trades, the standard deviation of the returns will also decrease. Thus, the system
will be less risky to trade, and consequently you can dare to take a larger position,
which in the end might result in a larger profit than what would have been the case
with the original system.
Note also that the profit target in Figure 17.5 actually managed to increase
the average profit per trade and decrease the standard deviation of the returns at
the same time. Although this is not very likely to happen in a real system, it is a
possibility that only proper research and trade management will give you. Yet
another reason why a system with a stop loss and a profit target might be more
profitable than a similar system without those features, and despite a lower aver-
age profit per trade, is that the shorter trade lengths will make it possible to trade
more often on more markets.
As a final observation, note that we have taken a set of close to normally dis-
tributed trades and turned them into a distribution with a few very distinctive out-
comes and a few trades scattered in between. That is, adding a set of thoroughly
researched exits, such as a stop loss and a profit target, makes it possible to calcu-
late exactly what to expect from each trade in precise quantitative terms, instead of
just “knowing” that each trade can either be a winner or a loser ending up some-
where within the normal distribution curve. More important, however, is that the
end result of your trading becomes a function of your skills as a trade manager,
instead of random market fluctuations. And you don’t want your trading results to
be a function of randomness, do you? Let’s state that a little differently: If the out-
come of your trades resembles the bell-shaped normal distribution curve, most like-
ly the outcome of your trading (no matter if you’re a winner or a loser) is a func-
tion of randomness and can be explained only as good or bad luck, instead of skill.
Now, we could stop here and conclude that it pays off to do the homework
and think about a problem all the way to its end. But have we thought about it all
the way to the end? No, we haven’t. The truth is, we haven’t even started. In fact,
all we’ve just learned is wrong for one very specific reason: Every time we make
a change to a system, no matter how small it is, we change its characteristics and
need to start the research all over. For example, when you add or change your stop
losses and profit targets, the trade length of several of the original trades will
change as well. Most likely, adding a profit target and a stop loss will shorten the
trade length. And shortening the trades means that we free up money and markets
that otherwise wouldn’t have been available for trading, which will produce a com-
pletely new set of trades that needs to be added to the mix.
For example, consider a system without a stop loss and profit target that
enters into its first trade, which then lasts for 10 days. Then nothing happens for
another five days before the system enters into a new trade that lasts for eight days.
Now take the same system, with the same entry rules, but add a stop loss and prof-
it target and do the research again. Now the first trade only lasts for five days
before it is stopped out. Then two days of no trading occur before the system enters198 PART 3 Stops, Filters, and Exits