Technical Indicators · Stochastic

Where price closes in its recent range.

Stochastic measures the close's position within the high–low range over a lookback window, and outputs two lines — %K and %D — pinned to a 0–100 scale. Vant8 exposes it as a rule in the Custom Strategy Builder so composed strategies can gate on overbought / oversold levels and %K/%D crosses alongside RSI, MACD, MAs, and Bollinger Bands.

What it measures

Position in the recent range, on a 0–100 scale

If the most recent close is near the top of the lookback's high–low envelope, %K is high. Near the bottom of the envelope, %K is low. The bounded scale means overbought / oversold levels work the same on any price. %D is a smoothed follow-up line that filters some of the noise out of %K.

Overbought (≥ 80)

Recent closes are clustering near the top of the range. A conventional overbought threshold; reversion setups look for a fall out of this band.

Neutral (20 – 80)

Neither extreme dominates. Trend strategies care about other indicators in this band; reversion setups typically wait.

Oversold (≤ 20)

Recent closes are clustering near the bottom of the range. Reversion setups look for a rise out of this band.

Computation

Full Stochastic — raw %K smoothed twice

Vant8 ships the Full Stochastic: raw %K over the lookback window, smoothed to produce the slow %K, then smoothed again to produce %D. That's the standard formulation you'd compare against on any charting platform.

raw %K = 100 × (close − lown) / (highn − lown)
%K = SMA(raw %K, smoothK)
%D = SMA(%K, smoothD)
Where n is the lookback period and lown / highn are the lowest low and highest high over that window.
Defaults

The parameters we ship

Textbook defaults out of the box. Override them per rule when you compose a Custom Strategy.

Lookback: 14

Length of the high–low envelope used to normalise the close. The standard reference.

smoothK: 3

Simple moving average applied to raw %K. Filters single-bar spikes into a slow %K that's more actionable.

smoothD: 3

Simple moving average of %K to produce the %D signal line. Cross events between %K and %D are the classic trigger.

What you can act on

Threshold levels, %K/%D crosses, and divergence

Three families of setups the Custom Strategy Builder can express directly using the Stochastic rule.

Threshold level

Fires while %K (or %D) is above or below a configured band. Compose rules like "%K below 20" for oversold filters or "%K above 80" for overbought filters.

%K / %D cross

Fires when %K crosses %D from below (bullish) or above (bearish). A smoother trigger than a raw threshold cross; better in choppy conditions.

Divergence

Price makes a new high or low that Stochastic doesn't confirm. Reversal signal in reversion contexts; useful continuation-doubt in trend contexts.

Where it shows up

Strategies that consume Stochastic

Custom (Strategy Builder)
First-class rule — expose %K, %D, cross events, and threshold conditions in your composed strategy alongside RSI, MACD, MAs, ATR, and Bollinger Bands.

The shipped native strategies (RSI, MACD, MA Cross, Momentum, Mean Reversion, HODL, DCA, Grid, and the perp roster) do not read Stochastic — they use their own primary indicators. Stochastic is available specifically through the Strategy Builder for users composing rules that stack it against those other indicators.

How to read it

Where Stochastic can mislead you

Strong trends pin it to extremes

In a real uptrend %K can sit above 80 for days without reversing. Fading every overbought reading in a trend is a slow way to lose money — pair Stochastic with a trend filter (ADX, MAs) before acting on an extreme in isolation.

Sensitive to the lookback

A 14-bar lookback on a 1-hour chart is a very different signal from the same 14-bar lookback on a 5-minute chart. Match the period to the timeframe your Custom Strategy is trading on, not a textbook default.

Pick a strategy that uses Stochastic.

The strategies listed above are deployed from the Strategies page on your dashboard. Each one ships with sensible defaults you can tune before deployment.