Testing systems on different time bars.

Discussions about the testing and simulation of mechanical trading systems using historical data and other methods. Trading Blox Customers should post Trading Blox specific questions in the Customer Support forum.
Post Reply
redbullpeter
Roundtable Fellow
Roundtable Fellow
Posts: 82
Joined: Mon Apr 28, 2003 9:12 am
Location: London, UK

Testing systems on different time bars.

Post by redbullpeter » Wed Apr 30, 2003 4:44 pm

Hi,

I was playing with EMA crossover on some currency pairs. On one pair the system did awful using hourly bars but on a 4 hour bar the system performed brilliantly with a win:loss ratio of 7:3! I used the same time frame but just varied the time bar from hourly to 4 hours.

Is this an optical illusion?

red

Forum Mgmnt
Roundtable Knight
Roundtable Knight
Posts: 1842
Joined: Tue Apr 15, 2003 11:02 am
Contact:

Post by Forum Mgmnt » Mon May 05, 2003 3:24 am

Did you use the same number of bars in your EMA in both cases?

redbullpeter
Roundtable Fellow
Roundtable Fellow
Posts: 82
Joined: Mon Apr 28, 2003 9:12 am
Location: London, UK

Post by redbullpeter » Fri May 16, 2003 11:15 am

Yes, the number of bars were the same. Would this just be a case of curve fitting? Admitedly the time period I tested it over was only 6 months.

But this brought a little thought to my mind. How does the risk/profit profile change when you scale in or out at different time bars?

The ATR on daily bars would be significanlty larger than for example hourly bars. If you are basing your stop position based on a derivative of ATR would using longer time bars increase your risk? However, utilizing the turtle method the monetary risk would be adjusted. So would the monetary risk be equalised?

However then, given that with a narrower ATR you are able to take a larger position, would it not be more desirable to trade smaller time bars?

Is my logic completely flawed here?

red

Post Reply