homework before you set out to trade the pattern in question and analyzed the
exits in such a way that they would work on average equally as well for all sig-
nals, no matter whether each specific signal would have been traded or not.
Therefore, the testing must be done in such a way that all signals will be tested
at one time or another, together with several different exit levels. To do that,
trade the system at random several times. (To learn more about this system and
how to test other systems the same way, see Active Tradermagazine, February
2002 to June 2002.)
Once the system and all its signals are fully tested, using a random number
generator, the system can be tested according to normal principles. Table 15.1
shows the results for the system, as featured in the original article, when tested on
65 different markets. As you can see, it’s doing really well, with an average profit
per trade of $348, and profit and risk factors of 1.49 and 0.30, respectively. The
risk–reward ratio of 1.06 also is very good, and indicates that at least 84 percent
of all trading sequences will end in a profit.176 PART 2 Trading System Development
FIGURE 15.1
Microsoft traded using original expert exits system—Fall 2001.