Trend & directionDirectional Movement Index · DMI
Wilder's +DI and −DI lines that measure upward versus downward directional pressure.
Works best in trending marketsEngine-computed on a fixed sample series
What it is
The Directional Movement Index, or DMI, is Welles Wilder's system for measuring whether buyers or sellers are winning the tug-of-war, and by how much. It consists of two lines that both travel between 0 and 100: the Plus Directional Indicator, written +DI, which measures upward pressure, and the Minus Directional Indicator, written −DI, which measures downward pressure. When up-moves dominate, +DI rises above −DI; when down-moves dominate, −DI rises above +DI. From the gap between these two lines Wilder also derived the ADX, which measures how strong the trend is regardless of direction, so the DMI and ADX are usually shown together. For a beginner, the DMI answers a simple but vital question: which side is in control right now, and is that control getting stronger or weaker? It separates the question of direction from the question of trend strength.
How it is calculated
For each bar, Wilder compares the current high and low to the previous bar's. The up-move is the current high minus the previous high, and the down-move is the previous low minus the current low. Plus directional movement is recorded only when the up-move is positive and larger than the down-move; minus directional movement only when the down-move is positive and larger than the up-move; otherwise that side records zero. These raw directional movements, along with the True Range, are then smoothed over 14 periods using Wilder's own smoothing method. Plus DI is 100 times the smoothed plus directional movement divided by the smoothed True Range, and −DI is the same for the down side. Finally, the DX is 100 times the absolute difference between +DI and −DI divided by their sum, and the ADX is a smoothed average of DX, tying the whole family together.
Reading it, step by step
Begin with which line is on top: +DI above −DI means buyers hold the upper hand, and −DI above +DI means sellers do. The width of the gap matters — a wide separation shows lopsided, one-directional pressure, while lines coiled close together show a balanced, indecisive market. The core event is a crossover: when +DI crosses up through −DI, directional control has shifted to the bulls, and the mirror shift to the bears when −DI crosses up through +DI. On their own the crossovers say nothing about whether a trend is strong enough to trade, which is where the ADX comes in: a rising ADX above 25 confirms that the dominant DI line reflects a real trend rather than range noise. So you read direction from the DI lines and conviction from the ADX, and you want both pointing the same way.
Best timeframes and settings
Wilder's default is 14 periods for both the DI smoothing and the ADX, and it remains the standard across the board. The DMI works on every timeframe, but because it is built to detect trends it is most reliable on higher timeframes such as daily and 4-hour, where trends are cleaner, and noisier on very fast intraday charts. Shortening the period makes the DI lines more responsive and the crossovers more frequent, which suits active traders willing to accept more false signals, while lengthening it produces slower, steadier lines with fewer but higher-quality crosses. Many trend traders keep 14 but pair it with a strict ADX threshold to filter the extra noise a shorter setting would create. The trade-off is the familiar one: speed versus reliability, with Wilder's smoothing already adding some lag that a shorter period only partly offsets.
When and where to use it
The DMI is fundamentally a trending-market tool and is designed to keep you out of directionless chop, so its natural regime is trending. It is used across equities, futures, FX, and commodities, and it is especially valuable on instruments prone to sustained directional runs. In a ranging market the DI lines cross back and forth repeatedly, producing a stream of false signals, which is precisely why the ADX filter exists — it tells you when the range has given way to a genuine trend worth acting on. Avoid trading DI crossovers in isolation during quiet consolidation, and be cautious right after a sharp reversal, since Wilder's smoothing means the lines lag the turn. Used properly, the DMI is less a standalone signal and more a two-part regime and direction gauge that other entries are timed around.
Strategies that use it
The first and most classic strategy is the filtered crossover: go long when +DI crosses above −DI while the ADX is above 25 and rising, and go short when −DI crosses above +DI under the same ADX condition, with stops beyond the recent swing. The second is the extreme-point rule Wilder himself suggested: on the crossover, note the high or low of the signal bar, and only act once price trades beyond that extreme, which weeds out crosses that never follow through. The third is a trend-strength overlay for an existing position: hold as long as your directional DI line stays on top and the ADX holds above 25, and consider trimming when the DI lines converge or the ADX rolls over, signalling the trend's grip is loosening. Each pairs the direction from the DI lines with the conviction from the ADX so that you only commit when both agree.
Combining it with other indicators
The DMI's most important companion is its own offspring, the ADX, which converts the DI relationship into a clean trend-strength reading and is almost always shown with it. Beyond that, it pairs well with other trend-onset tools such as Aroon or the Vortex Indicator, which time the birth of a trend and can corroborate a DI crossover. A moving average gives a visual trend backdrop and a dynamic stop reference for positions taken on DI signals. Because the DMI says nothing about location, support and resistance or pivot levels help you avoid buying a bullish crossover straight into overhead resistance. Volume tools can add conviction, since a DI crossover accompanied by expanding volume is more likely to mark a durable shift in control than one on fading participation.
Where it fails
The DMI's chronic weakness is the ranging market, where the DI lines whipsaw across each other and generate false crossover after false crossover — the reason the ADX filter is not optional but essential. Wilder's smoothing also makes all three lines lag, so at fast reversals the crossover arrives well after the turn, and by the time +DI overtakes −DI a chunk of the move may be gone. Beginners often trade the raw crossover with no ADX confirmation and get chopped up, or they misread a high ADX as a buy signal when ADX only measures strength, not direction. The fixes are to require an ADX above 25 and rising, to use the extreme-point rule so price must confirm the cross, and to avoid the tool during obvious consolidation. Remember too that ADX peaking and turning down does not mean reverse — it means the current trend's momentum is fading, which is a different message.
A worked example
Consider a stock that has been consolidating and then begins to break higher. On the breakout bar, +DI climbs to 30 while −DI slips to 18, so +DI has crossed above −DI and buyers now dominate. You check the ADX: the DX for this bar is 100 times the absolute difference of 12 divided by the sum of 48, which is 25, and the ADX, previously near 15, is now rising through 22 toward 25. Applying the filtered-crossover rule, you wait one more bar for the ADX to push above 25 and for price to trade above the crossover bar's high, then enter long with a stop under the consolidation low. Over the next two weeks +DI stays on top and the ADX climbs into the 30s, confirming a healthy trend, and you hold. When the ADX finally peaks near 40 and starts to roll over while the DI lines begin to converge, you tighten your stop, reading it as the trend's grip loosening rather than an outright reversal.