ADX and DMI — trend strength

Every tool in this part of the book — moving averages, crossovers, MACD — works beautifully in a trend and terribly in a range. You've seen it again and again: the same indicator that keeps you on the right side of a big move whipsaws you to death when price goes sideways. Which raises the most useful question you can ask before deploying any of them: is there actually a trend right now, or is this just chop?

That is the exact question ADX — the Average Directional Index — was built to answer. And here is its defining twist, the thing most beginners get wrong: ADX measures the strength of a trend, not its direction. A high ADX means "there's a strong trend" — it will not tell you whether that trend is up or down. Direction is carried by two companion lines, +DI and −DI, that come as part of the same system. This module untangles strength from direction, so you stop reading a strength gauge as a buy signal.

The tool was created by J. Welles Wilder, who designed it as a three-part system called the Directional Movement Index (DMI). Two of the parts answer direction, one answers strength:

• +DI (the plus directional line) measures how much of the recent movement has been upward. When buyers are dominating and each day's high pushes above the last, +DI rises.
• −DI (the minus directional line) measures how much of the recent movement has been downward. When sellers dominate and each day's low breaks below the last, −DI rises.
• ADX is derived from the gap between +DI and −DI, then smoothed. It ignores which line is on top and asks only: how far apart are they? A big, persistent gap — one side clearly dominating — produces a high ADX. Two lines tangled together — neither side winning — produces a low ADX.

That last point is the key to everything. ADX rises when one side is decisively winning, whether that side is the buyers or the sellers. A stock crashing hard has a big gap (−DI far above +DI) and therefore a high ADX — a strong trend that happens to be down. This is why ADX alone can never be a buy signal: strength and direction are genuinely separate questions, and ADX only answers the first.

Why bother separating them? Because it fixes the biggest failure mode of every other tool in this part. Trend-following tools only pay when a trend actually exists and runs on, and ADX is the closest thing technical analysis has to a "is there a trend?" meter. Read the strength first from ADX; only then let the direction tools speak.

The discipline is to read ADX and the DI lines as answers to two different questions, in order.

Question 1 — how strong? (ADX) The rough, widely-used bands are:

• Below ~20: weak or no trend. The market is ranging. This is a warning to stand down your trend tools — they will whipsaw here.
• Above ~25: a trend is present and worth respecting.
• Above ~40–50: a strong trend. Note that very high readings often come late and can mark exhaustion, because by the time the gap is enormous the move is already mature.

These thresholds are conventions, not laws — 25 is not a magic wall — but they're a fair rule of thumb for "trend versus no trend."

Question 2 — which way? (the DI lines) Only after ADX says a trend exists do you look at the DI lines:

• +DI above −DI: the trend is up.
• −DI above +DI: the trend is down.
• A DI crossover (one line crossing the other) marks a possible change of direction — though, like every crossover, it is late and whipsaws in a weak market.

So the full read is a sentence: "ADX is above 25 (a real trend exists) and +DI is on top (it's an uptrend)." Strip either half away and you have half a picture.

Look carefully at the ADX line (amber) in the figure. It is high on the left (strong uptrend), dips low in the middle (the choppy pause), and rises again on the right — even though the right side is a downtrend. That single line rising in both an up-move and a down-move is the whole lesson: ADX cares only about strength.

Read the composite in Figure 1 as three phases. illustrative

Phase one — the clean uptrend. Price climbs steadily. In the panel, +DI is riding high above −DI (direction: up), and ADX is elevated and rising (strength: real). Both questions answered, both agreeing: a genuine uptrend. This is the environment where your moving averages and crossovers actually earn their keep.

Phase two — the chop. Price stalls and wobbles sideways near the highs. Watch ADX: it falls, sliding down toward and below the 25 line, and the +DI and −DI lines squeeze together. This is ADX doing its most valuable job — telling you the trend has weakened into a range. A reader who trusts it now stops trading trend signals, because they know crossovers here will whipsaw. The falling ADX is the "stand down" order.

Phase three — the downtrend. Price rolls over and starts falling. The −DI line climbs above +DI (direction has flipped to down), and — the crucial part — ADX turns back up as the new down-move gathers strength. If you'd read the rising ADX as bullish, you'd have bought straight into a decline. Read correctly, ADX rising simply confirms "a strong trend is present," and the DI lines tell you it's now a strong trend down.

That three-phase walk is the entire skill: use ADX to decide whether to trust a trend tool at all, and use the DI lines (or the price) to decide which way.

ADX cannot tell you direction. This is not a limitation to work around; it is the definition of the tool. A high ADX is equally at home in a raging bull move and a brutal crash. Reading ADX as bullish because it is high is the single most common error, and it is a category mistake — asking a strength meter to point a compass.

It cannot tell you when a strong trend will end. A very high ADX often appears late, once a move is mature, and it can keep rising right up to the top. A high ADX raises the odds that a trend is real and persistent, but it guarantees nothing about how much longer it lasts — sometimes the strongest reading is the last gasp.

It cannot help in the very market it warns you about. When ADX is low, it is correctly telling you there's no trend — but that also means the DI crossovers are unreliable there, whipsawing just like every other trend tool. ADX is honest about the chop; it just can't turn the chop into a tradable signal. Its value in a range is the warning, not a strategy.

And it cannot tell you what to own. ADX and the DI lines are arithmetic on price highs and lows; they know nothing about the company's earnings, debt or quality. ADX describes the strength of a price move, not the strength of a business. It can tell you whether a trend is worth timing; it can never tell you whether the underlying is worth owning.

1. Reading high ADX as bullish. The headline trap of this whole module. A high ADX means a strong trend in whatever direction the DI lines show — which can be straight down. Always answer "which way?" from the DI lines or price, never from ADX.

2. Trading DI crossovers in a low-ADX chop. When ADX is under ~20, the +DI and −DI lines tangle and cross repeatedly, and each cross is a whipsaw. The low ADX is a signal to ignore those crossovers, not to trade them.

3. Treating 25 as a precise line. The thresholds are rough conventions. An ADX of 23 is not meaningfully different from 26; do not build a hair-trigger rule around a soft number. An ADX threshold tuned to look perfect on past data is usually fitted to that data's noise.

4. Buying the highest ADX reading. The strongest ADX often comes near the end of a move, not the start. Chasing a stock because its ADX is extremely high can be buying exhaustion.