The short answer
The EMAs most gold traders use are the 9 and 21 for fast intraday trends, the 20 and 50 for the intermediate trend and the 200 for the big picture. We tested six common setups on two years of hourly XAUUSD data with a $0.30 per ounce cost on every trade. All six finished positive, and the 9/21 crossover was the strongest and most consistent across both years.
Before you rush to trade it: every EMA rule won only 16% to 35% of its trades, sat through drawdowns of $700 to $1,100 an ounce, and was tested in one of the strongest trending periods gold has ever had. EMAs are good at riding trends and bad at ranges, and the table shows both sides.
What an EMA measures
An exponential moving average is an average of recent closing prices that gives more weight to the latest bars. It smooths out noise so you can see direction. A fast EMA (9 or 20 periods) hugs price; a slow one (50 or 200) shows the longer trend. When a fast EMA crosses a slow one, the short-term trend has changed direction relative to the longer one.
For the full method, including pullback entries, SMA versus EMA and the golden and death cross, read our lesson on moving averages for gold. This page is about which settings did what on real data.
The EMA settings gold traders use
| Setup | Typical use | Character |
|---|---|---|
| 9/21 EMA cross | Intraday trend changes | Fast, many signals |
| 20/50 EMA cross | Intraday and swing trend | Medium speed |
| 50/200 EMA cross | Long-term trend (golden or death cross) | Slow, few signals |
| Price vs 20 EMA | Short-term momentum | Very many signals |
| Price vs 50 EMA | Intermediate bias | Medium |
| Price vs 200 EMA | The big-picture filter | Slow; most often a filter |
Why gold’s volatility matters for EMA choice
Gold’s average daily range on COMEX futures was about $34 in 2024, $64 in 2025 and around $123 so far in 2026. On the hourly chart, ATR(14) averaged $17.49 over our test window. When volatility rises, price spends more time far from its averages and swings across fast EMAs more often. That creates more crosses and more small losses in choppy weeks, but also bigger wins when a trend runs. A slower pair trades less and gives back more of each move before it exits.
We tested 6 EMA 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. Crossover rule: long when the fast EMA crosses above the slow EMA, short when it crosses below, always in the market. Price rule: long when the close crosses above the EMA, short when it closes below. 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, though year 2 included a run to about $5,627 in January 2026 and a sharp fall.
Hourly (H1) results
| Setting | Trades | Win rate | Avg per trade | Profit factor | Net result | Max drawdown | Year 1 | Year 2 |
|---|---|---|---|---|---|---|---|---|
| EMA 9/21 cross | 493 | 33.3% | +$4.64 | 1.29 | +$2,287 | $901 | +$999 | +$1,288 |
| EMA 20/50 cross | 209 | 35.4% | +$8.83 | 1.39 | +$1,846 | $1,104 | +$1,354 | +$492 |
| EMA 50/200 cross | 69 | 31.9% | +$13.16 | 1.33 | +$908 | $883 | +$467 | +$442 |
| Close vs 20 EMA | 1,495 | 22.9% | +$0.50 | 1.06 | +$741 | $862 | +$1,140 | −$399 |
| Close vs 50 EMA | 885 | 20.6% | +$1.52 | 1.15 | +$1,347 | $881 | +$1,507 | −$160 |
| Close vs 200 EMA | 395 | 16.5% | +$2.26 | 1.19 | +$891 | $687 | +$618 | +$273 |
Daily (D1) results since 2015
| Setting | Trades | Win rate | Avg per trade | Profit factor | Net result | Max drawdown | 2015–20 | 2021–26 |
|---|---|---|---|---|---|---|---|---|
| EMA 9/21 cross | 128 | 31.2% | +$15.57 | 1.65 | +$1,992 | $452 | −$53 | +$2,046 |
| EMA 20/50 cross | 54 | 35.2% | +$34.92 | 2.06 | +$1,886 | $485 | −$247 | +$2,133 |
| EMA 50/200 cross | 18 | 22.2% | +$28.36 | 1.24 | +$510 | $1,275 | +$312 | +$199 |
| Close vs 20 EMA | 418 | 26.8% | −$0.18 | 0.99 | −$77 | $1,415 | −$880 | +$803 |
| Close vs 50 EMA | 250 | 22.0% | +$5.72 | 1.36 | +$1,430 | $953 | −$406 | +$1,836 |
| Close vs 200 EMA | 118 | 27.1% | +$16.17 | 2.02 | +$1,908 | $959 | +$74 | +$1,834 |
What the results say, in plain English
- Crossovers beat price-vs-EMA on the hourly chart. The three crossover rules made more per trade and held up in both years. The “close crosses the EMA” rules flipped far more often (1,495 trades for the 20 EMA) and the 20 and 50 versions lost money in year 2, when gold swung violently without making net progress.
- The 9/21 cross was the most even. It made $999 in year 1 and $1,288 in year 2. The 20/50 cross did better per trade, but most of its gain came from the cleaner trend of year 1.
