The stochastic oscillator measures where the latest close sits within the high-low range of the last 14 candles, on a 0–100 scale. Near 100 means price is closing near the top of its recent range; near 0, near the bottom. With 14,3,3 settings it plots two lines, %K and %D, and flags overbought above 80 and oversold below 20.
George Lane popularised it in the 1950s with an observation: in uptrends, closes tend to land near the top of the range; in downtrends, near the bottom. When that habit breaks, momentum is shifting. Like all oscillators, the stochastic oscillator is derived entirely from price, so it reorganises what the chart already shows rather than adding new information.
How the stochastic is calculated
Fast stochastic
- Raw %K = (close − lowest low of 14 periods) ÷ (highest high of 14 periods − lowest low of 14 periods) × 100
- Fast %D = 3-period simple moving average of raw %K
The fast version is jumpy, so most platforms default to the slow version.
Slow stochastic (14,3,3)
- Slow %K = 3-period SMA of raw %K (this is the first “3”)
- Slow %D = 3-period SMA of slow %K (the second “3”)
In other words, the slow %K is the fast %D, and the slow %D smooths it once more. The three numbers in “14,3,3” are the lookback, the %K smoothing and the %D smoothing.
Worked example
Over the last 14 candles, the highest high is $120 and the lowest low is $100. Today closes at $115.
Raw %K = (115 − 100) ÷ (120 − 100) × 100 = 15 ÷ 20 × 100 = 75.
The last three raw %K values are 75, 80 and 85 → slow %K = (75 + 80 + 85) ÷ 3 = 80.
The last three slow %K values are 72, 76 and 80 → slow %D = (72 + 76 + 80) ÷ 3 = 76.
Reading: %K at 80 is right at the overbought threshold and above %D, so short-term momentum is still rising. Note what the formula ignores: how far price moved. A close of $115 in a $100–$120 range scores 75 whether it got there in a grind or a spike.
Overbought and oversold: trends vs ranges
The 80/20 thresholds behave very differently depending on market regime.
In a range, the stochastic works as designed. Price oscillates between support and resistance, and the oscillator swings between the zones with it. A move below 20 near support, followed by %K crossing back above %D and above 20, is a reasonable long trigger. The mirror applies at resistance.
In a trend, the zones mislead. During a strong advance, closes keep landing near the top of the 14-period range, so the stochastic pins above 80 for long stretches. Selling there means fighting the trend. The better use in trends:
- Uptrend: ignore overbought readings. Wait for a pullback that pushes the stochastic below 20 (or into the 20–40 area), then buy the cross back up.
- Downtrend: ignore oversold readings. Wait for a bounce into 60–80+, then look for the cross back down.
A simple regime filter: if price is above a rising 50 EMA, only take stochastic buy signals; below a falling one, only sell signals.
Divergences
A divergence happens when price and the stochastic disagree:
- Bearish divergence: price makes a higher high, the stochastic makes a lower high. Closes are landing less firmly at the top of the range — buying pressure is fading.
- Bullish divergence: price makes a lower low, the stochastic makes a higher low. Selling pressure is weakening.
Two cautions. First, because the stochastic is fast, it produces many small divergences; the meaningful ones form between swings visible on the price chart, ideally with the first oscillator extreme in the overbought or oversold zone. Second, divergence is a warning, not a timing tool. Wait for price confirmation — a break of a minor swing, a bearish engulfing at resistance, or a neckline break in a double top.
Stochastic vs RSI
Both are bounded 0–100 momentum oscillators, which is why traders often stack them. They are not the same thing.
| Stochastic (14,3,3) | RSI (14) | |
|---|---|---|
| Measures | Close’s position within the high-low range | Average gain vs average loss on closes |
| Uses highs/lows | Yes | No, closes only |
| Speed | Faster, more extreme readings | Smoother, extremes rarer |
| Default zones | 80 / 20 | 70 / 30 |
| Best at | Timing turns inside ranges and pullbacks | Gauging momentum strength and trend health |
| Weakness | Many whipsaws in trends | Slower to flag short-term turns |
In practice, choose one as your primary momentum tool. Using both mostly confirms the same idea twice. If you want RSI’s side of the comparison, read RSI, MACD, ATR and EMA explained.
Settings cheat sheet
| Use case | Settings | Zones | Notes |
|---|---|---|---|
| Default, most charts | Slow 14,3,3 | 80 / 20 | The shared reference |
| More responsive | Fast 14,3 | 80 / 20 | Expect more false signals |
| Smoother swing signals | Slow 21,5,5 | 80 / 20 | Fewer, later crosses |
| Short-term pullbacks in trends | Slow 5,3,3 | 80 / 20 | Use only with a trend filter |
| Strong trends | Slow 14,3,3 | 85 / 15 or 70 / 30 by direction | Shift zones toward the trend |
As with all period-based indicators, the numbers are candles, not days, so the same settings work on any timeframe.
How to combine the stochastic with other tools
A practical four-step framework:
- Trend sets the bias. EMA slope and swing structure decide whether you only look for longs, only shorts, or fade both sides of a range.
- Levels set the location. Signals at real support and resistance beat signals in the middle of nowhere.
- Stochastic times the entry. A %K/%D cross out of the opposite zone, in the trend direction.
- ATR sets the stop. Structure plus a buffer of 0.5–1 ATR, sized with the position size calculator.
A combined example
EUR/USD on the 4-hour chart is above a rising 50 EMA. Price pulls back to a former resistance level at 1.0850 that has turned into support. The stochastic dips to 14, then %K crosses above %D and back above 20. A hammer closes at 1.0862. ATR is 18 pips.
Plan: entry 1.0865, stop 1.0832 (below the level and the hammer’s low with a buffer), target the recent high at 1.0935. Risk 33 pips, reward 70 pips, R:R ≈ 2.1:1 — check it with the risk/reward calculator. Invalidation: a 4-hour close below 1.0832, or the stochastic rolling back under 20 without price making progress.
Common mistakes
- Selling overbought in an uptrend. The stochastic can sit above 80 for weeks.
- Trading every crossover. %K and %D cross constantly; only crosses at levels and in the trend direction matter.
- Stacking it with RSI. Two momentum oscillators agreeing is one opinion, twice.
- Reading tiny divergences. Use swings that are visible on the price chart, and wait for confirmation.
- Confusing fast and slow. Know which version your platform shows; 14,3 and 14,3,3 give different lines.
Read it automatically with SnapPulse
SnapPulse turns a photo or screenshot of a candlestick chart into a plan in about five seconds: the detected pattern with directional bias and confidence, key support and resistance, RSI, MACD, ATR and EMA explained in plain English, an entry zone, stop, target and risk/reward ratio, plus a 1–10 risk score. It’s a quick second opinion to compare with your stochastic read, and the Coach can answer follow-up questions about the momentum picture. Download SnapPulse to try it on your own charts.
Educational content — not financial advice.
