Ninja Trader Strategy

[maxwe789]

Just installed Ninja trader, and I'm using it with free data at the moment from openticks.

I was quite impressed at ease at which I could plug in my strategy and back test it.

However I was not impressed when it told me my strategy wasn't profitable.

I'm still undecided whether I should concentrate on discretionary trading or system trading. I like the idea of back testing a system and putting figures to it for profitability draw downs etc.

I should probably pay and get some proper data but is Ninjatrader a good platform for this kind of back testing?

 

[phlipps]

Try yahoo.com they are delayed but you can get free market data during the day. It's good to test with. BTW I like technical trading fo what it's worth. Everything else is just emotion, and you can't predict that.

 

[Adamus]

Just installed Ninja trader, and I'm using it with free data at the moment from openticks.

I was quite impressed at ease at which I could plug in my strategy and back test it.

However I was not impressed when it told me my strategy wasn't profitable.

That's disappointing but essentially the truth. Serious back-testing can take hours and days if not weeks.

The reality is probably that your chosen system will only work some of the time. So you have to back-test it on different markets or time periods, firstly to find out how to recognise when to start using it, and secondly to see how long it generally lasts when it is profitable, so you know roughly when to bail out.

Either that or it has to be good enough during the profitable times to cover the losses during the drawdowns.

 

[intradaybill]

I should probably pay and get some proper data but is Ninjatrader a good platform for this kind of back testing?

IMO the important question is whether you got a startegy that will work in the future during actual trading.

Any backtesting platform can do the job.

 

[Adamus]

Just realised the original post was back 2007. Oh well.

But I'll add more: back-testing works best when followed by forward-testing to check you haven't curve-fitted or over-optimised.