The short answer
The standard ADX setting is 14 periods, and the most quoted rule is to trade trends only when ADX is above 25 (some traders use 20). We tested ADX 7, 14 and 21 with 20 and 25 thresholds on two years of hourly XAUUSD data, with a $0.30 per ounce cost on every trade. The surprise: adding the ADX 14 filter made the results worse. Taking every +DI/−DI crossover with no filter made $1,386 an ounce; with ADX above 25 the same signals lost $743.
The fast ADX 7 with a 25 threshold was the best hourly setting, positive in both years. On daily data since 2015, ADX rules were flat to negative. ADX is a lagging measure of trend strength, and on a fast market like gold that lag can filter out the very moves you want.
What ADX measures
The Average Directional Index comes as three lines:
- +DI: how strong upward moves have been, relative to the average true range
- −DI: how strong downward moves have been
- ADX: a smoothed measure of the gap between +DI and −DI, from 0 to 100
ADX does not show direction. A rising ADX can mean a strong uptrend or a strong downtrend. Readings below about 20 are usually described as a weak or absent trend; above 25 as a trending market. Because ADX is a smoothed version of an already smoothed series, it reacts slowly.
ADX is built from the true range, the same input as ATR. Our lesson on ATR and Bollinger Bands for gold explains that side of volatility in depth.
The ADX settings gold traders use
| Setting | Typical use | Character |
|---|---|---|
| ADX 14, threshold 25 | The classic trend filter | Slow, strict |
| ADX 14, threshold 20 | A looser version | Slow, filters less |
| ADX 7, threshold 25 | Intraday traders who want faster readings | Quick, noisier |
| ADX 21, threshold 20 | Swing traders | Very smooth, late |
Why gold’s volatility changes things
ADX is normalised by the true range, so in theory it adjusts to volatility. In practice gold’s character has shifted fast. The average daily range on COMEX futures went from about $34 in 2024 to $64 in 2025 and about $123 so far in 2026; the hourly ATR(14) averaged $17.49 over our test window. In such a market, trends start abruptly, often on US data or Fed decisions, and a 14-bar double-smoothed indicator is late to confirm them. You can check the current hourly and daily ATR on our gold volatility tool.
We tested 5 ADX setups on 2 years of hourly gold data
How we tested. COMEX gold futures (GC=F) 1-hour bars from 30 October 2024 to 30 October 2026, 11,445 bars, plus daily bars from January 2015. Rule: go long when +DI crosses above −DI while ADX is above the threshold; go short when −DI crosses above +DI while ADX is above the threshold. Exit on the opposite DI cross. The last row takes every DI cross with no ADX filter, for comparison. Signals are taken on the close and filled at the next bar’s open. Each trade is 1 ounce and pays $0.30 per ounce for spread and slippage. Results are in US dollars per ounce; maximum drawdown is the largest fall in closed-trade equity. Year 1 ends on 30 October 2025.
For context, holding gold gained about $1,194 an ounce in year 1 and about $14 in year 2, with a rally to about $5,627 in January 2026 and a heavy fall in between.
Hourly (H1) results
| Setting | Trades | Win rate | Avg per trade | Profit factor | Net result | Max drawdown | Year 1 | Year 2 |
|---|---|---|---|---|---|---|---|---|
| ADX 14, DI cross, ADX>20 | 403 | 27.0% | +$0.01 | 1.00 | +$4 | $593 | +$80 | −$75 |
| ADX 14, DI cross, ADX>25 | 217 | 24.4% | −$3.42 | 0.74 | −$743 | $860 | −$152 | −$592 |
| ADX 7, DI cross, ADX>25 | 693 | 32.3% | +$2.37 | 1.25 | +$1,641 | $776 | +$402 | +$1,239 |
| ADX 21, DI cross, ADX>20 | 215 | 27.4% | +$0.83 | 1.06 | +$178 | $449 | −$45 | +$223 |
| DI cross, no ADX filter (14) | 909 | 31.0% | +$1.52 | 1.14 | +$1,386 | $833 | +$908 | +$478 |
Daily (D1) results since 2015
| Setting | Trades | Win rate | Avg per trade | Profit factor | Net result | Max drawdown | 2015–20 | 2021–26 |
|---|---|---|---|---|---|---|---|---|
| ADX 14, DI cross, ADX>20 | 87 | 28.7% | −$6.88 | 0.73 | −$598 | $931 | −$344 | −$254 |
| ADX 14, DI cross, ADX>25 | 53 | 22.6% | −$8.28 | 0.72 | −$439 | $947 | −$311 | −$128 |
| ADX 7, DI cross, ADX>25 | 162 | 37.7% | −$4.08 | 0.83 | −$661 | $1,496 | +$99 | −$761 |
| ADX 21, DI cross, ADX>20 | 45 | 31.1% | +$2.77 | 1.11 | +$125 | $323 | −$323 | +$448 |
| DI cross, no ADX filter (14) | 281 | 26.0% | +$0.54 | 1.03 | +$152 | $1,572 | −$590 | +$742 |
What the results say, in plain English
- The ADX 14 filter removed good trades. With no filter, DI crossovers made $1,386 an ounce. Requiring ADX above 20 cut that to about zero, and above 25 turned it into a $743 loss. The filter kept the rule out of the early part of new trends, then let it in late.
