has taken its toll on all the systems, as with all the strategies tested on the NAS-
DAQ stocks, it seems as if the volatility of the stocks still manages to produce
good-enough trading opportunities to set a new equity high every now and then.
Also, note in Figure 28.17 that the longest flat period for this strategy came prior
to the bull-market run in the late 1990s.
Looking at Figure 28.18, it also is obvious that this strategy didn’t do all that
well for the first half of the tested period. Looking only at the second half of the
tested period, the average annual return should be at least twice as high as the aver-
age return for the entire period.
We already know that one major reason for this is that a few of the stocks in
this test actually didn’t start trading until late in the 1990s. However, looking at
Figure 28.18, unfortunately, it also is fair to say that the entire growth in the equi-
ty comes from one lucky period, the bull market run in the late 1990s. Of course,
this is not a good thing, but we should take a closer look at Table 28.18 to see what
we can learn from it.
If you compare the average annual return in Table 28.17 with the 12-month
rolling return in Table 28.18, you’ll see that the 12-month rolling return is much
higher. How can this be?
The result from one exceptionally good period will not only influence the
result for the year it happened, but also for the 12 rolling periods following it. To
take it from the beginning:370 PART 4 Money Management
FIGURE 28.17
The equity curve for Strategy 12.