Need your help.

Ninja Trader's the new kid on the block and becoming quite popular. However, it's still somewhat primitive compared to Easy Language....and translating from EL to C# without all of the facilities that Tradestation had built in will be expensive. You will have to essentially replicate some of the 1 million lines of code behind the entire Tradestation product to get the requisite level of functionality needed by an automated trading platform.

Ninja's next release should be "da bomb" so-to-speak, but you might have to wait 3-6 months or longer.
Warning with other APIs: many are buggy and/or poorly supported or documented.
 
Are you saying you execute through IB and IB is the one having problems? You are not clear about that. Since when has IB had anything but a very excellent execution platform? I'm asking seriously because I use them.
 
Quote from syswizard:

Ninja Trader's the new kid on the block and becoming quite popular. However, it's still somewhat primitive compared to Easy Language....and translating from EL to C# without all of the facilities that Tradestation had built in will be expensive. You will have to essentially replicate some of the 1 million lines of code behind the entire Tradestation product to get the requisite level of functionality needed by an automated trading platform.

Ninja's next release should be "da bomb" so-to-speak, but you might have to wait 3-6 months or longer.
Warning with other APIs: many are buggy and/or poorly supported or documented.

How is it more primative? At the language level C# offers full multi threading, and a much more performance runtime JITing compared to EL, which I believe is interpreted at runtime, no?

Maybe NT is more primative at the app level, but in terms of building robust systems, I would go with C# before EL.
 
Quote from Mike805:

This morning was the last straw. My current primary broker went down at the open and for the first 15 minutes of today. As a result I missed executions on about 300 orders that would've resulted in a significant profit. Calling the broker resulted in a "We've been having problems this morning..." standard type answer. I realize that data issues are more likely during "peak" intervals, but I have several automated strategies that usually trade very actively during those exact times. For reference purposes I use TradeStation to generate signals and then execute primarily through IB and TS. This morning, none of my TS strategies sent orders even though they should've sent about 330 total. I managed to manually execute about 20 and missed fills on the rest. All in all I'm up mid 4 figures when it should be 20 times that. Very frustrating as I believe that opportunity NOT taken is much worse than lost money.

Anyhow, enough of the sob story, I am ready to make a transition and I am willing to consider any option except the combination I am currently using. My automated systems require real-time data as well as the prior 5 days data for about 1000 symbols. Total, my systems take between 30-500 trades a day, sometimes 50-100 will trigger at once, as they did this morning, wherein everything failed...

So, what do you guys recommend I use to trade these systems? I absolutely need reliability and stability. I also monitor about 12 accounts, so I can't watch these things trade minute by minute. Ideally, I want fully automated all day with no intervention no problems. My internet connection reliability is not an issue. Robust and stable data and execution is the key. I've tried Multicharts and there are still too many automated trading issues. Besides, I don't want to use IB.

A good firm (I am willing to go prop or retail) with a good API and the like is what I'm looking for.

I am computer literate (I can write software), but, I don't have the time nor the inclination to redo all my current easylanguage into whatever language these programs require. I am willing to hire someone (obviously any free resources would be great) to do this, but, that person will have to agree to be paid via equity in my fund that vests in a time-frame we deem acceptable, i.e. you will get paid in fund equity for time spent on this project, that equity will then become available to you once whatever software you write is tested real-time with real-money for a period of X (weeks months etc).

Any suggestions for firms, resources and programmers would be helpful.

Thanks in advance.

Mike

EVERY broker will give you fits at some point. The solution is to split your account between 2 or 3 brokers.

If a broker collapses, like Refco, would you like to be prevented from trading for 6-12 months? Having money at other brokers gives you flexibility.

Put some of your money in IB. In a national financial crisis, they are much larger and more conservative than these little "$300 daytrading margin brokers.
 
Quote from Trader922:

If you can write software why not go with something like an OpenQuant or MarketCetera type platform on a hosted server? It sounds like you need the app to be written in a language that has better error handling capabilities and can be added onto later. I am working on something similar, and intend to tie in SMS messaging when there are problems such as connectivity or order problems.

It sounds like the strategies you have make money so focus on error handling & disaster handling and that should help your results quite a bit.

I have seen somewhere there is an Easy Language to C# converter but I have no idea how well it works.

Regards,
Eric

Thanks for the information. I will look into MarketCetera as I have never heard of them. I've tried OpenQuant and wasn't too thrilled about their level of support.

At first glance, MarketCetera appears to be modular and configurable. My concern is the amount of pre-requisite software knowledge. When I say that I can write software - it is at a very basic level - like MatLab and Easy Language type programs. I have limited knowledge of C# or .Net and do not feel too comfortable taking on a large project in C#.

Thanks,
Mike
 
Quote from syswizard:

Ninja Trader's the new kid on the block and becoming quite popular. However, it's still somewhat primitive compared to Easy Language....and translating from EL to C# without all of the facilities that Tradestation had built in will be expensive. You will have to essentially replicate some of the 1 million lines of code behind the entire Tradestation product to get the requisite level of functionality needed by an automated trading platform.

Ninja's next release should be "da bomb" so-to-speak, but you might have to wait 3-6 months or longer.
Warning with other APIs: many are buggy and/or poorly supported or documented.

Can NinjaTrader monitor 1000 stocks at once and manage an open portfolio of 200-300 positions at a time? Real time stops, profit targets etc etc... Assuming adequate computing resources of course.
 
Quote from snackly:

Are you saying you execute through IB and IB is the one having problems? You are not clear about that. Since when has IB had anything but a very excellent execution platform? I'm asking seriously because I use them.

I generate orders via charts in TradeStation. Those orders go to IB electronically and then also get executed in TS. The problems occur when the TS data feed/platform goes down - the source of my orders is no longer functioning. I have another source via Fidelity, but, this source is not automated, hence when I get a lot of orders I end up missing fills because I can't keep up with them manually. In terms of execution, I find IB to be good - no complaints. The point of failure in this case is Tradestation and that is where I need redundancy or another alternative altogether.
 
Quote from TraderZones:

EVERY broker will give you fits at some point. The solution is to split your account between 2 or 3 brokers.

If a broker collapses, like Refco, would you like to be prevented from trading for 6-12 months? Having money at other brokers gives you flexibility.

Put some of your money in IB. In a national financial crisis, they are much larger and more conservative than these little "$300 daytrading margin brokers.

I do have 3 brokers at the moment, however, I only have a single source for order generation - that's the issue - where can I get a reliable & robust order generation system?
 
Quote from snackly:

How is it more primative? At the language level C# offers full multi threading, and a much more performance runtime JITing compared to EL, which I believe is interpreted at runtime, no?

Maybe NT is more primative at the app level, but in terms of building robust systems, I would go with C# before EL.
Can you say "circular buffer" for indicators/functions and can you say "maxbarsback" ?
Those two features alone are just so "missing" in Ninja. I could name a dozen more.
If you don't know the above, then you have "no clue" into the technical details of a great trading platform.
Go Home. Do your homework.
 
Back
Top