The short answer
Williams %R shows where the latest close sits inside the high-to-low range of the last 14 candles, on a scale from 0 (top of the range) to −100 (bottom of the range):
- 0 to −20: “overbought” (closed near the recent high).
- −80 to −100: “oversold” (closed near the recent low).
The standard setting is 14 periods. On gold, the big lesson is that overbought does not mean “about to fall”. In our runs, a simple %R reversal rule hit its exit about six times in ten, yet the signals that fought a gold trend ran far against it. It behaved differently on the daily chart, so the timeframe matters as much as the setting. %R works best as a timing tool inside a trend, not as a reason to fight it.
What Williams %R measures
The formula, for a 14-period setting:
%R = (Highest high of 14 candles − Close) ÷ (Highest high − Lowest low) × −100
Example on gold: over the last 14 hourly candles, the highest high was $4,100 and the lowest low was $4,000. The latest close is $4,090.
%R = (4,100 − 4,090) ÷ (4,100 − 4,000) × −100 = 10 ÷ 100 × −100 = −10
That is in the overbought zone: gold closed in the top 10% of its recent range.
If the close were $4,015, %R would be (4,100 − 4,015) ÷ 100 × −100 = −85, oversold.
Williams %R vs the stochastic
They are almost the same indicator. Williams %R equals the fast stochastic %K minus 100. The stochastic adds a smoothing line (%D); Williams %R does not, so it is faster and jumpier. If you already use the stochastic, adding %R tells you nothing new. Our test of stochastic settings for gold found the same weakness.
Settings traders use
| Setting | Behaviour | Typical use |
|---|---|---|
| 7 | Very fast, many signals | Scalping, with a strong filter |
| 14 | Standard | Most charts; what we tested |
| 21–28 | Slower, smoother | Swing trading on 4-hour and daily |
The zones (−20 and −80) can also be moved. Some gold traders use −10 and −90 to see only extreme readings. Fewer signals, but each still needs a reason.
How much time gold spends overbought and oversold
| Chart | Above −20 (overbought) | Below −80 (oversold) |
|---|---|---|
| Hourly, Jun 2024 – Nov 2026 | 27.4% of bars | 14.8% of bars |
| Daily, Jan 2015 – Nov 2026 | 26.3% of bars | 20.4% of bars |
Our data: COMEX gold futures (GC=F) via Yahoo Finance, 14-period %R.
Gold spent almost twice as long overbought as oversold on the hourly chart. That is what an uptrend looks like on this indicator. A trader who sold every overbought reading would have been selling into strength a quarter of the time.
The overbought trap, with data
We checked what happened after gold was overbought on the daily chart since 2015.
| Daily readings | Average gold change over next 10 trading days | Share of times gold was higher 10 days later |
|---|---|---|
| After %R above −20 (758 days) | +0.38% | 53.3% |
| All days | +0.45% | 55.1% |
After an overbought reading, gold was still higher 10 days later more than half the time, and the average 10-day move was almost the same as on any other day. “Overbought” on its own told us almost nothing about what came next.
How a simple %R reversal rule behaves
Rules:
- Buy when %R crosses up through −80 (leaving oversold).
- Sell when %R crosses down through −20 (leaving overbought).
- Exit a buy when %R rises above −20, and a sell when %R falls below −80, or after 20 candles.
- Cost: $0.30 per ounce per trade (0.01 lot is usually 1 ounce).
| Chart | Trades | Share of trades that reached the opposite zone |
|---|---|---|
| Hourly, Jun 2024 – Nov 2026 | 831 | about 62% |
| Daily, Jan 2015 – Nov 2026 | 191 | about 65% |
Our own run, 17 November 2026. Past behaviour does not predict future results.
What the run means
- A high hit rate is not the same as a good rule. On the hourly chart, many small wins sat next to a few trades that kept running when gold trended, especially sells against the 2024–2026 uptrend. One of those trend days can undo many small wins, so the hit rate alone tells you very little.
- The daily chart behaved differently. Daily signals were fewer and slower, and gold rose strongly over those years, which favours the buy side. That is a feature of the period, not proof of an edge.
- The same tool behaved very differently on two timeframes. That is a warning against trusting any indicator result you have not tested on the chart you trade.
- The rule had no stop and fixed exits. A real plan would add a stop, a level and a trigger, and that changes the outcome. Results depend on settings, timeframe and costs, so test it on your own data before risking money.
How to test %R yourself in 20 minutes
- Open the gold chart you actually trade and add Williams %R (14).
- Scroll back three months. Mark every time %R leaves oversold (crosses up through −80) or leaves overbought (crosses down through −20).
- For each mark, write down whether gold was trending or ranging on the higher timeframe.
- Note where price was 10 and 20 candles later, and take off your real spread for each trade.
