BL-SD-032: Post-Exhaustion Reversion, Entered Passively

BIJECTIVE LABS

BIJECTIVE LABS . SEP 8, 2026

BL-SD-032: Post-Exhaustion Reversion, Entered Passively

Status: RELEASED

Classification: RESTRICTED

Revision: 1

Date: 2026-09-08

Venue: Hyperliquid perpetuals Programme: I Verdict: OPEN Blocking axis: Power Transferability: not applicable, OPEN verdicts carry no transferability classification

1. Identification

Field	Value
Internal code	BR1
Family	Cascade exhaustion, maker-fill executable reversion
Venue	Hyperliquid perpetuals
Programme	I
Pre-registered	2026-07-14, written before a single result was seen
Verdict issued	2026-07-14, preliminary and event-limited
Pre-registration document	`2026-07-14_BR1_maker_reversion_preregistration.md`
Verdict document	Same document, section 5, ledger row BR1-1
Predecessor	`BL-SD-022`, whose falsified fade thesis this hypothesis attacks from the cost side rather than the mechanism side

This dossier's classification is RESTRICTED. Sections 1, 2, and 8 are stated in substance, because the fact of an open, event-limited hypothesis and the reasoning behind its OPEN verdict belong in this record. Sections 3 through 7 and 9 through 12 are withheld in full under BL-FM-004 section 4, stated openly rather than left blank, because this hypothesis retains live monetisation potential and has not reached a settled verdict.

2. Economic thesis

A prior audit finding, referenced in the asset register as fact A4, measured a real gross reversion on Bitcoin after liquidation-cascade exhaustion, on the order of several basis points, that failed to clear a round-trip cost built from crossing on both legs. The mechanism itself was not in question: the multi-confirmation reversion hypothesis in BL-SD-022 and its momentum counterpart in BL-SD-028 had already established that neither fading nor following a cascade at the moment of exhaustion, entered by crossing, produces a tradeable directional edge. What those measurements had not tested was whether a passive entry, a resting order placed at the best price on the fade side at the moment exhaustion is detected, filled by the very continuation flow that follows rather than by crossing to meet it, changes the cost side of the arithmetic enough to matter. This hypothesis is an attack on execution cost, not a new directional mechanism.

Who is forced to lose. The same structural loser named in the predecessor hypothesis: a wallet whose position is liquidated, whose flow is forced and price-insensitive by construction. What differs here is the execution posture of the counterparty on the other side of that flow, which is the substance withheld under this dossier's classification.

3. Falsifiable claim

Withheld under BL-FM-004 section 4. The pre-registered claim and its kill-criteria table exist in the source pre-registration document and are not reproduced in this public record while the hypothesis remains open and event-limited.

4. Mathematical specification

Withheld under BL-FM-004 section 4. The signal, entry, exit, sizing, and cost specification exists in the source instrument and pre-registration document and is not reproduced here.

5. Operating envelope

Withheld under BL-FM-004 section 4. The parameter table for this measurement is not reproduced here.

6. Instrument

Withheld under BL-FM-004 section 4. The instrument name, test coverage, and statistical method are not reproduced here.

7. Results

Withheld under BL-FM-004 section 4. Result tables are not reproduced here.

8. Verdict, blocking axis, and transferability

Verdict: OPEN. Axis: power. Transferability: not applicable.

An OPEN verdict means the interval produced by this measurement spans its pre-registered threshold and nothing has been concluded, which is not the same finding as a kill and is never recorded as one. Two of this hypothesis's kill criteria were evaluated on the sample available at the time of measurement. The economic criterion, a confidence interval on net return required to lie entirely above zero, did not clear: the sample was event-limited and the resulting interval spans zero. The second criterion, that the event mean must beat a dislocation-matched passive-entry null, did clear, at a significance level worth taking seriously rather than dismissing as a near-miss. Exhaustion-conditioning, in other words, adds real, measurable value over a generic passive fade at this dislocation, even though the economic question at the pre-registered horizon remains unresolved on the sample available.

The binding constraint is sample size, not a sign reversal and not a cost that structurally exceeds the edge. The event class this hypothesis draws on accumulates at a rate that places the pre-registered target sample size well outside this programme's runway, which is why the verdict is recorded as OPEN rather than pushed toward a premature FALSIFIED or CLOSED classification on an underpowered interval. BL-D-002's discipline on this point is explicit and is honoured here: labelling an underpowered result a falsification poisons every decision that later inherits the label, and this programme has made that mistake once before and does not intend to repeat it.

This hypothesis and its sibling on a different venue, both entries in the OPEN section of the kill ledger, are held at RESTRICTED classification and appear in the ledger by identifier and status only, for the reason stated in section 1: the mechanism this hypothesis measures retains monetisation potential that a full public specification would compromise while the verdict remains unsettled.

9. Cost and capacity analysis

Withheld under BL-FM-004 section 4.

10. Envelope sensitivity and reopening conditions

Withheld under BL-FM-004 section 4. What is stated openly, because it belongs to the verdict rather than to the withheld specification, is the nature of what would resolve this hypothesis one way or the other: additional accumulation of the underlying event class to reach the pre-registered sample target. That accumulation is mechanical and passive; no design change is required to produce it, only time.

11. Reference implementation

Withheld under BL-FM-004 section 4.

12. Evidence register

Withheld under BL-FM-004 section 4.

13. Lessons extracted

The lesson this hypothesis contributes to the record does not depend on the withheld detail. An OPEN verdict, honestly reported as event-limited rather than forced toward a premature kill or a premature success, is itself a discipline worth naming: this programme's own history includes a prior instance of mislabelling an underpowered interval as a falsification, and the rule against repeating that mistake is exercised here on a hypothesis where one of two kill criteria did clear, which made the temptation to round up to a positive verdict correspondingly stronger, not weaker. The verdict was not rounded up.

14. Cross-references

BL-SD-022 for the falsified fade mechanism this hypothesis attacks from the cost side. BL-SD-028 for the falsified momentum counterpart on the same event class. BL-D-002 for the discipline against mislabelling an underpowered interval, exercised in section 8. BL-FM-004 section 4 for the classification and disclosure policy governing this dossier's withheld sections. Ledger row BR1.