in consecutive order. The winning or losing streak ends when a trade results in a gain or loss in the opposite direction of the series. Calculations based on consecutive series data can produce two very powerful evaluation tools:

1. Average gain (or loss) of a winning (or losing) streak.

2. Average gain or loss for the trade that ends the streak.

The first calculation takes an average net gain of all winning streaks or an average net loss of all losing streaks. In general, the more winning (or losing) trades there are within a streak, the larger the magnitude of the streaks average performance. This measurement can determine a systems profitability during periods of extended strength or weakness.

The second calculation centers on the average return for the trade that ends the streak. This figure can determine how abruptly a system reverses its course of action after a consecutive series of wins or losses. We like to end a streak with a small loss or large gain. Therefore, a small losing reversal is preferred for ending a winning streak, while a large winning reversal is preferred for ending a losing streak.

Table 9.11 shows the consecutive winning series for a real trading system used by the author. All numbers in this table are quoted in S&P 500 Index points. This system generated a number of consecutive winning trades, and as expected, the longer the streak, the larger the average net gain. In addition, notice the inverse relationship between the average streak gain and the loss size of the trade that ends the streak. As the average streaks net gain increased, the loss size of the ending trade decreased. An explanation for this is that a system that trades exceptionally well has most likely found its preferred trading zone, and it would be unusual for a system that is trading "red hot" to immediately turn "ice cold." Consequently, the system ends its winning streak with a relatively small loss. This is the mark of a well-designed system. Knowing a systems historical tendency during and after a consecutive series of trades will help you build confidence in your trading system.

Table 9.11 Consecutive winning series

To put this into a real world trading perspective, consider the following true

story.

An investor, after months of comparing trading services, finally decided to subscribe to a trading service with an impressive track record. The service was in the midst of a small winning streak, making the investment decision all the easier. As time passed, the service continued to extend its winning streak. The investor, thrilled with these results, felt reassured in his decision to subscribe to the service. After a few more winning trades, the investor began to feel a little uneasy, knowing that all winning streaks must eventually end. As the winning streak persisted, the investor completely centered his attention on the next losing position. The more the service won, the more uncomfortable the investor became. Eventually the investor, consumed with fear of a major loss, decided to pull out of the service. When asked why the investor had left the service, the response was "Why wait for a major loss?"

The only thing that changed in this case was the investors perception. Trading logic did not dictate that the next losing trade had to be a major loss. If the system had been evaluated using the consecutive series data in the first place, then the investor would have been better prepared to trade using this system. Fear of the unknown always weighs heavier in our mind than that which is known. Preparing for the unknown begins with a thorough system evaluation.

Time Analysis""4

-ftlissection centers its evaluation strictly from the standpoint of time in or out of the marketrThe longer a position is exposed to the market, the more risk it assumes. This form of analysis can be used on the entire system or on individual trades (see Box 9).

BOX 9

TIME ANALYSIS

Total time in the market (periods): The time in which the system is actually trading.

Percentage of time in the market: The total time in the trades divided by the total time for the test period (in a percentage). If a system trades more then 80 percent of the time, make sure its reward/risk ratios are in line with other comparable systems. Time in the market is another measure of risk.

Longest flat period: This figure notes the longest period in which the system did not trade. Consider this to be a patience factor. Note that the longer the flat period the more patience a trader must possess to follow the system.

Table 9.12 shows a simple side-by-side comparison of two trading systems. We can evaluate them based on percent-in-the-market.

With all things equal, and with time as the only variable, System A made its profits in 40 percent less time in the market than System B. This smaller exposure to the market translates into less risk and ultimately, a more efficient trading system. If you were to account for the interest earned on the idle cash position, then System A becomes the clear winner. By factoring time into the evaluation process, traders will improve their overall understanding of a system and its true performance.

Time can also be used in the evaluation of a systems individual trades. These calculations serve as historical reference points, especially useful for option traders where time to expiration is a critical factor in evaluating price.

Table 9.13 shows the results of a trading system that remains in an average trade for approximately 10 days and is then idle for about 48 days. This information should be used to compare open (in the market) positions to the systems historical tendencies. Trading action may be required, should the duration of the open position strongly deviate from the average. This type of review increases trading confidence in the underlying position.

These calculations can also reveal potential system design flaws. Although the intended trading strategy may be correct, your systems design may not match the time-based trading results. For example, the numbers in Table 9.13 reflect a trading strategy that trades in short-term spurts. If the system had been originally designed as a longer-term, trend-following system, then the design would definitely have a problem.

Quite often the difference between a professionally evaluated system and a quick review performed by a novice investor is the use of time. If traders are going to profit in the real world, they must factor time into their evaluation.

Table 9.13 Time analysis