Direct answer
A worked example of “ADX Trend” is a step-by-step numerical scenario that uses the Average Directional Index (ADX) framework to illustrate how trend strength is computed from price movements. ADX is commonly used alongside Directional Indicators (DI+ and DI−). In this article, the goal is not to predict market moves, but to show the mechanics clearly with explicit assumptions.
Mechanism or definition
ADX (Average Directional Index) is an indicator that aims to quantify trend strength. It is typically built from:
- Directional Movement: how much price movement is in the up direction vs. the down direction.
- Directional Indicators (DI+ and DI−): scaled versions of the directional movement.
- DX: a measure of how different DI+ and DI− are.
- ADX: a smoothed average of DX over a chosen lookback period (often called “length”).
Material assumption for any worked example: different charting platforms may implement smoothing and initial values slightly differently (for example, how the first smoothed averages are seeded). A worked example can therefore teach the conceptual workflow, but exact numeric outputs can differ by implementation.
Evidence or example
Scenario setup (all assumptions stated)
To keep this fully verifiable without live data, assume we have a simplified series of 5 bars with the following true range (TR) and directional movement components already computed from high/low/close (the ADX math then uses these components). This is a common way to structure a worked example: separate “raw calculation” from “ADX smoothing.”
Assumptions:
- Lookback length for smoothing is 3.
- We will compute DI+ and DI− using the standard ratio:
- DI+ = 100 × (smoothed +DM) / (smoothed TR)
- DI− = 100 × (smoothed −DM) / (smoothed TR)
- For smoothing we use a simple exponential-style recursive idea described as “Wilder-style smoothing” at a conceptual level: each new smoothed value is mostly the prior smoothed value plus the current value spread across the length. (Exact seeding can vary, so treat the numbers below as one consistent worked set.)
- We will compute DX from DI+ and DI− each bar:
- DX = 100 × |DI+ − DI−| / (DI+ + DI−)
- ADX at bar t is the smoothed average of DX over the length; here we demonstrate the workflow across the bars, not a trading outcome.
Table of inputs
Assume the following per-bar values for directional movement and true range:
- Bar 1: +DM=6, −DM=2, TR=10
- Bar 2: +DM=3, −DM=5, TR=9
- Bar 3: +DM=7, −DM=1, TR=11
- Bar 4: +DM=4, −DM=4, TR=10
- Bar 5: +DM=2, −DM=6, TR=12
Step 1: compute “smoothed” +DM, −DM, and TR (length = 3)
We pick one consistent smoothing approach for illustration. Using length 3, the update concept is:
- smoothed_new = smoothed_old − (smoothed_old / 3) + current_value
Seed assumption:
- At Bar 1, smoothed values start equal to the Bar 1 values.
So:
- Bar 1 smoothed +DM=6, −DM=2, TR=10
- Bar 2 smoothed +DM = 6 − 2 + 3 = 7
- Bar 2 smoothed −DM = 2 − 0.666… + 5 = 6.333…
- Bar 2 smoothed TR = 10 − 3.333… + 9 = 15.666…
- Bar 3 smoothed +DM = 7 − 2.333… + 7 = 11.666…
- Bar 3 smoothed −DM = 6.333… − 2.111… + 1 = 5.222…
- Bar 3 smoothed TR = 15.666… − 5.222… + 11 = 21.444…
Continue similarly for Bar 4 and Bar 5 (numbers will be approximate if shown with rounding).
Step 2: compute DI+ and DI−, then DX
For Bar 3 (as an example calculation):
- DI+ = 100 × (11.666…) / (21.444…) ≈ 54.4
- DI− = 100 × (5.222…) / (21.444…) ≈ 24.4
- DX = 100 × |54.4 − 24.4| / (54.4 + 24.4)
- DX ≈ 100 × 30.0 / 78.8 ≈ 38.1
At each subsequent bar, you compute DI+ and DI− from the smoothed values, then compute DX from those. Finally, ADX is a smoothed average of DX over the chosen length.
What this shows (conceptually)
In a “ADX Trend” framing, the ADX value is interpreted as the strength of directional movement (how decisive DI+ vs. DI− has been). Whether DI+ exceeds DI− affects which direction has been stronger, but ADX itself is not a direct direction label. A worked example like this clarifies that: the computation relies on relative movement and scaling, then smoothing.