[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: What Constitutes Acceptable System Performance?



PureBytes Links

Trading Reference Links

Ross,

A couple of comments/questions on all of this. I have heard that Sharp
ratio is bad because it punishes upward volatility of your equity
distribution. This MAR ratio sounds very interesting, but I would have a
question about adjusting for changes in your account size; as the
corresponding growth rate and percentage DD would all change. 2*max
drawdown seems a little subjective. Just my 2 cents.

Gabriel 

-----Original Message-----
From: Kevin Sven Berg [mailto:ksberg@xxxxxxxxx] 
Sent: Tuesday, January 20, 2004 1:03 AM
To: omega-list@xxxxxxxxxx
Subject: Re: What Constitutes Acceptable System Performance?


Ross,

One stat that I'm using more frequently these days is MAR ratio. I give 
credit to Dean Hoffman for bringing this to my attention 
(Synergy/Checkmate). It is simply the compound annual growth rate (CAGR)

divided by the maximum percentage draw down.

I'm using more often because it has a very intuitive feel, unlike some 
other metrics. A ratio of 1.0 means you can expect a draw down equal to
the 
growth rate, invariant of your position sizing. [note: that means you
can 
use MAR to assist dialing in size vs. % risk].

What is a good MAR ratio metric? IMHO, MAR < 1.0 is unacceptable; MAR >
1.0 
is better ; MAR > 2.0 looking good, MAR > 3.0 is mighty fine, and MAR >
4.0 
is walking on water.

Other metrics? Closely related to Mathematical Expectancy is R-Multiple 
Expectancy, or Average Reward-to-Risk (Van Tharp). The calculation is 
slightly different than ME, and the by-products are useful for risk 
analysis. I prefer to see R-ME > 1.0, hopefully >> 1.0 (much greater
than 1.0).

I consider the %Win to be a "feel good" metric. By that I mean it feels
a 
heck of lot better to win than lose ... but it may not be the most 
profitable system. Another "feel good" metric is % New Highs on the
equity 
curve. Actually, I'm not demeaning these stats at all by calling them
"feel 
good" ... trading psychology is a very important consideration in system

selection.

Kevin

At 07:08 PM 1/19/2004 -0800, you wrote:
>   Would anyone care to share their views on what constitutes
>   acceptable system performance: what baseline values they consider
>   would make a system a candidate to move from testing into real
>   trading?
>
>   (If there is interest in this thread we may need to break it into
>   Position trading and Day trading.)