- A faster ADX did better. ADX 7 reacts sooner, so its filter allowed trades earlier in each move. It was the only filtered setting positive in both years. Most of its gain came from short trades (about $1,250 of the $1,641), and most of it arrived in the volatile year 2.
- Win rates were low everywhere, between 22% and 38%. DI crossovers whipsaw a lot; the rules relied on a few large trends.
- On the daily chart, ADX did not help. The unfiltered DI cross was roughly flat over almost 12 years, and the ADX 14 filters lost money in both halves.
- The lesson is about lag. ADX is not “wrong”, but on gold, where trends often start with a single news-driven burst, a slow strength filter confirms the trend after much of it has happened.
How to use ADX on different timeframes
Scalping (M1 to M5). ADX on the scalping chart itself jumps with every spike. A more useful approach is to read ADX on H1: a rising ADX there suggests a trending session, where breakouts tend to follow through; a falling ADX below 20 suggests a range, where fading the edges makes more sense.
Intraday (M15 to H1). If you want an ADX filter, our data suggests a faster setting (around 7) or a lower threshold (20) filters less of the move. Many traders watch the slope of ADX rather than a fixed level: rising means strengthening directional movement.
Swing (H4 to D1). On daily gold, ADX filters added nothing in our test. If you use it, use it to describe the market (trending versus ranging) rather than as an entry rule.
A combination idea: ADX as a regime switch, not a trigger
Instead of using ADX to approve entries, use it to choose which playbook to run:
- On H4, note whether ADX 14 is rising and above 20 (trending) or falling below 20 (ranging).
- Trending: look for pullback entries in the trend direction, for example to the 20 EMA, as covered in the moving averages lesson.
- Ranging: mark the range edges and wait for rejections there, with targets back toward the middle.
- Size every trade from the stop distance with the position size calculator.
For a trend tool that held up better in our tests, compare the Ichimoku cloud results, where the cloud did the filtering instead of ADX.
Common mistakes with ADX on gold
- Reading ADX as direction. ADX rising during a sell-off is a strong downtrend, not a bullish sign.
- Waiting for ADX 25 before entering a breakout. Our test suggests that on gold you are often entering late.
- Using one threshold forever. A level that suits a quiet market may filter everything out, or nothing, in a volatile one.
- Ignoring the DI lines. Most of the information about direction is in +DI and −DI, not in ADX.
- Stacking ADX with other trend filters. ADX plus a moving average plus the cloud can leave you with almost no trades and no evidence any of them helped.
These results come from simple mechanical rules on past gold data and are not recommendations. A filter that hurt in this period could help in another. Always test with your own costs and data.
Test it yourself
Run a simple experiment on your own platform: take any entry signal you already use and test it twice, once with an ADX filter and once without. If the filter removes more profit than loss, it is not earning its place. Use at least a year of XAUUSD data, include your real spread, and compare two separate periods. Our backtesting lesson shows how to set that up, and the gold indicator settings hub has the same test for 11 other indicators.
Test it with other traders
Filters like ADX are where traders most often fool themselves, because they feel safe. In the XAUUSD Global community, traders compare filtered and unfiltered versions of their setups and review each other’s journals. The first month is free, then $19.99 a month.
Educational content only, not financial advice. Trading gold with leverage carries a high risk of loss.
Frequently asked questions
What is the best ADX setting for gold?
In our two-year hourly test, the fast ADX 7 with a 25 threshold gave the best result, plus $1,641 per ounce over 693 trades after a $0.30 per ounce cost, positive in both years. The standard ADX 14 with a 25 threshold lost $743. On the daily chart since 2015, no ADX setting produced a meaningful edge.
Does ADX above 25 mean gold is trending?
ADX above 25 means directional movement has been strong over roughly the last few weeks of bars, but it says nothing about direction and it lags. By the time ADX 14 rises above 25, a good part of the move has often happened, which is why the filter cut profitable trades in our test.
Should I use ADX 20 or ADX 25 on XAUUSD?
With ADX 14 on the hourly chart, the 20 threshold broke even (plus $4 over 403 trades) and the 25 threshold lost $743 over 217 trades. Both did worse than taking every DI crossover with no ADX filter, which made $1,386. A lower threshold filters less and hurts less.
What do +DI and −DI mean?
+DI measures upward directional movement and −DI downward movement, each as a share of the average true range. When +DI is above −DI, upward moves have been stronger. ADX is a smoothed measure of how far apart they are, which is why it shows strength but not direction.
Is ADX useful for gold scalping?
ADX on M1 to M5 charts reacts to every burst of volatility and gives little useful context. Scalpers who use it usually read ADX on H1 or H4 to decide whether the day is trending or ranging, then trade lower timeframes accordingly.
Sources and further reading
- CME Group — Gold futures overview — Contract specifications and trading hours for COMEX gold
- CFTC — Commitments of Traders — Weekly futures positioning, including gold
- Investor.gov — Stop order — How stop orders work, and their limits in fast markets
External sites are independent of XAUUSD Global. Check dates and data yourself.