The short answer
The standard Bollinger Band setting is 20,2: a 20-period simple moving average in the middle, with an upper and lower band 2 standard deviations away. Some gold traders widen it to 20,2.5 so the bands are touched less often. Both are reasonable settings for reading the chart.
We ran both as a simple “fade the band” rule on gold. On two years of hourly XAUUSD data, about 6 band touches in 10 came back to the middle band, but the ones that kept running went much further. The lesson is not “find a magic setting”. It is that a band touch alone is not a reason to trade against gold.
What Bollinger Bands measure
Bollinger Bands have three lines:
- Middle band: a 20-period simple moving average (the average close of the last 20 candles).
- Upper band: the middle band plus 2 standard deviations.
- Lower band: the middle band minus 2 standard deviations.
A standard deviation is a measure of how spread out prices have been. When gold is calm, closes stay near the average, so the bands squeeze together. When gold is wild, closes spread out, so the bands widen.
So the bands do two jobs at once:
- They show stretch: how far price is from its recent average.
- They show volatility: how wide the recent swings have been.
The two numbers in a setting such as 20,2 are the length (20 candles) and the width (2 standard deviations).
How often gold closes outside the bands
Here is how often a candle closed outside the bands on our data. Hourly data is COMEX gold futures (GC=F) from 11 June 2024 to 12 November 2026, about 13,900 bars, via Yahoo Finance. Daily data is the same contract from January 2015 to 12 November 2026.
| Setting | Hourly gold: closes outside bands | Daily gold since 2015: closes outside bands |
|---|---|---|
| 20,2 | 12.7% | 13.3% |
| 20,2.5 | 4.5% | 3.4% |
Textbooks often say that about 95% of prices sit inside 2-standard-deviation bands. On gold, it was closer to 87%. Gold has “fat tails”: big moves happen more often than a neat bell curve expects. That is the first reason fading the bands is dangerous on this market.
How the band fade behaves on gold
The rule we ran is the one most beginners try:
- Buy when a candle closes below the lower band. Sell when a candle closes above the upper band.
- Exit when price closes back at the middle band, or after 20 candles, whichever comes first.
- Cost: $0.30 per ounce on every trade (a rough spread-plus-slippage figure).
- One trade at a time. Results are in dollars per ounce (0.01 lot is usually 1 ounce, so $1 per ounce = $1 on 0.01 lot).
| Chart and setting | Trades | Share of trades that reached the middle band |
|---|---|---|
| Hourly, 20,2 | 567 | about 62% |
| Hourly, 20,2.5 | 352 | about 59% |
| Daily since 2015, 20,2 | 119 | about 63% |
| Daily since 2015, 20,2.5 | 53 | about 60% |
Source: our own run on COMEX gold futures (GC=F) data via Yahoo Finance, 17 November 2026. Past behaviour does not predict future results.
What the run shows, in plain English
- A high hit rate is not the whole story. Six touches in ten came back to the middle band. But the middle band is close, so those moves were small. When gold kept trending, the trades that did not come back ran much further. That is the shape of every fade rule: many small wins, a few big losses. Whether it adds up depends on your stop, your costs and the market you test in.
- Wider bands traded less often. 20,2.5 gave far fewer signals, so it paid less in costs and stood in front of fewer trend days.
- The rule had no stop. A real trader would use one, and that changes everything about the result. That is why we show how the rule behaves, not a dollar total.
The same pattern shows up with the RSI and the stochastic: on a trending market like gold in 2024–2026, “it’s stretched, so it must come back” is a weak idea by itself. Results depend on settings, timeframe and costs, so test any band rule on your own data before risking money.
Settings by timeframe
There is no proven best setting for each timeframe, but these are the common choices and why traders use them:
| Timeframe | Common setting | Why |
|---|---|---|
| 1–5 minute (scalping) | 20,2 | Default; many touches, so use only with a level and a trigger |
| 15 minute – 1 hour (intraday) | 20,2 or 20,2.5 | 2.5 cuts the number of touches on fast news days |
| 4 hour – daily (swing) | 20,2 | The default most traders watch; useful for squeeze and trend reading |
| Weekly | 20,2 | Shows when gold is historically stretched over months |
Changing the length (for example 10 or 50) changes how fast the bands react. Shorter is faster and noisier; longer is slower and smoother. If you change it, test it, and keep one setting for a long sample before judging it.
Where the bands stood on gold (12 November 2026)
On the daily chart, gold closed at $3,985.0 on 12 November 2026 (COMEX futures). The 20,2 bands were:
| Band | Level |
|---|---|
| Upper band | $4,202.0 |
| Middle band (20-day average) | $4,054.9 |
| Lower band | $3,907.9 |
Band width was about 7.3% of price. Price was below the middle band but inside the bands: no stretch signal either way. These numbers move every day, so check your own chart.
The squeeze: does low volatility predict a big move?
The Bollinger squeeze is when the bands come very close together. The idea is that calm periods are followed by big moves.
