Best Stochastic Settings for Gold: 5,3,3 vs 14,3,3 Tested on XAUUSD

The short answer

The most common stochastic settings on gold are 14,3,3 (the default), 5,3,3 for fast scalping, 8,3,3 as a middle ground and 21,5,5 for slower, smoother signals. We tested all four on two years of hourly XAUUSD data with a $0.30 per ounce cost per trade. All four lost money on the hourly chart, and so did a version that only traded in the direction of the 200 EMA.

The useful lesson is not “find a better setting”. It is that the stochastic, like RSI, is a poor standalone reversal signal on a trending market. It works better as a timing tool inside a plan that already knows the trend and the level.

What the stochastic measures

The stochastic oscillator shows where the latest close sits within the recent high-to-low range, on a scale of 0 to 100. With a 14-period setting, a reading of 90 means the close is near the top of the last 14 bars’ range. The three numbers in a setting such as 14,3,3 are:

• 14: the lookback for the high-low range (%K)
• 3: smoothing applied to %K (the “slow” stochastic)
• 3: the moving average of %K, called %D, the signal line

Readings above 80 are called overbought and below 20 oversold. The classic signal is %K crossing %D inside one of those zones.

The stochastic settings gold traders use

Why gold’s volatility changes things

The stochastic is built from the high-low range, so it reacts directly to volatility. Gold’s average daily range on COMEX futures went from about $34 in 2024 to $64 in 2025 and about $123 so far in 2026. Over our test window the hourly ATR(14) averaged $17.49. When ranges are large and gold trends, the close sits at the top or bottom of the range for many bars in a row, so the stochastic pins near 100 or 0 and every small dip produces a “signal”. Faster settings suffer most.

We tested 5 stochastic setups on 2 years of hourly gold data

For context, holding gold gained about $1,194 an ounce in year 1 and about $14 in year 2, although year 2 included a rally to about $5,627 in January 2026 and a steep fall.

Hourly (H1) results

Daily (D1) results since 2015

What the results say, in plain English

• Every hourly setting lost, in both years. That consistency is the most important finding. It was not one bad stretch; the rule was on the wrong side of gold’s moves throughout.
• The default 14,3,3 was the worst hourly setting, losing about $2,742 an ounce over 811 trades, with a drawdown of about $2,845.
• Shorts did more damage than longs. In all four unfiltered hourly settings, the short trades lost more. In a rising market, “sell when overbought” is fighting the tide.
• The 200 EMA filter did not rescue it on H1. Taking only trades in the trend direction cut the number of trades by more than two-thirds, but the result was still a loss. Buying a stochastic turn in an uptrend often meant buying a dip that kept dipping.
• One daily setting was positive. 8,3,3 on the daily chart made $857 an ounce since 2015 and was positive in both halves. With one success out of four plain settings, that could easily be chance. It is a candidate to test further, not a discovery.

The stochastic and RSI behaved in a very similar way in our tests. Compare the RSI settings results: both oscillators lost as reversal rules on hourly gold.

How to use the stochastic on different timeframes

Scalping (M1 to M5). 5,3,3 is popular because it moves quickly. On these charts it will cross many times an hour. Use it only at a level you marked beforehand, in the direction of the higher-timeframe trend, and accept that many crosses should be ignored. Check how much each move is worth with the gold pip calculator.

Intraday (M15 to H1). 14,3,3 is fine for reading momentum. Rather than trading every cross in the extreme zones, watch for the stochastic to turn back up from below 50 in an uptrend: a sign that a pullback is losing steam.

Swing (H4 to D1). Slower settings such as 14,3,3 or 21,5,5 give fewer readings. Our daily results suggest a turn from an extreme is a weak trigger on its own; combine it with weekly support and resistance.

A combination idea: stochastic for timing, structure for location

1. Define the trend on H4 with market structure or the 200 EMA.
2. Mark the zone where you would expect a pullback to stop, such as a prior breakout level.
3. Wait for price to reach it. Only then look at the stochastic on M15 or H1.
4. Take the trade when %K crosses back above %D, with the stop beyond the zone and size set by your risk, using the position size calculator.

Here the stochastic is the last item on a checklist, not the first. In ranging conditions, the ideas in our mean reversion lesson explain when fading extremes has a better chance, and the RSI divergence lesson shows how momentum divergence is combined with a structure break.

Common mistakes with the stochastic on gold

• Selling every reading above 80. In a trend, the stochastic can stay above 80 for most of a session.
• Running RSI and the stochastic together. They tell you almost the same thing; two agreeing oscillators are not two confirmations.
• Using fast settings without counting costs. The 5,3,3 hourly rule traded 868 times, paying about $260 an ounce in costs.
• Cherry-picking the one setting that worked. The daily 8,3,3 result looks good in isolation. Look at its neighbours: 5,3,3 and 14,3,3 lost.
• Fixed-size stops. Gold’s hourly range in 2026 has often been several times what it was in 2024. Check current ATR on the gold volatility tool.

Test it yourself

Write your stochastic rule down exactly, including the stop, target and exit. Run it on a year or more of your broker’s XAUUSD data, with your real spread. Split the sample in half, as we did, and check whether both halves agree. If you want to test the daily 8,3,3 result, do it on a different period or a different gold feed: that is the only way to tell an edge from a fluke. Our backtesting lesson walks through the process, and the gold indicator settings hub shows all 12 indicators side by side.

Test it with other traders

Testing alone makes it easy to fool yourself. Inside the XAUUSD Global community, traders run settings like these on their own charts, post their results and challenge each other’s conclusions. 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