Short timeframe mechanical system

[meanreversion]

Anyone out there running a purely mechanical system based on shorter timeframes (e.g. 15 mins - 1hr)? If so, which markets are favoured?

 

[grimesd]

I run my automated strategy on 2min timeframe. I'm running it exclusively on Forex at the moment, but it will work just as well on Futures. I haven't got the capital to trade futures at the moment, but if it keeps going as it is at the moment I'll be trading Futures next year.

 

[meanreversion]

Is this using Ninjatrader?

 

[grimesd]

Yes.

 

[Adamus]

2 mins time frame!? Is the ATR at that level enough to pay the transaction costs?

m.r.: I'm working on a 15 mins strategy (in NinjaTrader). It was an hourly system but then out of curiosity I tested it on 30 mins, 15 mins, 10 mins, 5 mins, 1 mins - and it worked on 30 & 15mins, but not any shorter.

Well, I say 'worked' very loosely. I mean 'showed promise but needs seriously tweeking'.

 

[meanreversion]

Hmmm, have you looked at higher timeframes, i.e. daily? A lot of noise on lower TF.

 

[Adamus]

Hmmm, have you looked at higher timeframes, i.e. daily? A lot of noise on lower TF.

Me? Yes, a lot. I don't enough capital for it. At least not for the Holy Disposable Polystyrene Cups style systems I invent. (As opposed to Holy Grails).

 

[grimesd]

2 mins time frame!? Is the ATR at that level enough to pay the transaction costs?

m.r.: I'm working on a 15 mins strategy (in NinjaTrader). It was an hourly system but then out of curiosity I tested it on 30 mins, 15 mins, 10 mins, 5 mins, 1 mins - and it worked on 30 & 15mins, but not any shorter.

Well, I say 'worked' very loosely. I mean 'showed promise but needs seriously tweeking'.

Just because the strategy run over the 2min time frame, that doesn't necessarily mean that it's in and out of the market that frequently. My trades usually last anywhere from 30min to 3 days so transaction costs aren't really a problem. Entries and exits base on moving averages not price.