Concept

Squeeze Release Direction

Squeeze Release Direction is a Volatility concept. The Library holds 1 implementation — a working definition you can pull into Quant.

Top Squeeze Release Direction indicator

The top custom implementation, built on the original standard Squeeze Release Direction formula.

1 total

Want to trade Squeeze Release Direction? The implementation below is one prompt away from a backtested strategy in Quant.

Squeeze Release Direction preview

Squeeze Release Direction

Indicator

What is Squeeze Release Direction?

A squeeze flags compressed volatility and likely expansion, but it is direction-neutral. Squeeze release direction is the follow-on question: which way the move resolves once the squeeze fires. In the TTM Squeeze convention popularized by John Carter, the release is traded in the direction of a momentum histogram at the moment the bands re-emerge from the Keltner Channel; in the Bollinger Squeeze tradition, the cue is the side on which price breaks out of the bands.

Direction is the unreliable half of the squeeze trade. John Bollinger documented the head fake, where the first post-squeeze move runs one way before the real expansion goes the other, which is why most release rules demand confirmation: a close beyond the band rather than a tick beyond it, a structure break, volume expansion, or agreement from the higher-timeframe trend. The compression itself is the dependable observation; the direction call remains a hypothesis until confirmed.

The direction evidence stacks from several independent sources. The momentum histogram supplies sign and slope, expanding momentum into the fire being the strongest version of Carter's cue; the coil's own geography testifies, since price compressing against the top of its range, riding the basis line's upper side, leans differently than a coil hugging support; BandWidth trajectory times the fire itself; structure adds the skew, compression directly beneath resistance resolving downward more grudgingly than compression atop reclaimed support; and the higher-timeframe trend remains the heaviest single vote, releases with it enjoying the wind, releases against it fighting for every point.

Execution styles split on whether to call the direction at all. The directional camp pre-commits on the evidence stack and demands a confirmation event, a close beyond the band with follow-through, before full size; the agnostic camp brackets the coil with stops on both sides and lets the market self-select, paying for the neutrality whenever a head fake triggers both legs. Timing-focused builds add clocks and filters around the fire, and every style shares the failure protocol: a release that reverses back through the coil is the head fake announcing itself, exit or reverse, never argue.

How to identify the likely release direction

The squeeze is a fact; the direction is an evidence-weighing exercise with a confirmation gate.

  1. 1Confirm the squeeze is active and maturing, by band-in-channel state or a long-lookback BandWidth low.
  2. 2Assemble the directional evidence: momentum histogram sign and slope, which side of the basis price coiled on, and the structure above and below the compression.
  3. 3Weigh the higher-timeframe trend as the heaviest vote, discounting release calls that would fight it.
  4. 4Pre-commit the plan: a directional entry awaiting its confirmation event, or a two-sided bracket that lets the break choose.
  5. 5Manage the release: full size only after a confirmed close beyond the compression with follow-through, and treat a reversal back through the coil as the head fake it is.

How it's calculated

Marks the bar where a Bollinger Band squeeze releases from inside the Keltner Channel and uses the sign of a momentum value to label the expected breakout direction.

UpperBBt=SMA⁡n(Ct)+mbb×StdDev⁡n(Ct)\mathrm{UpperBB}_t = \operatorname{SMA}_n(C_t) + m_{\mathrm{bb}} \times \operatorname{StdDev}_n(C_t)
LowerBBt=SMA⁡n(Ct)−mbb×StdDev⁡n(Ct)\mathrm{LowerBB}_t = \operatorname{SMA}_n(C_t) - m_{\mathrm{bb}} \times \operatorname{StdDev}_n(C_t)
UpperKCt=SMA⁡n(Ct)+mkc×ATR⁡n\mathrm{UpperKC}_t = \operatorname{SMA}_n(C_t) + m_{\mathrm{kc}} \times \operatorname{ATR}_n
LowerKCt=SMA⁡n(Ct)−mkc×ATR⁡n\mathrm{LowerKC}_t = \operatorname{SMA}_n(C_t) - m_{\mathrm{kc}} \times \operatorname{ATR}_n
SqueezeOnt=(UpperBBt<UpperKCt) and (LowerBBt>LowerKCt)\mathrm{SqueezeOn}_t = (\mathrm{UpperBB}_t < \mathrm{UpperKC}_t) \text{ and } (\mathrm{LowerBB}_t > \mathrm{LowerKC}_t)
Releaset=SqueezeOnt−1 and not SqueezeOnt\mathrm{Release}_t = \mathrm{SqueezeOn}_{t-1} \text{ and not } \mathrm{SqueezeOn}_t
Midt=HHn+LLn2+SMA⁡n(Ct)2\mathrm{Mid}_t = \frac{\frac{\mathrm{HH}_n + \mathrm{LL}_n}{2} + \operatorname{SMA}_n(C_t)}{2}
Momt=LinReg⁡n(Ct−Midt)\mathrm{Mom}_t = \operatorname{LinReg}_n(C_t - \mathrm{Mid}_t)
Directiont=up if Momt>0, down if Momt<0, read on the bar where Releaset is true\mathrm{Direction}_t = \text{up if } \mathrm{Mom}_t > 0 \text{, down if } \mathrm{Mom}_t < 0 \text{, read on the bar where } \mathrm{Release}_t \text{ is true}
C_t: close price at bar t
t: bar index
n: shared lookback length for all averages (default 20)
m_bb, m_kc: Bollinger and Keltner multipliers (defaults 2 and 1.5)
SMA_n, StdDev_n, ATR_n: simple moving average, standard deviation of close, and average true range over n bars
UpperBB_t, LowerBB_t: Bollinger Band upper and lower bounds
UpperKC_t, LowerKC_t: Keltner Channel upper and lower bounds
HH_n, LL_n: highest high and lowest low of the last n bars
SqueezeOn_t: true while both Bollinger Bands sit inside the Keltner Channel
Release_t: true on the first bar the squeeze condition turns off
Mid_t: reference midline, average of the Donchian midpoint and the SMA
LinReg_n: endpoint value of a linear regression fitted to the last n values
Mom_t: squeeze momentum histogram value
Direction_t: breakout direction label on the release bar

