certainty, fall somewhere in the interval 0.83 to 1.19, while the true percentage of
profitable trades will, with 68 percent certainty, fall somewhere in the interval
48.48 to 54.05.
In Table 21.2, 21 markets have a profit factor above one, and 19 markets
have a higher profit factor than in Table 21.1. Of all 30 markets, we also now can
say that 11 of them also have a high enough profit factor so that we can be 68 per-
cent sure that their true profit factors will be above one. Using the trend filter, 21
markets also had a higher percentage of profitable trades, as compared to with-
out the filter. Of these 21 markets, for 14 of them, we also can say that we can be
68 percent certain that the true percentage of profitable trades is over 50 percent.
For all stocks combined, the true profit factor will, with 68 percent certainty, fall
somewhere in the interval 0.85 to 1.35, while the true percentage of profitable
trades will, with 68 percent certainty, fall somewhere in the interval 48.77 to
56.69.
Although the test shows that we still can’t be sure that the average profit fac-
tor for all stocks will be above one, it still shows that most stocks will benefit from
a trend filter, even as rudimentary and arbitrarily chosen as this one. We still can’t
say with 68 percent certainty that the profit factor for all stocks will be above one
because a few stocks didn’t like this trend filter at all and therefore contributed to
increase the standard deviation of the results.
In the same article, another test showed that it was possible to increase the
performance of a short-term system traded on the S&P 500 futures market, by only
taking the trades in the direction indicated by a 200-day on-balance-volume indi-
cator. In short, this test lowered the number of trades from 332 to 177, while at the
same time it increased the profit factor from 1.21 to 1.56 and the number of prof-
itable trades from 53.31 percent to 58.19 percent.
While the on-balance-volume indicator is one of my favorite trend indicators,
because it incorporates all the concepts of technical analysis (price and volume)
into one simple-to-understand number, the lookback period for this indicator too
was arbitrarily chosen. Plenty of other tests in Trading Systems That Work, on other
futures markets, using other systems and filtering techniques, confirm the impor-
tance of always trading with the trend.
That said, this isn’t exactly what this chapter will be about. Yes, we will apply
three different filters to the five short-term systems we have worked with in this
section, the three filters being the more long-term relative strength systems we
introduced in Part 2. However, the major difference between the testing we will do
here and the tests just mentioned is that, for the tests to come, we are not all that
interested in the actual long-term trend of the markets, but rather if the market we
intend to trade shows a greater potential for moving the way we want it to, as com-
pared to one or several other markets. As a consequence of that, we might or might
not trade with the direction of the long-term trend, and at this point I have no idea
what the results will be.258 PART 3 Stops, Filters, and Exits