Backtesting & Optimizer Software Recommendation ?

[octrout]

Hi, I currently use CQG backtesting & optimizer tool.
But i don't think this cqg trade system software is good to use.
if you want to create simple trade system, it is pretty easy to use cqg trade system.

But when it comes to writing code just like any programming language (eg: starting with declaring variables and initializing them and write if then statement, looping.. etc.), I don’t even know if you can write it like ordinary programming language.
And also, I don’t think they even have cqg syntax guide or grammar book to teach you how to write code with some examples. they just tell you call for support, but sometimes it is really hard to get a hold of one of cqg backtesting specialist.


Is there any software that you recommend over CQG?
I think tradestation is very popular in backtesting, but I’m not sure about data integrity such as speed when you set your system in alert mode.


Thanks

 

[fat chance]

Some CQG help can be found at http://www.cqg.com/Support/User_s-Guide.aspx

I gave up using CQG for backtesting and trading system development some time ago because the "programming" language simply isn't up to the job - I now do everything in my own software. They've recently created an API so that you can link your software to CQG but I haven't tried it. You may want to ask them about what can be done with their API and Excel/VBA because it may be quite straightforward to generate signals in Excel from CQG's data and have CQG execute them, without you having to write your own application. This way you could have the benefits of CQG's good datafeed and the flexibility of VBA.

 

[johnpaterson]

I think tradestation is very popular in backtesting, but I’m not sure about data integrity such as speed when you set your system in alert mode.

Thanks

Tradestation is the way to go.

John

 

[MarkeGene]

Try TradersStudio (eod) and/or Tradecision.

 

[tony4966]

I understand that CQG backtesting result is that not reliable because it doesn't use tick data instead it uses bar data. So the best one can get in CQG is to see the result on 1 min bar (this should be fairly close to the tick data).
I want to ask those who have the experience of using CQG to do the backtest that how 'wrong' it is if we backtest a system on a 30min bar? I understand the backtesting result will be better than the real result if using tick data. But I don't think it can be too wrong, can it? I mean, for example, if on 30 bar the backtest results shows the system makes 100k over 1 year, I assume the real result on tick data would be around 70k or 75k.
Does anyone know about this issue? It would be great if someone can share their experience.

Many thanks.

 

[jamesboo]

You can also check out TraderCode. It performs backtesting and technical analysis in Microsoft Excel.

http://www.spreadsheetml.com/technicalindicators/index.shtml