Chapter
Systems as Filters
You can determine the long-term trend several ways. The most straightforward
way is simply to use your own fundamental and discretionary judgment about
where the economy in general and your stocks in particular are heading, and then
stick to that prognosis over a prolonged period of time. Other ways could be to
trade only in the same direction of a long-term moving average, or some other
indicator suitable for more long-term analysis.
Over the years I proved that this works, in several articles for Futuresand
Active Tradermagazines, and also in my first book, Trading Systems That Work.
For example, in Trading Systems That Work, I examined 16 different futures mar-
kets over the period January 1980 to October 1999. First, I tested all markets 12
times each with a system that entered the market randomly in either direction, and
stayed in the trade for five days. Then I added a 200-day moving average and
altered the rules for the system so that it was allowed to enter randomly only in the
same direction as the long-term trend.
Without the trend filter, only seven markets had an average profit factor
above one, but for none of them could we say with 68 percent certainty that the
true profit factor also would be above one. For all markets combined, trading
without a trend filter, the true profit factor was, with 68 percent certainty, likely
to be found somewhere within the interval 0.88 to 1.12. With the trend filter,
however, all markets but one had an average profit factor above one, and for a
total of nine we also could say that we could be 68 percent sure that the true prof-
it factor, trading with the trend filter, also would be above one. For all markets
combined, the true profit factor could with 68 percent certainty be found in the
interval 1.03 to 1.29.
255
Copyright 2003 by The McGraw-Hill Companies, Inc. Click Here for Terms of Use.