I have been trading with Wilder's Averaged Directional Index (ADX), sometimes known as the Directional Movement Indicator (DMI), for more than twenty years and have written and lectured about my findings throughout the world. I would hope that my public fondness for this indicator has contributed to its increasing popularity among knowledgeable technicians. However I continue to see evidence that the ADX is not well understood and is often used incorrectly.

In this brief article I would like to point out a very common misconception about ADX and explain how to correctly interpret the vital information provided by this most valuable technical tool.

As most technicians already know, the ADX is an indicator that measures trendiness. But to do its job most effectively it needs to be interpreted correctly. Unfortunately the inventor, J. Welles Wilder Jr., got our basic understanding of his ingenious indicator off to a bad start by explaining that the level of the ADX is what we need to be concerned with.

If you are acquainted with Wilder's book, New Concepts in Technical Analysis that first introduced the ADX/DMI and I asked you: "Which is more predictive of a trending market; an ADX of 20 or an ADX of 30?"

You would probably not hesitate to respond that the higher level of 30 would obviously be more predictive of trendiness than the lower level.

More important than the ADX level is its direction

However that is not necessarily correct because it is not the level of the ADX but its direction that provides the predictive information we seek. An ADX of 20 that is rising steeply is much more predictive of trendiness than an ADX of 40 that is flat or declining. Traders and technicians need to understand that regardless of the absolute level of ADX, if it is sloping upward the market is increasing its trendiness and if the ADX is declining the market is losing trendiness.

Once this vital distinction between the level of the ADX and the direction of ADX is understood we can carry this directional logic an important step further. An ADX that is rising rapidly is more predictive of trendiness than an ADX that is rising slowly. Let's get into some specifics now. In my experience with ADX using Wilder's default period of 14 bars I have found that any upward movement from the previous bar of 0.25 or greater is significant and tells us that a trend is underway. If the change from the previous bar is greater, say 0.50 or more, then that higher rate of change indicates that an even stronger trend is in place. In fact I have observed that many of the strongest and longest lasting trends will start with a rise in the ADX of 1.00 or more from bar to bar.

The problem with rounding

Unfortunately, a few of the popular software providers follow Wilder's original instructions which called for rounding the ADX to the nearest whole number. Rounding causes serious problems when trying to measure the critical rate of change. For example a significant upward move from 25.05 to 25.35 leaves the rounded ADX unchanged at 25 while an insignificant move from 25.45 to 25.55 jumps the ADX a whole point from 25 to 26.

To use the ADX correctly we must carry the calculations out to at least two decimal places. I have pointed out this problem to various charting services over the years and most of them have understood and modified their formulas.

Fortunately most software providers now carry their calculations out to the required two or three decimal places and there are many places on the web where ADX can be obtained for free.

I primarily rely on the ADX for entry signals and have not found it to be of much value after the trade has been initiated. Exit strategies generally require more precise timing than the slow moving ADX can provide.

The ADX is a very robust indicator and I have had good results using ADX settings as short as 7 bars and as long as 30 bars. The fact that the ADX produces good results regardless of its exact parameter setting is a tribute to its robustness and reliability. Reliable indicators should work well over a wide range of settings. Beware of indicators that require very precise parameter settings to work effectively.

A few more tips that may be helpful to users of ADX:

Profit targets: When the ADX is rising you should be applying trend-following strategies and looking for fairly substantial profits of at least four Average True Ranges or more. When the ADX is flat or declining there is very little trendiness, so trend following strategies are likely to fail. Under these conditions you will need to find another market where the ADX is rising or temporarily abandon trend following strategies and apply short term counter-trend methods with modest profit targets of only 1.5 ATRs or less.

Major trend reversals: When the ADX is above 35 and it is higher than either the Plus DI or the Minus DI then extreme caution is in order because many important trend reversals occur at these extremes. When the ADX is above both the Plus DI and the Minus DI try using a very sensitive indicator (such as the Parabolic) to quickly signal a possible change in the primary direction. This technique can identify many major "V" shaped tops and bottoms particularly in stock index markets.

I hope that this brief article will help you get better results with my favorite technical indicator. I don't know how anyone can trade without looking at ADX beforehand. It provides critical information that all traders can use.

A PowerPoint presentation about ADX is available at no charge if you send an email request to me at clebeau2@cableone.net.

Charles LeBeau is a highly respected futures industry professional with over 40 years of trading experience. He is presently Director of Trading for Tan LeBeau LLC, a hedge fund management company. Chuck is co-author of a best-selling book on futures trading titled Computer Analysis of the Futures Market (McGraw Hill, 1992). This highly acclaimed book has been translated into six languages and is now considered a classic text on how to use technical indicators.