Choosing between two similar systems

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
MarkS
Roundtable Fellow
Roundtable Fellow
Posts: 70
Joined: Wed Dec 15, 2010 12:10 pm
Location: Chicago, IL

Choosing between two similar systems

Post by MarkS » Thu Jan 06, 2011 11:19 am

I wanted to see what people's thoughts are on the following:

I have two systems that I have been backtesting for a long time, as well as running them "on paper" for over a year. Having recently bought TB, I've re-run the backtests and walk-forwards as I find this technology much easier to use than my former software. This testing has all been done at the request of some peers that will be putting their money into a small CPO that will lock everybody up for a decade (to avoid anyone getting cold feet and trying to pull out in the middle of a drawdown) and trade the system with minimal discretion on the GP's part when it comes to the trades. [As an aside, I'm not a neophyte to derivatives, having spent time on the CBOE floor and then later trading futures/forex as a risk manager; however, trend following ("TF") as a strategy is both new and intriguing to me]

Both systems are TF systems. Across multiple timeframes and 52 markets, system A has a MAR of 1.2, while system B has a MAR around 1.0. The main difference is the total number of trades: system A has 2x the number of system B.

The initial reaction is to use System B -- less slippage, less commissions, less active management of trades, less change for human error, and so forth.

Is there any arguments to be made for using System A? I'm thinking that the potential negative costs in the real world stemming from a doubling of trades is not worth the incremental increase in MAR.

Thoughts?

Paul King
Roundtable Knight
Roundtable Knight
Posts: 207
Joined: Mon Feb 23, 2004 9:13 am
Location: Vermont, USA
Contact:

Post by Paul King » Thu Jan 06, 2011 11:35 am

Mark,

Is there any reason why you can't trade both of the systems simultaneously with some portion of the available capital allocated to each? I'd try a TBB Simulation Suite with a 50% allocation to A and B and see how the results compare to either system traded individually.

Hope this helps

Paul

MarkS
Roundtable Fellow
Roundtable Fellow
Posts: 70
Joined: Wed Dec 15, 2010 12:10 pm
Location: Chicago, IL

Post by MarkS » Thu Jan 06, 2011 11:46 am

Paul, thank you for the suggestion. I haven't thought of doing that because in essence they are the same system; not in the logic or mechanics, but in taking the same trends. For example, an ATR-based entry vs a Bollinger vs an RSI: all these systems will go long and short the same trends, just at different times. I figured that trading both wouldn't be of benefit due to this innate similarity, but that's why I now have TB - I will do what you have suggested and report back.

Paul King
Roundtable Knight
Roundtable Knight
Posts: 207
Joined: Mon Feb 23, 2004 9:13 am
Location: Vermont, USA
Contact:

Post by Paul King » Thu Jan 06, 2011 11:54 am

OK, if the 2 systems tend to take similar entries then I suggest delaying the start date by, say, 25 days for one of the systems to avoid simply doubling up on positions.

If that doesn't work then you really need to think about a sound "measure of goodness" that incorporates all of the important aspects of a trading system into one number that will tell you which one you should select.

I have a measure I call King Score that I use, but there have been many posts on the forum regarding "measures of goodness" that may be useful to you - do a search.

Hope this helps

Paul

nonpareil
Roundtable Fellow
Roundtable Fellow
Posts: 62
Joined: Fri Nov 06, 2009 11:39 am

Post by nonpareil » Thu Jan 06, 2011 11:58 pm

You may also consider the fact that the higher number of trades, the more robust it is and the higher MAR is the cherry on top.

Post Reply