trade over the long haul. Furthermore, a market that is losing money by itself can
still add positively to your equity when traded together with other markets in a
portfolio. It all depends on its correlation with the other markets. Or, to put it in
less scientific terms: It all depends on how and when each market zigs and zags in
relation to all the other markets.
When I build a system for the Trading Systems Lab pages in Active Trader
magazine, I usually test it on 30 to 60 markets, not expecting it to be profitable on
more than two-thirds of those. Preferably, I also want it to be only marginally prof-
itable on the profitable markets, but, by the same token, only marginally unprof-
itable on the losing markets. Also, the fewer the markets that stick out (whether for
good or bad), the better the system, as far as I’m concerned.
Hopefully, in the end, this will mean that all markets will go through both
good and bad periods, but at each instance there will be, on average, more markets
doing well than doing badly. Also, when one market moves from good to bad,
which is easily done, most likely one or two markets will move the other way to
pick up the slack.
So, which living creature do I want my trading systems to resemble? The
cockroach, which can stay alive a week on the fat left behind by a fingerprint, with
almost no concern for the environment or habitat.PERFORMANCE MEASURES
Have you noticed that the financial press and television frequently report the daily
price changes for a stock or an index in percentage terms, together with the dollar
changes? Believe me, this has not always been the case. When I moved to Chicago
from Sweden, back in 1997, hardly anyone presented the percentage changes, and
I was just amazed at how the supposedly financially most sophisticated country in
the world didn’t know how to calculate percentages.
Everyone I tried to talk to about this just stared at me with blank eyes, and I
found myself in constant arguments with both common men and distinguished
system designers and money managers about why doing the analysis in percent-
age terms instead of dollars will result in better, more reliable, forward-looking
trading systems. Most, however, seemed to be too blinded by the almighty dollar
to understand what the hell I was talking about.
So, while writing for Futuresmagazine, I decided to start my own little per-
centage crusade by stressing the importance of measuring a price change in rela-
tion to the price level around which the change took place. After two years of
doing that, I ended up writing a book on how measuring trading performance in
percentages pertains to systematic trading using mechanical trading systems in the
futures markets (Trading Systems That Work, McGraw-Hill, 2000).
As a senior editor for Active Tradermagazine, I continued to stress the
importance of using relative rather than absolute measurement techniques. At the4 PART 1 How to Evaluate a System
Copyright 2003 by The McGraw-Hill Companies, Inc. Click Here for Terms of Use.