tem as broad in scope as possible, work with as few optimizable variables as pos-
sible (preferably more than three), and avoid aiming for the stars when it comes to
the performance of the system. A small but steady average profit per trade, result-
ing in modest profit and risk factors and high risk–reward ratios, is the way to go.
All the analysis we did when looking for the optimal and on average best-
working stops really was to assure stability. We did this using several surface
charts, which helped us compare several input combinations with each other and
find a combination that was surrounded by other combinations, that produced sim-
ilar results, or at least not results that were catastrophically different.
A robust system is a system that works equally as well, on average and over
time, on several different markets and market conditions. For example, if you use
a nine-day moving average on Microsoft, you should be able to use a nine-day
moving average on General Electric as well without running the risk of being
ruined. Maybe a nine-day setting has worked much better on Microsoft historical-
ly, but you must test the system in such a way so that you can trust it on General
Electric as well. At the very least, you should feel confident that if and when
Microsoft starts to behave as General Electric has in the past, you should still be
left with plenty of time to observe and correct.
The two markets might not always be profitable with this setting and General
Electric might only be so under very special conditions (such as a prolonged bear
market). But when these conditions do occur, you want to be prepared. The price
you pay for being this prepared is to allow yourself to lose a little trading General
Electric during normal conditions, and perhaps not make as much on Microsoft as
you could have using the optimal setting for Microsoft, because the nine-day set-
ting you’re working with is the one that works best on average over several mar-
kets. The robustness of each system has been examined using the performance-
summary tables throughout the book, especially when we also compared the aver-
age profit per trade with the risk–reward ratio.
But it’s also fair to say that we looked at each system’s consistency, or how well
the system is likely to perform in several continuous and well-defined time periods,
such as quarterly or yearly. It all depends on how you look at it. If you look at each
series of trades on a specific market as only that, then you are evaluating robustness.
But if you look at each series of trades without taking into consideration where and
when they were produced, then you evaluate the system’s consistency.
There also is another way to research and analyze consistency. Once you’ve
decided on your basic entry and exit rules, and tested the system over several mar-
kets to make sure it works equally as well, on average, over several types of mar-
ket conditions (uptrends, downtrends, low–high volatility periods, etc.), there is
one more thing you can do to make sure the results aren’t just a fluke. So far, a
trading sequence has contained trades only from one market. In real life, however,
a trading sequence over a certain period of time can hold trades from several mar-
kets. There is also no saying in what order these trades will occur.272 PART 3 Stops, Filters, and Exits