The Sample Analysis

The Proof Is in the Back-Testing

Short-term system traders make their money, leave the game, and wait to execute the next setup; they don’t have to worry about being an economist. They get in when the signals indicate and get out when the signals dictate. It has been stated many times and proven by history that traders who consistently make money are those who follow a program or a set of predefined rules. These rules have been tested, and they correspond in a variety of time periods and with diversified or non-correlated markets.

You may have made observations on changes in the market, such as when the stock index futures are trending up by 10:30 a.m. (CT) on a Friday, the market has a tendency to trade higher through the end of the day. With that said, once you have a trading idea about the way a market moves, through your own observation, you can apply it to computer code and have a computer generate an alert to inform you when a signal is generated.

Through the proficiency of computers, we can take historical data and back-test the results to see how reliable and accurate the theories are. Not all systems work well. Some are good in trending market conditions, and others will work well in non-trending market environments. Therefore, it is important to back-test a methodology for various time periods and in various noncorrelated markets to see if the principles are sound and stand the test of time. So far, with what we have gone over, you can develop your own rules based on pivot point and candlestick patterns.

One such example is if we assign a positive (+) value to a candle with a higher close than open, if you have a sequence of three positive values after a succession of, say, five negative (-) values, and if the low of a given candle is within two ticks, or 5 percent of, the average daily range on a 10- day average, then a buy signal is triggered. Another way of stating this is:

• For a positive assigned candle:

             Close (C) is greater than open (O) ( C > O)

• For a negative assigned candle:

            Close is less than open (C < O)

• For a doji candle, with a neutral (=) assigned value:

            Close is equal to open (C = O)

• You can program your own system to generate an exit strategy based on an average holding period of, say, seven or eight periods or until the market forms a lower close than open (C < O) or a negative assigned value and if that time period closes below the prior time period’s low.

Another methodology or system that is a simpler program would be to combine a three-period pivot point moving average with a five-period pivot point moving average. After a series of lower closing lows, once the three- period moving average crosses above the five-period moving average and if the market was within five points above or below a pivot point first support level (S-1) target number, then a buy signal would be generated. For added optimization, you could add a filter that if the stochastics indicator was below the 20 percent line, once %K and %D crossed and closed back above the 20 percent level, a buy signal would trigger.

Using the value of the pivot point resistance target of R-1 while combining the three-period pivot point moving average with the five-period pivot point moving average, once the three-period crossed and closed below the five-period pivot point moving average, within five points either above or below the R-1, a sell signal would be generated. For added fine tuning, you could add another element, such as a stochastics reading, so if both moving averages were above the 80 percent level and once %K and %D both closed back below the 80 percent level, a sell signal would be triggered.

The hard part in coding is comprehending or understanding the software language and then thinking like a computer. You can’t just state “sell” when stochastics gives a sell signal. The software or computer program does not know what your interpretation of a sell signal is.

If you recall, in Chapter 1, I suggested that traders need to ask more questions. That is what a system trader does. In order to develop a trading plan or a mechanical system, you need to search for answers to question such as:

• Will this be for day or position trading?
• Will volume studies be used and, if so, how?
• What will be the average holding period?
• What methods do we want to use and what parameter will we choose?

A trader needs to use various tools to get the job done, so to speak. That is why I have several trading platforms and quote services. Genesis Software allows me to develop my own “black box” of algorithms and indicators and to back-test the program. In order to develop a good system, you need to decide if the trading concepts are valid. Once you have come up with your set of rules or criteria, then you can start to optimize the program. You have heard of the “Keep It Simple” rule? Leonardo da Vinci is credited with saying “Simplicity is the ultimate sophistication.” If Leonardo liked things simple, well then, it is good enough for me, too!

A trading system should be designed from simple yet effective rules that blend key concepts of momentum changes and trending conditions. A well-defined trading strategy can be developed to generate buy and sell signals in most markets in various time periods. Believe it or not, designing a trading system is easy to do once you have an idea of what the call to action, or the trigger, will be to enter a position. In this book, we have gone over several good techniques with which you can start, such as when indicators correspond near pivot points like the stochastics 80 percent/20 percent close line technique or the moving average convergence/divergence (MACD) zero-line cross signal. You can develop a system based on these two indicators and then determine how many time periods, or the length of time, a position should be in the market. Now all you need to determine is how many holding periods (such as 4 or 8 periods) are needed before an exit strategy is triggered. You can develop a trigger on moving average crossover to initiate a signal, just like the pivot point moving average we went over in Chapter 6 or a simple moving average using 5-period versus 10-period parameters. You will want to test the system over a lengthy time frame and with various non-correlated markets. If there is any validity to the rules or methods, then you should see positive performance across the board.

Not all systems that show results of extremely high profit levels are best. A system also depends on drawdowns, or periods of equity loss. Also, it is important to have software that will show you a trading system’s weakness or will help point out specific time periods or time frames in which you should not trade. Being able to closely follow your drawdown patterns is a huge benefit to the computer-friendly active systems trader, finding answers to such questions as: What is the recovery period to gain back losses? What is the depth or magnitude of the losses? These are very important questions because if a system has great returns but absurd equity draw downs for long periods, then it is most likely a poor system for an average investor to trade. When you are designing your system, ask yourself what you want this system to do. The obvious answer is to make money; but you also want to know how long and how deep the drawdown periods are or whether there are consistently more winners than losers. As a trader you need to evaluate new techniques, especially ones that increase your profitability while reducing your risk and exposure in the market. Systems need to be developed with the idea of triggering a turning point in the market earlier than other indicators and other traders do. Considering that most indicators are lagging, you need a signal that identifies turning points as they occur, and you need to use indicators to confirm that the position is valid. Pivot point analysis does just that, especially when combining support and resistance levels and by implementing a moving average approach

One of the best features or benefits to using mechanical trading systems is that they alert a trader to initiate a trade or to act on a signal as it occurs, rather than on a hunch or on the two main destructive emotional elements, fear and greed. Based on a system that is tested, you have statistical evidence to validate the entry; and that should help eliminate trading on a hunch, a rumor, or a feeling, which is what drives most trading decisions from inexperienced traders. As this book has demonstrated, it is not my intention to portray the ability to buy the exact low or sell the exact high. In most cases, I am looking for a definitive trend or price reversal to occur; and then I act on that signal and carry a position until the market demonstrates a loss in momentum, the time to exit the trade. The keys to winning are the ability to practice patience and wait for the setup to develop and the discipline to act on the trigger. Mostly, it is the ability to exit a losing trade as the signals dictate, not to hang onto the loser. So when using a system, you will know and understand that the mechanics are never 100 percent right and that you will have a certain number of trades that do not work out. Once you come to grips with that fact, it will be easier to embrace your losses rather than become emotionally wrangled when they occur. Hopefully, you can learn from a trader with long-term positive experiences. Good advice will get you ahead more quickly than if you have to learn everything on your own, and it will help you avoid costly mistakes. Take my advice: Learn a system, back-test it to see the strengths and weakness of the method, and then trade on these testable signals.