J. Welles Wilder Jr.
ADX + DI: one reading for direction, a separate one for whether the trade is worth taking
Every crossover system has the same weakness: it keeps crossing in markets that are going nowhere. Wilder's answer in 1978 was not a better crossover but a second, independent reading — ADX — that measures how strongly a market is moving without caring which way. The cross supplies the direction; ADX decides whether the direction is worth anything. Neither reading works alone, and the system is the requirement that both agree.
- Style
- Trend following
- Approach
- Mechanical
- Difficulty
- Intermediate
- Horizon
- Position (weeks to months)
- Holding period
- Weeks to months
- Time needed
- 15 minutes after the close
- Markets
- Futures · FX · Commodities · Index ETFs
- Source
- New Concepts in Technical Trading Systems (1978) — J. Welles Wilder Jr.
The rule set
- Compute the 14-period ADX together with its +DI and -DI companion lines
- Go long when +DI crosses above -DI while ADX is above 25
- Exit when -DI crosses back above +DI
- Trail a stop from the highest high since entry rather than a fixed percentage
- No ADX reading above 25 means no trade, whatever the DI lines are doing
What makes it distinctive
- Keeps separate the two questions most indicators merge: which way is this moving, and is it moving enough to trade at all
- The ADX-above-25 gate removes the chop that ruins plain crossover systems
- Wilder built it for commodities, and it still reads futures better than it reads slow equities
When it works
Commodities, FX and index futures in a sustained directional move — the markets Wilder designed the calculation on, where a trend that clears the ADX gate tends to run for weeks.
When it fails
ADX itself lags: by the time it clears 25, a good part of the move is already done. And in a fast reversal both DI lines cross repeatedly while ADX is still elevated, so the filter that protects you from chop does not protect you from a V-turn.
How a decision moves through it
Input
Daily bars
Price only — highs, lows and closes. The calculation needs no volume, which is one reason it travels to FX, where reported volume is not real.
Measure
ADX and the two DI lines, all on 14 periods
One calculation, three outputs. +DI and -DI say which side has been moving price; ADX condenses their disagreement into a single strength reading with no direction of its own.
Decide
Bullish DI cross in a trending market
+DI crosses above -DI and ADX is above 25 — Wilder's threshold for a market that is trending rather than chopping. Both conditions, not either.
Act
Buy
A single full entry at the signal. The system holds until the direction reading flips — there is no target, because a trend's length is not knowable in advance.
Decide
Bearish DI cross ends the trade
-DI crossing back above +DI is the exit. Wilder's original reversed into a short here; this implementation simply steps aside.
Size & protect
Trail off the highest high since entry
A chandelier stop — three ATR(14) below the highest high made while the trade is on. It is the modern rendering of Wilder's own extreme-point rule, which anchored the stop to the extreme reached during the trade.
Direction and strength are different questions, and Wilder refused to merge them
Most indicators hand you one number and let you infer everything from it. Wilder's directional movement system hands you two distinct readings. The +DI and -DI lines measure which side — buyers or sellers — has been moving price over the last 14 bars. ADX, computed from their disagreement, measures how one-sided that contest has been, and deliberately throws the direction away.
| Reading | Question it answers | What it cannot tell you |
|---|---|---|
| +DI vs -DI | Which way has price been moving? | Whether the movement amounts to a trend or just noise |
| ADX | Is this market trending at all? | Which direction the trend points — ADX is direction-blind by design |
A bullish DI cross with ADX at 15 is not a weak buy signal. Under this system it is no signal — the strength reading has vetoed the direction reading.
That veto is what separates this from an ordinary crossover system. Crossovers fail in ranges, where the lines tangle and every cross reverses within days. The ADX gate does not make the crosses better; it declines to act on them until the market has demonstrated that crosses currently mean something.
ADX rises in downtrends too — the most misread fact about it
Because ADX is built from the absolute gap between the two DI lines, it climbs whenever a market moves persistently in either direction. An ADX of 40 during a collapse means exactly what it means during a rally: strong trend. Reading a rising ADX as bullish is the single most common misuse of the indicator.
ADX answers 'is there a trend worth trading?'. The DI lines answer 'which way?'. Neither reading can substitute for the other, which is why the entry condition names both.
The same separation explains why Wilder could use ADX for a second job: comparing markets. A futures trader with twenty charts can rank them by ADX and spend attention only on the few that are actually trending — a use that has nothing to do with entries and is arguably the reading's most durable value.
Who this actually suits
Badly suited
- Wants to catch the turn, not the middle
- Cannot sit flat while ADX stays under 25
- Trades one slow large-cap stock
- Will override the gate 'just this once'
Well suited
- Checks charts once, after the close
- Trades futures, FX or liquid ETFs
- Accepts entering after the trend is visible
- Wants a written reason to not trade
The right column describes someone whose main leak is overtrading quiet markets. The gate exists for exactly that person: for long stretches it will say no to everything, and the system's value is precisely that the no is written down.
Five ways into this system
- The DI cross, the ADX gate, and why the system needs both to fireOne calculation produces three lines. Two of them decide direction, the third decides whether direction currently means anything — and the entry requires agreement.8 min read
- Sizing off a volatility stop when the exit is a moving crossThe stop is three ATR below the highest high since entry — a distance that varies by market and by week. Size is whatever makes that distance cost a fixed fraction of the account.6 min read
- Built on commodities, still best where markets trend hard and oftenThe system was calibrated on 1970s commodity trends and still wants the same diet: liquid markets that spend real time above the ADX gate, read once a day after the close.6 min read
- The gate opens late, and it does not close fast enough for a V-reversalEvery weakness of the system is a price paid for one of its strengths. The lag buys reliability, the gate buys selectivity, and both bills arrive at the same two moments: the start of a trend and the end of one.7 min read
- ADX for beginners: two questions before any trade — which way, and is it worth itThree lines on one panel, but really just two questions: which side is winning, and is the contest one-sided enough to bet on. Everything else is detail.7 min read
The ideas behind it
This system assumes you already know these. Each one is explained from scratch in Investing 101.
Compare with
- MACD Signal-Line CrossoverBuy when the MACD line crosses above its signal line, sell when it crosses back below — one indicator, one rule, on daily bars.
- Ichimoku Cloud TrendFive lines that show trend, momentum and expected support in one picture — with the support levels drawn a month ahead of current price.
These are documented methods described for study. Nothing here is investment advice, a recommendation, or a claim about future returns — every system on this page has losing periods, and the pages say where.
Reading about a system is not having one.
Plutux is where you write your own rules down, test them against real data, and keep the record your memory would otherwise rewrite. Join the waitlist for early access.