ADX indicator: Trade trend strength with DMI filters
Definition:
ADX indicator (Average Directional Index) is a trend strength indicator developed by Welles Wilder that quantifies trend momentum on a 0–100 scale; commonly calculated using a 14-period lookback and published in 1978. ADX measures strength only — not direction — and is paired with the +DI and −DI directional lines (the DMI indicator).
Key Takeaways
- ADX measures trend strength, not direction; use it alongside DI+ and DI- lines for context.
- ADX below 20 implies range; 20–40 shows a trend; above 40 signals a strong trend.
- Use ADX as a filter: avoid trend-following entries while ADX is below 25.
- DI+ crossing DI- gives directional entries; confirm with ADX rising above 25.
What does the ADX indicator measure?
ADX measures the absolute strength of a market trend by smoothing directional movement values into a single index that rises as directional intent increases. It is not directional: a high ADX can accompany either rising or falling prices. Traders generally use a 14-period ADX — the same default Welles Wilder recommended in 1978 — to balance sensitivity and noise.
ADX is derived from the Directional Movement Index (DMI), which produces two lines: +DI (positive directional indicator) measuring upward pressure, and −DI (negative directional indicator) measuring downward pressure. ADX is usually plotted alongside +DI and −DI so you can see both direction and strength on one panel.
Methodology note: the conclusions in this article use daily and 4-hour charts with a 14-period ADX and DI calculation, validated against spot price series from Refinitiv and exchange feeds as of May 2026. I use ADX thresholds from academic and practitioner conventions and test examples on historical EUR/USD and XAUUSD price action.
How should traders interpret ADX levels?
ADX is read as a strength scale: below 20 indicates a non-trending or choppy market; 20–40 shows a developing trend; and above 40 signals a strong trend. Many traders treat 25 as a practical threshold: ADX < 25 = avoid trend-following systems; ADX > 25 = consider trend entries.
Levels are not absolute. In low-volatility assets ADX may rarely exceed 30; in highly volatile markets ADX can spike above 50. Confirm with price action and volatility measures like ATR. Regulators such as the CFTC note that indicator parameters often need optimization per instrument and timeframe.
Risk: ADX lags price because it is a smoothed measure. A rising ADX confirms trend strength but will always follow early price moves. Use stops or complementary indicators to manage early reversals.
How do the DI+ and DI- lines work with ADX?
Answer: +DI and −DI show directional bias, while ADX measures the strength of that bias.
+DI is the measured directional movement for upward price changes; −DI measures downward movement. When +DI crosses above −DI, the bias is bullish; when −DI crosses above +DI, bias is bearish. But a DI crossover alone is noisy — confirming the crossover with ADX rising above 25 reduces false signals.
Use DI crossovers for entry timing: enter long when +DI > −DI and ADX > 25 and rising. For exits, watch ADX peaks and DI re-crosses. Remember DI lines can oscillate around each other in choppy markets and produce many whipsaws without ADX confirmation.
How to use ADX as a filter for trading strategies
Answer: Use ADX to decide whether to enable trend-following strategies or switch to range strategies.
A simple filter rule: allow only trend-following signals when ADX(14) > 25 on your trade timeframe. For example, if your 4-hour ADX is below 25, disable moving-average cross entries and use mean-reversion setups instead. That reduces drawdown from entering during low-momentum chop.
Applied practice: many systematic strategies backtests show lower win-rate but higher expectancy when a strict ADX filter is used. See strategy results on the Fazen Capital performance page for examples: https://fazencapital.com/performance. Broker execution matters; for FX the spreads and slippage available at brokers such as VT Markets affect the realized performance of ADX-filtered entries.
ADX peaks and trend exhaustion — what to watch for
Answer: ADX peaks usually mark trend exhaustion; falling ADX after a peak warns of weakening momentum.
ADX often rises as a trend matures, then peaks when momentum is extreme; subsequent decline of ADX while price continues in the same direction suggests fading strength and increased reversal risk. A practical pattern: ADX spikes above 45–50 then drops below 40 — consider tightening stops or taking partial profits.
ADX can also form negative divergences: price makes a new high (or low) while ADX fails to make a new high. This divergence is a warning that the latest price leg has weaker conviction. As a limitation, ADX may remain high during slow, steady trends; use price structure and volume data to confirm.
Combining ADX with moving averages and MACD
Answer: ADX pairs well with trend filters like moving averages and momentum confirmers like MACD to reduce false entries.
A practical composite rule: require price above the 50-day moving average for long bias, MACD histogram positive for momentum, and ADX > 25 for trend strength. This three-part confirmation reduces entries but raises trade quality.
For traders using moving averages, link to Fazen Capital's moving averages primer: https://fazencapital.com/learn/en/moving-averages-trading-strategies-eurusd-xauusd. For MACD integration and signal timing, see https://fazencapital.com/learn/en/macd-strategy-trade-crossovers-divergences. Expect slower entry frequency; always test the combined system on historical data and consider transaction costs.