- Compare the signals taken with the trend against the signals taken against it.
Most traders who do this see the same thing: %R signals in the direction of the bigger trend are far more useful than signals that fight it.
How to use Williams %R better on gold
1. Trade with the trend, use %R for timing
In an uptrend (for example, price above the 200 EMA with higher highs and lows), ignore overbought and wait for %R to drop below −80. When it turns back up above −80, look for a buy trigger at a support level. In a downtrend, do the opposite.
2. Look for failure swings at levels
When gold reaches a clear resistance and %R fails to get back above −20 on the second push, momentum is fading. That is extra evidence, not a signal by itself. Divergence works in a similar way; see RSI divergence on gold.
3. Use one oscillator, not three
RSI, the stochastic and %R all measure similar things. Using all three makes you feel more certain without adding information. Pick one.
4. Respect news
On CPI, NFP and Fed days, %R can hit 0 or −100 in one candle and stay there. Readings around big news are not overbought or oversold in any useful sense.
A worked example (how to read it)
Gold is in a daily uptrend. On the hourly chart it pulls back from $4,120 to $4,070, and %R falls to −92. Price reaches a previous breakout level at $4,068. The next hourly candle closes at $4,078, and %R rises to −74, back above −80.
A trader buys at $4,078 with a stop below the level at $4,060 (risk $18) and a target at the recent high, $4,120 (reward $42). The %R reading did not create the trade. The trend and the level did. %R only helped with the timing.
This example is hypothetical, to show the method.
Where %R stood (12 November 2026)
On the daily chart, gold’s 14-day %R was −62.0 on 12 November 2026 (COMEX futures close $3,985.0): in the middle of the range, neither overbought nor oversold. This changes every day, so check your own chart.
Common mistakes
- Selling every overbought reading in an uptrend.
- Ignoring costs on low timeframes, where %R gives many small signals.
- Changing settings to fit last week’s chart.
- Using %R without a level or a trend filter.
- Stacking oscillators that all say the same thing.
Checklist
- What is the higher-timeframe trend?
- Is there a real support or resistance level here?
- Has %R left the zone in the trend’s direction (for example, back above −80 in an uptrend)?
- Is there a trigger candle?
- Stop beyond the level, lot sized for 1% risk or less?
- No major news soon?
Our gold indicator settings page collects every indicator setting we have tested on gold.
This page is education, not financial advice. Trading gold with leverage carries a high risk of loss.
Frequently asked questions
What is the best Williams %R setting for gold?
14 periods is the standard and the one we used. Shorter settings such as 7 react faster and give more signals; longer ones such as 21 are smoother. No setting turns the indicator into a reliable standalone signal; test any setting on your own chart first.
What does Williams %R above -20 mean?
A reading between 0 and -20 means gold closed near the top of its recent range, which is usually called overbought. It does not mean price must fall. On daily gold since 2015, the 10-day return after a reading above -20 averaged plus 0.38%, close to the plus 0.45% average for all days.
Is Williams %R the same as the stochastic?
Nearly. Williams %R is the fast stochastic %K turned upside down and shifted: %R = %K minus 100. It has no built-in smoothing line, so it moves faster and looks noisier.
Does Williams %R work on XAUUSD?
Not as a stand-alone reversal rule. In our hourly run, about six signals in ten reached the opposite zone, but the trades that fought a gold trend ran much further against the rule. It behaved differently on the daily chart, which shows how much the timeframe matters. Use %R for timing within a plan, not as the plan, and test it on your own data.
How do I use Williams %R in a trend?
In an uptrend, look for %R to dip into oversold (below -80) and then turn back up, as a timing signal to join the trend. Ignore overbought readings, because strong uptrends spend a lot of time there.
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.
Comments
No comments yet.
More in Strategies & technical analysis
- London Opening Range Breakout on Gold: Rules, Timing and Fakeouts on XAUUSD
- How to Backtest a Gold Strategy in Python: Free Data, Simple Code, Costs and Overfitting
- Williams Alligator and Fractals on Gold: Rules, Settings and a Tested XAUUSD System
- ICT Unicorn Model on Gold: Breaker Block Plus Fair Value Gap, Explained With XAUUSD Rules
- How to Find Daily Bias on Gold: ICT Daily Bias for XAUUSD, With Real Data
Latest on the blog
- Gold Down Days in a Row: What Happened Next After 4, 5 and 6 Straight Falls
- The Philly Fed Survey for Gold Traders: Diffusion Index, Prices Paid and the Thursday Reaction
- NR7 Breakout on Gold: How the Narrowest Day of the Last Seven Behaves
- What Is the Fed Beige Book? A Gold Trader's Guide to District Reports, Price Clues and Market Reaction