- Low win rates are normal. The close-vs-200-EMA rule won only one hourly trade in six and still finished positive, because a few long trends paid for many small losses. If you cannot sit through five or six losers in a row, you will abandon a rule like this at the worst moment.
- The daily results are a bull-market story. Almost every daily gain came after 2020. From 2015 to 2020, most of the same rules were flat or negative. The 9/21 daily cross made about $2,402 an ounce on longs and lost about $410 on shorts.
- Drawdowns were large. The hourly crosses had drawdowns of $900 to $1,100 per ounce. On a 0.10 lot position (10 ounces), that is a $9,000 to $11,000 equity dip.
How to use EMAs on different timeframes
Scalping (M1 to M5). The 9 and 21 EMAs are popular for direction, but on these timeframes they cross many times an hour. Use them to stay on the right side of the move, and take entries from levels rather than from every cross. Check what each move is worth at your lot size with the gold pip calculator.
Intraday (M15 to H1). The 20/50 pair or the 9/21 pair works as a trend map. A popular approach is to trade only in the direction of the H1 9/21 or 20/50 alignment and to enter on pullbacks to the 20 EMA.
Swing (H4 to D1). The 50 and 200 EMAs define the bigger trend. Our daily data shows the price-vs-200-EMA rule did well from 2021 onwards but only broke even before that, so treat it as a filter, not a guarantee.
A combination idea: EMA filter plus a structure entry
The cleanest use of EMAs on gold is to let them decide which side you trade, and let price structure decide where:
- On H1, only look for longs when the 20 EMA is above the 50 EMA (shorts when below).
- Mark the last swing low and any support zone the pullback is heading into.
- Enter when price rejects that zone, with the stop beyond the swing low.
- Size the trade from the stop distance with the position size calculator.
We saw the value of a direction filter in other tests too: adding a 200 EMA filter to the Heikin Ashi rule improved its profit factor from 1.12 to 1.23. If you prefer a volatility-adjusted trend line, compare the Supertrend results, which behaved quite differently in year 2.
Common mistakes with EMAs on gold
- Trading every cross in a range. In sideways weeks, fast EMAs cross back and forth and each cross is a small loss.
- Judging a rule by its win rate. A 20% win rate can be profitable and a 60% win rate can lose. Look at average result per trade and drawdown.
- Optimising to the last trend. The 20/50 cross looked best in year 1, and much weaker in year 2.
- Ignoring the short side. On daily data, shorts lost money for most EMA rules because gold rose so much. A rule that only works on one side is really a bet on the trend continuing.
- Forgetting costs on fast rules. The close-vs-20-EMA rule paid about $449 an ounce in costs over 1,495 trades.
These results come from simple mechanical rules on one instrument and one historical period, which was unusually strong for trend-following. They are not recommendations. A different period, broker feed or cost level will give different numbers.
Test it yourself
Pick one EMA setup, write the exact rule, and run it on your own platform’s XAUUSD data. Include your real spread, split the test into two halves, and record the worst losing streak as well as the net result. Our lessons on backtesting a gold strategy and trading math and expectancy show what to measure. The gold indicator settings hub compares the EMA results with 11 other indicators tested the same way.
Test it with other traders
The best way to find out whether an EMA setup suits you is to trade it in a demo account alongside other people and compare notes. In the XAUUSD Global community, traders share their tested settings, journals and charts every day. 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 EMA for XAUUSD?
There is no single best EMA. In our two-year hourly test, the 9/21 EMA crossover produced the largest net result (plus $2,287 per ounce over 493 trades after costs) and was positive in both years, while the 50/200 cross traded only 69 times. The 200 EMA is most useful as a trend filter rather than a trigger.
Is the 9/21 or 20/50 EMA cross better for gold?
On hourly gold over two years, the 9/21 cross made more in total and was more even across the two years, while the 20/50 cross made more per trade but most of its gain came in the first year. Both won only about a third of their trades.
Does the 200 EMA work on gold?
As a simple rule (long when the close crosses above the 200 EMA, short when it crosses below), it was positive on both hourly and daily data in our test, but it won only 16.5% of hourly trades and its daily gains came almost entirely after 2020. Most traders use it as a filter for direction.
Why do EMA strategies have such low win rates?
Trend-following rules lose small amounts many times when the market chops sideways and make most of their money on a few large trends. In our hourly test, the close-versus-200-EMA rule won one trade in six, and still finished positive because the winners were much larger.
Which EMA is best for gold scalping?
Scalpers often use a fast pair such as the 9 and 21 EMA on 1 to 5 minute charts. Our test was on the hourly chart; on lower timeframes there are many more crosses and the spread takes a bigger share of each trade, so test on your own data first.
EMA or SMA for gold?
The EMA reacts faster because it weights recent prices more heavily, which is why it is more popular for intraday gold trading. The SMA is smoother and slower. Our moving averages lesson compares them in detail.
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.