The trick is to test your system with several different fictive risk levels and
decide on one that makes sense to you. As already mentioned, the optimal f, as
defined by Ralph Vince, only looks at the fastest historical equity growth, which
could make the system very risky to trade on future unseen data. Other ways to
select fcould be to go for the fgenerating the highest profit factor, the highest
risk–reward ratio, the smallest average drawdown, the shortest flat time, the most
profitable time periods (such as weeks or months), or the best compromise based
on all this information. Remember never to trade a system on its historical optimal
ffor the fastest equity growth, as that most likely will be too risky.
Figure 26.5 shows the evaluation variables we will learn to calculate with this
spreadsheet. These are far from all the variables needed, but it would be impossi-
ble for me to describe them all. Figures 27.2 and 27.3 show you a more profes-
sional set of evaluation variables, which are derived with the help of the spread-
sheet I use to build the systems for Active Tradermagazine. But for now, let’s stick
to Figure 26.5 and describe each evaluation variable one at a time.
The ending equity in cell H2 is the value of our portfolio at the end of the
testing period, including the value of any open positions, based on the price of the
stock, the number of shares purchased, and the profit or loss of the open position.
The total return in cell H3 is the percentage difference between the starting
balance in cell C2 and the ending equity in cell H2. In this case, we only examine312 PART 4 Money Management
FIGURE 26.4
Input variables to be altered.FIGURE 26.5
Results of varied input variables.