Summary. В истории американской фантастики немало случаев, когда один и тот же рассказ или роман получал две самые высокие награды

Summary

In this chapter we examined seven major systems, each exploring a dif­ferent philosophy. The 65sma-3cc system is a simple trend-following system that offers solid performance. We examined this system in great detail. You should perform such a detailed analysis of your systems. We used many important ideas such as maximum favorable and adverse ex­cursions, frequency distribution of trades, and the effect of a volatility-

Summary 145

based initial stop. You should endeavor to understand these ideas and use them often. In summary, the 65sma-3cc system was a robust and profit­able system that makes money in trending periods. About 4 percent of the trades were mega-trades, and you should pick your exit strategies carefully so as not to cut off such trades prematurely.

The channel breakout-pull back (CB-PB) system was our first pat­tern-based system for long trades. The system gives reliable long entries;

it was profitable about 55 percent of the time without stops or exits over many markets. Our tests with different exit strategies found that the CB-PB system provides low-risk entries and can be easily adapted to many trading styles. CB-PB does well in short- to intermediate-term systems on active markets. In summary, we showed how we can develop very dif­ferent systems of the same entry signal.

The ADX burst trend-seeking system uses a filter based on the strength of the trend. It successfully provides entries into strong trends. The system has a large neutral zone so it is in the market only 35 to 45 percent of the time.

We next looked at the trend-antitrend (T-AT) system, which auto­matically switches between antitrend and trend-following modes. This system often picks tops and bottoms with surprising accuracy. It is essen­tial to use good risk control with a system such as this because the vola­tility can trap you on the wrong side of the market.

The gold-bond system allows us to use intermarket analysis on positively or negatively correlated markets. This system tries to capture weak intermarket relationships. It can be used with or without a trend filter. We found several interesting relationships between the U.S. bond market and the gold, copper, soybeans, and currency markets. We extended the gold bond system to look at correlated markets such as deutsche mark-Japanese yen and corn-live hogs. At a minimum, you can use this system to evaluate a favorable or unfavorable background for your trading.

The bottom-fishing pattern started off as a market-specific system for trading the S&P-500 futures. We found that the pattern can be gen­eralized to other markets, as well as to stocks. The pattern also works on daily, weekly, or monthly data. Thus, we captured a fundamental feature of price behavior in markets with this pattern.

The extraordinary opportunity system is primarily a flag to vary bet size during trading. It is an effective trend-following system in its own right. It assumes that during major moves, a significant portion of the move will be outside 1-percent or 3-percent trading bands drawn around a 50-day SMA.

146 Developing New Trading Systems

You will notice that we did not follow the usual practice of dividing the test period into two or more subintervals, "perfecting" our system on one interval, and then testing it on the remaining data. This proce­dure exposes the system to new data, and its performance on such "out-of-sample" data is used to determine how well it might work in the fu­ture. If you do test many combinations of parameters, try to find the most "stable" combination. Stability means that the system is profitable across most of the test data, and results do not change suddenly for small variations in test conditions. A detailed discussion of this approach is found in Babcock, Chande and Kroll, Pardo, and Schwager among oth­ers (see bibliography).

Instead of relying on testing alone, here we have focused on the ideas driving the trading strategy. We tested these systems on all avail­able data over 20 years without optimization. Thus, we stressed our sys­tems to check if they were based on valid market dynamics that can be tapped for future trading profits. These systems are examples of the art of building new strategies around a particular market intuition. In the next chapter we examine how you can modify well-known, proven ideas to create useful variations for your collection of trading systems.

Chapter

Developing Trading System Variations

Even change is never the same.