ADDING STOPS

To make Intermarket Three a more usable trading system, we should try to reduce drawdown. We can do this by experimenting with various stops, including profit targets, breakeven stops, money management stops, and trailing stops. For this example, we will use a trailing stop, to try to keep drawdown low while locking in profits from trades that are going well.

Figure 14.17 is an optimization table for every trailing stop from $2,000 to $20,000 at intervals of $1,000. All the lower trailing stops, those under $4,000, keep drawdown under $28,000. Excessive drawdown is the major factor that scares traders into abandoning a trading system. For that reason, many traders would have peace of mind in trading a system such as this with a small stop. However, the smaller stops greatly affect system performance figures.

The larger trailing stops, say from $15,000 up, maintain system performance, but at the expense of drawdown. The larger stops affect only a few trades over time, and results stay consistent, no matter what stop is used. Trading with such a large stop may provide some protection in the event of a market catastrophe, and for many people, it adds a measure of psychological security.

Thus, the addition of yet another optimized input, the trailing stop, increases complexity, but still does not greatly deteriorate the system. The decision whether to

Figure 14.17 Optimization: Intermarket three.

Optimization: Intermarket Three

Lenl = 3 Len2 = 9. Optimize Trailing Stop $2,000 - $20,000. Subtract Commissions=$25, Slippage=$50

Source: Printed using TradeStation by Omega Research Version 4.02.15-July 09 1996.

utilize a stop, and, if so, what type and at what value, is a personal one. In this example, the trade-offs can be easily seen. The choice depends on the traders evaluation of the factors of maximizing profitability and minimizing risk.

Evaluating the System

For the purpose of further analysis, we will choose the values of 3 and 9 for Lenl and Len2 respectively, and $17,000 as the trailing stop. We will deduct $75 per trade for commissions and slippage. For the evaluation, and the graphs that follow, we will use Performance Summary Plus (PSP), by RINA Systems, Inc. The results are shown in Figure 14.18.

The performance factors of net profit and percent profitable allow for a ready comparison of systems. In this case, profit statistics of $408,675, with 70.77 percent profitable trades, are favorable. However, the ratio of the average win divided by average loss is only 1.18, which is low.

The profit factor, which is the gross profit divided by the gross loss, shows how much money was made for every dollar lost. A good system has a profit factor of 3 or more. Our example, with a profit factor of 2.85, approaches that figure.

Return on maximum drawdown, otherwise known as return on account, is calculated as the net profit divided by maximum drawdown. A good system has this number at 150 percent per year. Since our test was for 14.6 years, the 2,153.67 percent return on account averages about 148 percent per year, which is not bad.

The average trade is calculated as the net profit divided by the number of trades. With the stop we have chosen, the average trade is an excellent $2,095.77.

To get a better idea of how this system performs, it is necessary to look at how its trades deviate from the average. This can be seen in a graph of the total trades, as shown in Figure 14.19-

It stands to reason that the closer the trades are to this profitable average, the more consistent and better the system. If we add and subtract one standard deviation from the average trade, we are able to see the range within which 64 percent of the trades reside. Here, the range is from $8,462.95 to -$4,271.42.

The coefficient of variation is calculated by expressing one standard deviation as a percentage of the average trade. The smaller the percentage, the more stable and consistent is the system. Our example system showed a coefficient of variation of303.81 percent. This is quite high as a good system would have this figure at 200 percent or less.

A good analysis of a system focuses on drawdown. The maximum drawdown is the largest intraday drawdown experienced by the system. The $37,600 drawdown of this system is somewhat high, even for a contract such as the S&P, which has a value in excess of $350,000.

The maximum drawdown does not tell us the whole picture about drawdown, since it may have only occurred once in the many years of history. The average drawdown is a very important figure. This is the average maximum loss potential of all trades. It gives you an idea of what can normally be expected during trading. Here,

Intermarket Three S&P 500 Index - CME-Daily 04/21/82-12/19/96 System Analysis

Net Profit Gross Profit Gross Loss

Percent profitable Ratio avg. win/loss

$408,675.00 $629,975.00 ($221,300.00)

70.77% 1.18

Open Position

Profit factor

$5,225.22

2.85

Return on Max. Drawdown

2153.67%

Total Trade Analysis

Number of total trades 195

Average trade $2,095.77

1 Std. Deviation (STDEV) $6,367.19

Total stopped trades Avg. trade ± 1 STDEV Coefficient of variation

$8,462.95 / ($4,271.42) 303.81%

Drawdown

Maximum Drawdown

Average Drawdown

1 Std. Deviation (STDEV)

Reward/Risk Ratios

Largest Loss Ratio

($37,600.00) ($3,491.55) $4,106.98

24.28

Avg. trade ± 1 STDEV Coefficient of variation

$615.43 / ($7,598.53) 117.63%

Outlier Trades

Positive outliers Negative outliers Total outlier

Total Trades 1 1 2

Profit/Loss $47,025.00 ($17,050.00) $29,975.00

Copyright© 1996, RINA Systems, Inc. Source: RINA Systems, Inc. Used by permission.

Version 2.0