imum holding period. This makes perfect sense in regard to the memory of the
market. We’ll take a close look at this second volatility method shortly, but first
we need to learn how to analyze the results.Surface Charts
When trying to figure out where to place the stops in the system I’m working with,
I like to use a surface chart, which provides an excellent method for getting a feel
for how the output variable (for example, the average profit per trade) reacts to two
different input variables (for example, the stop-loss and profit-target distances)
and how the inputs are interacting with each other. In the upcoming analysis of our
systems, we look at the following output variables and how they change with the
addition and changes to the various stops we already have defined:
The average profit per trade
The standard deviation of all trades
The average percentage of winning tradesOf course, one can use other evaluation parameters and output variables as
well, such as the final net profit, drawdown, profit factor (gross profit divided by
gross loss), average trade lengths, and the average number of trades. But at some
point it becomes information overload, and we won’t be able to digest it all—not to
mention the redundancy of the output variables basically telling us the same thing.
Figures 19.1 and 19.2 show what a couple of typical surface charts look like.
In Figure 19.1, the output variable is the average profit per trade. In Figure 19.2,
the output variable is the risk–reward ratio (the average profit per trade, divided by
the standard deviation of all trades). The different values of the output variable are
represented by the different colors. The legend to the right of the chart tells us
which specific value each color represents. The values for the input variables are
represented by the numbers on the horizontal and vertical axes. The output vari-
ables for both figures are the maximum allowed trade length on the horizontal
axis, and the stop loss on the vertical axis. The numbers on the axis tell us that this
particular system has been tested for maximum trade lengths varying from one to
ten days, in steps of one day, and for stop-loss distances away from the entry price
varying from 0.2 percent to 2 percent, in increments of 0.2 percent.
The grid inside the charts helps us find the output value that corresponds to
a specific combination of the input variables. For example, in Figure 19.1, the lit-
tle square surrounds the intersection on the grid that represents a system with a
four-day maximum allowed trade length and a 1 percent stop loss. Matching the
color that covers this area of the chart with the colors in the legend, we see that this
version of the system produced an average profit per trade somewhere between 0.4
and 0.5 percent. To be prudent, we will go with the lower value and say that the
average profit per trade is 0.4 percent.212 PART 3 Stops, Filters, and Exits