Worst drawdown over fixed time periods DM 65sma-3cc

1 2 3

5 6 12

-2000

-4000

S

-6000

Quot;• -8000

-10000

-12000

-14000

Time (month*)

Figure 6.15 Maximum losses in DM over fixed monthly intervals from February 1988 through June 1995.

198 Equity Curve Analysis

will show what percentages of consecutive months were profitable on a monthly basis, a useful measure of system performance. Figure 6.16 shows these data for the DM test with 65sma-3cc. More than 50 percent of the 90 monthly intervals from February 1988 through June 1995 were profitable. The proportion of winning intervals increases as the period increases. This could be interpreted as the longer you are in a drawdown mode, the more likely you are to come out it.

You should take a good look at the proportion of profitable inter­vals for each market when you combine different markets hoping to in­crease the proportion of profitable intervals. A good measure of your successful diversification processes is upward changes in the proportion of profitable intervals. Here diversification includes multiple markets, multiple trading systems, and different money-management strategies.

You should also look at the standard deviation of monthly equity changes. You can use it to project drawdowns for this system. This idea is explored in the next section. For now, we will test the effect of adding a $1,500 trailing stop to the 65sma-3cc model using actual DM con-