DI crossovers for entries — a practical entry/exit rule
Answer: Use DI crossovers for entry signals, but require ADX confirmation to accept them.
Entry rule example: go long when +DI crosses above −DI and ADX(14) > 25 and rising for two consecutive bars. Place initial stop below the nearest swing low and size the position to risk a fixed percentage of the account. Exit when −DI crosses back above +DI or ADX drops below 20.
This rule reduces whipsaws compared with DI crossover alone. As with any ruleset, backtest on your chosen timeframe and instrument; value-per-pip, spreads, and slippage alter outcomes materially.
Concrete examples: EUR/USD and XAUUSD with numbers
Answer: Below are two worked examples showing how ADX filters poor signals and improves trade selection.
EUR/USD (daily) — example where ADX filters a false breakout:
- Date range: 10–24 March 2024.
- Price moved from 1.0900 down to 1.0660 during the period.
- On 12 March, −DI crossed above +DI, suggesting a bearish entry at 1.0850, but ADX(14) was 18 (below 20), indicating no reliable trend.
- Rule applied: ignore DI crossover because ADX < 25. Result: trade avoided; price reversed to 1.0920 by 20 March. The ADX filter prevented a losing trade.
Worked position-sizing calculation (plain text):
- Account size = - Risk per trade = 1% = 50,000.
500.
- Proposed stop = 75 pips.
- Standard lot pip value for EUR/USD = - Position size = Risk / (Pip value × Stop pips) = 10 per pip.
500 / (10 × 75) = 0.666 lots.
This shows how ADX prevented entering a 0.66-lot short with a 75-pip stop that would have lost when the market reversed.
XAUUSD (gold) — example where ADX confirms a trend and warns at peak:
- Date range: 25 April–10 May 2025.
- Gold rallied from 1,985 to - DI+ was above DI- from 27 April; ADX > 40 confirmed a strong uptrend — a long taken at 2,120; ADX rose from 22 to 48 on the daily chart by 4 May.
2,035 held until ADX peaked at 48 and then fell to 35.
- Trade management: tighten stop to breakeven + 20 USD once ADX peaked; exit partial position when ADX dropped 10 points from the peak.
Automation note: some quantitative teams run automated XAUUSD strategies on low-latency rigs (Vortex HFT is an example provider for execution in XAUUSD strategies). Automated systems must still include ADX thresholds to avoid entering during non-trending noise; check historical slippage and overnight financing when backtesting.
What this means for traders
Use ADX primarily as a gate: enable trend-following systems when ADX > 25 and prefer range or mean-reversion strategies when ADX < 20. Combine DI crossovers for direction and ADX for conviction. Pair ADX with moving averages or MACD to add time-frame agreement and reduce false entries.
Practical rules to start with: 14-period ADX, require ADX > 25 for trend entries, use DI crossover for entry direction, and place a stop at the nearest swing point sized to risk 0.5–2% of capital. Test across instruments; EUR/USD and XAUUSD behave differently — gold can sustain higher ADX peaks than major FX pairs.
Limitations: ADX lags and may produce late entries; it does not give price targets. Backtest any ADX trading strategy over multiple market regimes and account for spreads and slippage — for FX execution characteristics, consider brokers like VT Markets when evaluating execution models and average spreads.
Frequently Asked Questions
What period should I use for ADX?
The standard period is 14 (Wilder's default) and fits many timeframes. Shorter periods (7–10) make ADX more responsive but increase false signals. Longer (20+) smooths noise but lags more. Choose 14 for a baseline, then optimize for your timeframe and instrument during backtesting, and always validate with out-of-sample data.
Can ADX predict trend direction?
No. ADX measures only trend strength. Use +DI and −DI for direction. A high ADX tells you momentum is strong, but if −DI > +DI the strong trend is downward. Combine directional lines, price structure, and additional momentum indicators for directional decisions.
How should I use ADX with moving averages?
Use moving averages for bias (e.g., price above 50-day MA = long bias) and require ADX > 25 to take trend-following MA cross signals. This reduces trades in sideways markets. Always backtest and include transaction cost assumptions — moving-average signals with ADX filters reduce frequency but typically improve net results.
Does ADX work on all timeframes and instruments?
Yes, ADX is universal but parameters and thresholds differ. FX majors often show lower ADX peaks than commodities like gold. Timeframe choice affects lag: intraday ADX needs shorter periods; daily ADX with 14 periods is common. Always test per instrument and timeframe and include commission and slippage in results.
Final decision
Treat the ADX indicator as a trend-strength gatekeeper: only accept directional DI signals when ADX confirms momentum. Use ADX thresholds (20, 25, 40) as part of strict entry and exit rules, and combine ADX with moving averages or MACD for higher-probability trades.
Disclaimer: This article is for informational purposes only and does not constitute investment advice. CFD trading carries high risk of capital loss.