The Stochastic Oscillator, Drawn Out
Stochastic gets filed next to RSI as a momentum oscillator. Both run 0 to 100, both come with overbought and oversold lines, so it's easy to treat them as interchangeable.
The point of this lesson is that they aren't. On identical data at the same period, the two give opposite answers to "was this market ever oversold" — because they measure different things.
The charts use the same window as the earlier lessons, with one difference: stochastic is the first indicator here that needs highs and lows, not just closes.
1. %K is a position inside a range
The question %K asks is simple. Across the last 14 bars there's a highest high and a lowest low; where in that range did this bar close? At the very bottom it reads 0, at the very top 100.
Where RSI looks at the balance of gains against losses, stochastic looks at position within the recent price range. Same 0-100 scale, different subject.
Over the last 14 bars, where did this close land between the highest high and the lowest low?
- Price (close)
- 14-bar high-low range
- %K
09/05: range $77,101 to $82,300, closed $79,609 → %K 48.2
BTC/USDT · 4h · Binance · 2026-08-03 → 2026-09-05
The shaded box is the last 14 bars' high-low range and the dot is this close. How high the dot sits inside that box is exactly the %K below.
2. %D and the crossings
%K on its own is jagged, so it's usually paired with %D — a short average of %K. Most stochastic strategies treat the two crossing as the signal.
%D is just a 3-bar average of %K. Most stochastic strategies trade where the two cross.
- %K
- %D
69 of 69 crossings reached.
BTC/USDT · 4h · Binance · 2026-08-03 → 2026-09-05
A marker lands at each crossing. Press play and watch how quickly they accumulate.
3. Changing the periods
The %K lookback sets how recent a range you're measuring against; the %D smoothing sets how much of %K's jitter survives. Both directly change how many signals you get.
A shorter lookback reacts faster and crosses far more often.
- %K
- %D
At %K 14 / %D 3: 69 crossings in this window.
BTC/USDT · 4h · Binance · 2026-08-03 → 2026-09-05
The figure below reports how many crossings that setting actually produces in this window.
4. Same data, opposite conclusions
Here's the part worth remembering. Below, %K and RSI are drawn on one panel. Both run 0 to 100 and both are set to 14, so they genuinely share a scale — which is what makes overlaying them legitimate rather than misleading.
Both are called momentum oscillators, both run 0 to 100, both set to 14 — on identical data.
- %K
- RSI 14
%K fell as low as 7.0 and spent 20 bars below 20. RSI never got below 33.5 and spent 0 bars below 30. Same window, same period, opposite conclusions about whether this market was ever oversold.
BTC/USDT · 4h · Binance · 2026-08-03 → 2026-09-05
Solid is %K, dashed is RSI. Watch how differently deep the two go through the middle of August.
%K fell to 7.0 and spent 20 bars below 20. Over that same window RSI never dropped below 33.5 and spent zero bars below 30.
Neither is broken. When price drifts down slowly inside a narrow range, closing near the bottom of that range is easy — so %K goes low. But the drops themselves are small, so the balance of gains against losses barely tilts — so RSI stays mid-scale. They're answering different questions.
5. It crosses constantly
One practical characteristic of stochastic is sheer signal count. For comparison, MACD's signal line crossed 6 times across this same window.
Every crossing in the window, marked on price. Most of them happen nowhere near an extreme.
- %K
- %K crosses up
- %K crosses down
69 crossings in this window — 47 of them (68%) with %K between 20 and 80, far from either band.
BTC/USDT · 4h · Binance · 2026-08-03 → 2026-09-05
Every crossing, marked on price. The shaded band below is the 20-80 middle, and most of them land inside it.
Two thirds of the crossings happen in the middle, nowhere near either band. That's why stochastic strategies rarely trade raw crossings — they usually add a condition like "only when %K is below a certain level", which is exactly what the Simulator's stochastic filter provides.
6. It pins to one end in a trend
The trap from the RSI lesson applies here unchanged. In a strong advance, most bars close near the top of their recent range, so %K simply stays above 80.
The longest unbroken stretch above 80, found in the data.
- Price (close)
- %K
%K held above 80 for 16 bars straight from 08/19, and 39 bars above it overall, while price went $68,554 to $78,411. The same trap RSI sets: a strong trend parks an oscillator at one end and leaves it there.
BTC/USDT · 4h · Binance · 2026-08-03 → 2026-09-05
The highlighted stretch is the longest unbroken run above 80 in this window.
7. Common misreadings
- Stochastic and RSI are not interchangeable. They share a scale, not a subject.
- A low %K means price closed near the bottom of its recent range, not that it fell a long way. In a narrow range a small dip reaches 0.
- 80 and 20 are conventions, and they differ from RSI's 70/30 for no deeper reason than convention.
- Crossings are common — 69 in this window alone, most of them far from either band.
- A strong trend parks stochastic at one end. Above 80 does not imply a pullback is due.
8. Seeing it run in the Simulator
Stochastic appears in the Simulator in two roles: a template that trades the %K/%D crossing directly, and an optional filter that gates another strategy's entries on the %K level.
To compare %K and %D combinations rather than guess, the Sweep tool runs a real backtest across every combination in the ranges you pick and ranks them.
- Stochastic Oscillator Crossover — trades the %K/%D crossing, with optional overbought/oversold zone filtering.
- Stochastic filter — an optional gate on another strategy's entries, keyed to the %K level.