To code or not to code?

[gman997]

First off I started a few months ago in my spare time to work on an ATS in C# using the IB API. It's been painful in terms of just trying to come up with an entry/exit strategy (have not found one yet that would produce even a slight profit after commissions). Writing code to test, writing code to actually do the trades in a simulated environment, etc.

So today, after doing some research on ET I came to the realization that there is software out there that you can buy/use for automated trading such as I seen NinjaTrader and others.

Here comes the crossroad junction where I need to make a choice on a path to go....

So the one question right now is what is the pros of creating your own ATS when it seems there are others out there, why spend all the time and headaches in doing your own????????

 

[trade4ever2day]

Quote from gman997:

So the one question right now is what is the pros of creating your own ATS when it seems there are others out there, why spend all the time and headaches in doing your own????????

I have been developing and testing system for 10 years. The reason you want to have a system is because that is supposed to offer you an edge. As soon as you plug it in a platform of some company, nothing can assure you that they cannot see its logic or attempt to fade its signals.

For me the best route is developing you own code and keep separate from the platform that places the trades. Just feed you trades to it.

Alex

 

[Shreddog]

Quote from gman997:

First off I started a few months ago in my spare time to work on an ATS in C# using the IB API. It's been painful in terms of just trying to come up with an entry/exit strategy (have not found one yet that would produce even a slight profit after commissions). Writing code to test, writing code to actually do the trades in a simulated environment, etc.

So today, after doing some research on ET I came to the realization that there is software out there that you can buy/use for automated trading such as I seen NinjaTrader and others.

Here comes the crossroad junction where I need to make a choice on a path to go....

So the one question right now is what is the pros of creating your own ATS when it seems there are others out there, why spend all the time and headaches in doing your own????????

Might I suggest that you put the horse back before the cart and develop a system that works before trying to code it up?

For that there's no need to write your own backtester, there are plenty out there. Then, once you have a strategy that works you can decide whether to use an existing ATS system or build your own.

Automating without a working strategy is simply a way to lose money faster.

Good luck!

 

[maxpi]

If you can do C# Ninja Trader's environment should work fine for you although it is not without it's bugs, but they get fixed... once you have your system(s) designed you can code them in Easy Language and run them in Multicharts, it can run without phoning home so the chart developers can be kept away from your strat if you really think they would want it...

 

[rosy2]

Quote from gman997:


So the one question right now is what is the pros of creating your own ATS when it seems there are others out there, why spend all the time and headaches in doing your own????????

gmann997,
you have come to the realization that it is just not worth the effort to actually work. I totally agree with you and can help. I happen to have an automated system that hooks into IB or any FIX datafeed and it is yours for a special low price. Feel free to PM me.

 

[silvermotion]

Quote from rosy2:

gmann997,
you have come to the realization that it is just not worth the effort to actually work. I totally agree with you and can help. I happen to have an automated system that hooks into IB or any FIX datafeed and it is yours for a special low price. Feel free to PM me.

gmann997,

Actually it is worth all the trouble and the headaches to program your ATS system. Of course you need to start with a strategy you know works, because automating a bad strategy will give you bad results. automating a good strategy will give you good results.

I also strongly advise you stay away from strangers offering you candy on the internet especially when theres a ''special low price'' offer. If this guy autotrader would be working and profitable he would not be selling it because once you have a ATS that works its basically a money generating machine and its worth millions, he would jealously be keeping it for himself just like the greedy money hungry traders we all are (or wish to become) on this forum.

 

[decade]

Quote from gman997:

So the one question right now is what is the pros of creating your own ATS when it seems there are others out there, why spend all the time and headaches in doing your own????????

Modern trading platforms are fully customizable - you can create any complex strategy and even specify optimization algorithm and custom strategy performance measurements.
While they free you from unneeded and hard-to-code details.

You should give a try to NeedForTrade Studio (www.needfortrade.com). They have free version and a fully functional trial. Our company started using it 2005. We've implemented a broad range of strategies - from very simple to very complicated(we are using wavlets, neural networks and so on).

In NeedForTrade you can write strategy logic on C# or VB without using any object-oriented construcions - just simple logical operations, loops, etc. But if you need it object-oriented power is instantly available.

Even if you are an experienced programmer - why not to leverage speed boost provided by such systems. They let simple things to be simple, yet allowing you to write very complicated strategies with ease.