performance quite a bit. Again, the risk–reward ratio is almost one-third the size
of the original system, and the average profit per trade is cut in half. And by now,
perhaps, the time spent in the market is a little too low to make the results reliable
for real-life trading on unseen data.
Let’s round off by looking at two more versions of the expert exits system.
Table 21.17 shows the results for the stop-loss version of the system, to be com-
pared with Tables 20.9, 21.6, and 21.12. This system has a very high profit factor,
which means that it probably trades really well on a few markets. Unfortunately,
the high profit factor also is accompanied by the lowest average profit per trade
and risk–reward ratio for all stop-loss versions looked at in this test, although
they’re not as low as for the trailing-stop version in Table 21.18 (to be compared
with Table 20.10). In both cases, however, the average profit per trade is still very
good, given the short average trade length of these systems.CHAPTER 21 Systems as Filters 269