC .NET. But if you want to research in your free time, do not go with C# and just go with MultiCharts PowerLanguage. Otherwise, in a year from now, you would still be learning programming , while with PowerLanguage you can get up to speed quickly.
It's also a matter of personality. For example, do you like studying programming books? If so, then C# might be an option. Are you disciplined and conscientious enough to work on programming projects? And do you like debugging? Then go with R and Phyton, where you need to build more from scratch.
You also mentioned the costs. While that is a valid point, I would also take into consideration the cost of the learning curve. For example, PowerLanguage (costs money) is easy to learn and start with. R and Phyton (both free) take more time to learn and more time to get something practical out. So while it seems like a good idea to start with something free like R or Phyton, you might need to pay more in hours.
You're right my job is a progammer and I just plan to do trading in spare time. Maybe I need a very strong reason to switch the platform. Before that, I better focus on finding a profitable way that is within my risk acceptance range.