This follows John Carter's TTM Squeeze convention; the widely used open-source ports apply the same conditions.

Some platforms build the Keltner Channel or the momentum midline from an EMA instead of an SMA, or average plain range instead of true range, which shifts squeeze timing slightly.

The label only reads the momentum sign; many traders also require the histogram to keep expanding for a few bars after release.

How traders use it

  • As a momentum-sign rule: when the squeeze fires, enter in the direction of the momentum histogram and stay while momentum keeps expanding, exiting as it wanes.
  • As a breakout rule: trade the side on which price exits the compression range, with confirmation filters to avoid the head fake, a squeeze-specific form of false breakout.
  • As a bias overlay: pre-assign the expected release direction from the higher-timeframe trend and take only the releases that agree with it, skipping counter-trend fires.
  • As a bracket alternative: entry stops on both sides of the coil with cancel-one logic sidestep the direction call entirely, accepting double-trigger whipsaw risk as the price of neutrality.
  • With a head-fake protocol: a release that fails and reverses through the compression zone converts to an exit or a stop-and-reverse, since the failed first move is itself the classic prelude to the genuine expansion opposite.

Release direction vs related squeeze concepts

TTM Squeeze: The TTM tool answers when: bands inside the channel, squeeze on, bands out, fired. Release direction is the follow-on question the tool's momentum histogram was bolted on to answer, and the histogram's sign at the fire remains that tradition's canonical cue.

Bollinger Squeeze: The band tradition detects compression from BandWidth's own history and reads direction from the breakout side, with Bollinger's head-fake warning attached. Same two-part structure, different machinery: internal width statistics versus channel comparison, break side versus momentum sign.

Volatility Contraction Pattern: The campaign-scale cousin: weeks of successively tighter pullbacks inside a base, where the direction question is largely pre-answered by the base's structure and the buy point sits at its boundary. Squeeze releases ask daily what the VCP asks once per base.

Concept family

Volatility

57 concepts mapped · 57 in the Library

Squeeze Release Direction FAQ

How do you know which way a squeeze will break?

You don't with certainty. The momentum sign at the fire, the prevailing higher-timeframe trend, and the structure around the range each provide a bias, and combining them helps, but head fakes are common enough that most traders wait for a confirming close or structure break before committing, and keep a stop in case the release reverses.

What is a head fake after a squeeze?

John Bollinger's term for the initial post-squeeze move that breaks one way, draws traders in, then reverses into the true expansion in the opposite direction. It is the main failure mode of release-direction rules and the reason many systems require a confirmed close beyond the band or a follow-through bar before entering.

Should the direction come from momentum or the breakout side?

The two traditions each have a case. Carter's momentum rule fires at the moment of release with the histogram's sign, earlier and more exposed to head fakes; the Bollinger tradition waits for the band-side break, later and confirmation-flavored. They agree most of the time, and their disagreement is itself information, a fire whose momentum and break side conflict being exactly the release to treat with suspicion or skip.

What counts as sufficient confirmation?

The working ladder: a close beyond the compression rather than an intrabar poke, a follow-through bar extending it, expansion in range or volume backing the move, and ideally a structure event, a swing or level breaking in the release direction. Each rung filters more head fakes and costs more entry price. Where a trader stops on that ladder is a temperament decision, but the first rung, close not tick, is close to consensus.

Is bracketing both sides better than calling the direction?

It is a different bargain. The bracket never argues with the market and catches every genuine release, while paying double whenever a head fake triggers one side and reverses through the other; the directional call avoids that whipsaw tax and misses or fights the fires it reads wrong. Choppy instruments punish brackets, cleanly trending ones reward them, and the honest choice comes from the instrument's own head-fake frequency.

Do release-direction rules work on every timeframe?

The mechanics compute anywhere, but the signal-to-noise follows the compression's scale: daily and multi-hour squeezes resolve with follow-through worth trading, while very low timeframes fire constantly and head-fake proportionally more, their coils being microstructure as often as accumulation. The higher-timeframe vote also gets heavier as you descend, a five-minute release against the daily trend being the configuration that costs beginners the most.

Turn Squeeze Release Direction into a trading strategy.

Take the implementation from this page into Quant, then build on it, backtest it on real data, and keep refining it with AI.