We checked this on daily gold since 2015. We marked every day when band width was the narrowest of the last 120 days (60 such days) and measured the high-to-low range over the next 20 days.
| Days | Average high-to-low range over the next 20 days |
|---|---|
| Squeeze days (120-day low in band width) | 7.12% |
| All days | 7.20% |
In this sample, a squeeze did not predict a bigger move. It was followed by a normal-sized move. That does not make the squeeze useless. It tells you volatility is low right now, so a breakout may come with a small stop. But it does not tell you the direction or the size, so you still need a level and a trigger.
How traders use Bollinger Bands better on gold
1. “Walking the band” means trend, not reversal
In a strong uptrend, gold can close near or above the upper band again and again. That is called walking the band. It is a sign of strength. Selling every touch of the upper band in that trend means standing in front of the move again and again.
2. Use the middle band as a pullback zone
In a trend, the 20-period middle band often works as a rough pullback area. A trader in an uptrend waits for price to dip toward the middle band, then looks for a bullish trigger candle at a real level. Our lesson on moving averages for gold explains pullbacks to an average step by step.
3. Fade only at a real level, in a range
Band touches work best when gold is clearly in a range and the band lines up with a real support or resistance level. Then the band is extra evidence, not the whole reason. Learn to find those levels in support and resistance on XAUUSD.
4. Read band width as a volatility gauge
Wide bands mean wide stops. Narrow bands mean the market is calm. Size your position from the stop distance, not from a fixed lot. Our ATR and Bollinger volatility lesson shows how.
A simple checklist before trading a band touch
- Is gold trending or ranging on the higher timeframe? If trending, do not fade the band against the trend.
- Is there a real level at the band (a previous high or low, a round number)?
- Is there a trigger candle (a clear rejection or a break of a small swing)?
- Is a big news release due soon? Bands widen fast on CPI, NFP and Fed days.
- Is your stop beyond the level, and is the target at least as far as the stop?
- Is your lot size set so the stop loses no more than your planned risk (often 1% or less)?
If you cannot tick all six, skip the trade.
Common mistakes
- Treating a band touch as a signal. It is a measure of stretch, not an instruction.
- Changing settings after every loss. You will end up with a setting that fits the past and fails next week.
- Ignoring news. On release days a 2-standard-deviation band can be broken in one candle.
- Using bands plus RSI plus stochastic. They often show the same thing. One stretch tool is enough.
- Forgetting costs. On low timeframes, the spread can be a big share of the distance to the middle band.
Test it yourself
Open the hourly gold chart, add 20,2 bands, and scroll back three months. Mark every close outside the bands, then mark where price was 20 candles later. You will see trend days where the band fade would have hurt, and range days where it worked. Keep notes in a trading journal and compare. Our full list of common settings is on the gold indicator settings page.
This page is education, not financial advice. Trading gold with leverage carries a high risk of loss. Never risk money you cannot afford to lose.
Frequently asked questions
What is the best Bollinger Band setting for gold?
20,2 (a 20-period average with bands 2 standard deviations away) is the standard and a fine default for reading gold. In our runs, neither setting turned a simple ‘fade the band’ rule into a reliable signal on hourly gold. The setting matters less than how you use the bands. Test any rule on your own data before risking money.
Is 20,2.5 better than 20,2 on XAUUSD?
Wider bands give fewer signals. On hourly gold, 20,2 had 12.7% of closes outside the bands and 20,2.5 only 4.5%. The wider setting traded less often, so it paid less in costs and stood against fewer trend days. Neither setting is a signal by itself.
What Bollinger settings are good for gold scalping?
Scalpers often keep 20,2 on the 5-minute or 15-minute chart. Lower timeframes produce more band touches and more costs, and gold often ‘walks the band’ in trends. Use the bands to see stretch and volatility, and take entries only at a level with a clear trigger.
Does the Bollinger squeeze work on gold?
A squeeze shows that volatility is low, not which way price will break. On daily gold since 2015, days when band width was at a 120-day low were followed by an average 20-day range of 7.1%, almost the same as the 7.2% average for all days. Treat a squeeze as a warning to get ready, not a signal.
Why is a 60% hit rate not enough for a Bollinger fade?
Because the wins are small and the losses can be big. The fade exits at the middle band, which is close by, but when gold keeps trending the losing trades run much further. A high win rate with a poor reward-to-risk ratio can still lose money, so check both before trusting any rule.
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
- Money Flow Index (MFI) on Gold: How It Works, MFI vs RSI, and the Volume Problem on XAUUSD
- Gold Trend Pullback Strategy: How to Buy the Pullback on XAUUSD
- How to Trade Gold in a Downtrend: Sell the Rally, Stops Above Structure, No Bottom Picking
- How to Find Daily Bias on Gold: ICT Daily Bias for XAUUSD, With Real Data
- ZigZag Indicator on Gold: Settings, the Repaint Problem and How to Use It to Mark Structure
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