Trend Indicators and Price Components

Trend Indicators and Price Components
Since trading with the trend is essential to exploiting the trend component of market price action, successful systems and approaches employ some method of identifying trend. A popular idea is a separate trend filter. This is an algorithm that preliminarily looks at recent price action and determines whether the trend is up, down or neutral.
It is not essential that every approach have a separate trend filter. Some methods, such as moving averages, incorporate a trend indicator into the entry technique. Others (not recommended by me) try to predict an imminent change in trend and are therefore entering when the trend is against them at the time.
Many traders attempt to complicate the problem of trend identification. They invent all kinds of fancy mathematical equations and methods of massaging past price action to more precisely determine whether the trend is up or down. I have long argued that this is pointless. Like all things in trading, simplicity is the best route.
Of course, trend is only relevant in conjunction with a particular time frame. But once you identify the time frame, there is nothing fancy about the concept of trend. The price is either moving up or it is moving down during that time period. You can look at a chart and in most cases quickly determine whether price is moving up or down in your time frame. Why should it be any more complicated than that? Perhaps in cases of price congestion, it is slightly more difficult to say with certainty what the trend is by quickly looking at a chart. However, this is no reason to unnecessarily complicate the trend determination process.
In creating simple trend indicators, there is one intriguing issue. What price components should you use to construct the indicator? There are four choices: the open, high, low and close. You could use them separately or in combination. Can research help determine whether one or more price components does a better job of indicating trend than the others? Or are they allequivalent in trend-identifying ability?
I first performed this exercise in my 1989 book, The Dow Jones-Irwin Guide to Trading Systems. The testing period there was the five-year period ending June 30, 1987. Some time ago, I thought it would be instructive to bring that research up to date.
The procedure is to create four simple trend filter indicators. Each will use different price components. To test their effectiveness, I created for each a simple trading system. When the trend filter is pointing up, the system goes long. It holds the position until the trend filter reverses to down. At that point the system closes out the long position and establishes a short position. The system continues to trade that way. It is always in the market and always positioned in the direction of the trend filter. In order to make the system test as realistic as possible, I deducted $100 from each trade to cover slippage and commission. Theoretically, the filter that is most effective in determining trend should yield the highest profits as a trading system.
I tested each filter in ten diversified markets over a ten- year period. The ten markets were the same I used in my book: Cattle (LC), Sugar (S), Soybeans (SB), Swiss Francs (SF), Comex Gold (GC), T-Bonds (US), Japanese Yen (JY), Heating Oil (HO), Eurodollars (ED) and the S&P 500 (SP).
I used a similar approach to measure historical market trendiness in my book, Trendiness in the Futures Markets. Because the time period is so crucial in measuring trend and because I did not want to make a subjective choice, for purposes of that book I tested every time period between 5 and 85 days. I update the research every year, so the book is constantly up to date.
For this limited project, I chose to use 34 days as a constant time period. From my research I have found this area to be an excellent overall trend-measuring period for all markets. It did yield profitable results for the average of all ten markets, even after deducting $100 per trade for slippage and commission.
Momentum
Momentum is perhaps the simplest form of trend indicator, but it is theoretically sound and ought to be psychologically pleasing. It looks at the direction of the closing price. It uses only the closing price in its calculation.
To calculate Momentum, first determine the time period over which you will measure trend. Then compare today's close with the close at the beginning of your chosen time period. In this case you would compare today's close with the close 34 days ago. If today's close is lower, the trend is down. If today's close is higher, the trend is up. If today's close is the same, the trend is the same as it was the previous day.
This comports with common sense. If the 34-day trend is up in the market, wouldn't you expect today's close to be higher than that of 34 days ago? Isn't that what trend means? This is a simple indicator to use while eyeballing charts. Pick your time frame and become an instant trend expert.
Using the Momentum-based trading system, half the markets were profitable, and the overall average trade was $42.
Directional Movement
This method of identifying trend was created by Welles Wilder and described in his 1978 book,New Concepts in Technical Trading Systems. Unlike Momentum, which uses only the closing price, Directional Movement uses only the high and the low price. It is fairly complicated mathematically, and therefore, you cannot eyeball it on charts.
Directional movement compares the portion of today's price bar that moves beyond yesterday's. If today's high and low are greater than yesterday's, the portion of the today's bar that extends above yesterday's is the up directional movement. If today's high and low are less than yesterday's, the portion of the today's bar that extends below yesterday's is the down directional movement. For an inside day (where today's high is lower and today's low is higher than yesterday's), there is no directional movement. For an outside day (where today's high is higher and today's low is lower than yesterday's), the day's directional movement is the larger of the up or down directional movement as defined above. Up directional movement is positive; down directional movement is negative.
To calculate the indicator, you find the total directional movement over the time period you selected for measurement. If it is positive, the trend is up. If negative, the trend is down. Just as with Momentum, the concept makes intuitive sense. If price is trending up, you would expect the preponderance of daily bars to have higher highs than the day before, and vice versa.
Using the Directional Movement-based trading system, half the markets were profitable, and the overall average trade was $54.
Open/Close Indicator
This is an alternate trend indicator I created about nine years ago. It looks at the relationship between the opening price and the closing price on the same day. Thus, it has the advantage for our purposes of incorporating the opening price in the trend calculation. It is the only trend indicator I know of that uses the opening price.
In an uptrending market, you will find that the close is most often higher than the open. In a downtrending market, the close tends to fall below the open. By comparing the sum of the closes with the sum of the opens over the trend time period you have selected, you get a good indication of the trend based on the relationship of the opens and closes.
Using the Open/Close Indicator-based trading system, half the markets were profitable and one broke even. The overall average trade was $31.
Combination Indicator
The last test attempts to use all four available data points (open, high, low and close) to determine trend. Its indicator is the combination of Momentum, Directional Movement and Open/Close. When all three indicators (measured over 34 days) point up at once, the Combination trend is up. It stays up until all three indicators point down at once. At that point, the trend changes to down.
Using the Combination Indicator-based trading system, again, half the markets were profitable, but the profits were considerably larger. The overall average trade was $185.
The results of these tests indicate that the best trend indicator was the Combination. Next best was Directional Movement. Momentum was third best. Open/Close was the least effective.
That such simple trend-following trading systems were profitable on average over ten diversified markets for a ten-year period shows that there is an exploitable trend component in market price action. By using a more sophisticated trading system and trading it in a more carefully chosen selection of markets, you can improve results substantially. That is precisely the way a diversified, trend-following program, such as I use in my personal trading, works.