Liquidation Intelligence Archive

Post-Mortem Database

Every significant liquidation cascade documented with pre-event LSI conditions, trigger classification, cascade mechanics, and recovery analysis. The only archive that maps what the signal detected, what it didn't — and precisely why.

Category A · 6 events

Derivatives-Driven

LSI is the primary early-warning tool. These events originate in accumulated leverage. Consistently preceded by elevated derivatives stress in all documented cases.

Category B · 4 events

External Trigger + Derivatives Amplification

LSI detects the loaded weapon. It cannot predict the external finger on the trigger. Severity gap explained by the amplifier.

Category C · 3 events

Off-Chain Structural Failure

LSI blind spot by design. These events did not originate in derivatives. LSI correctly read the derivatives layer. The risk was elsewhere.

Event Index

14 events

#	Event	Coverage	LSI Pre-Signal	Category	Phase	Liquidations	Recovery
003	Black Thursday Mar 12–13, 2020	Detected	~68 ~88th · persist	Cat A	Bull Mid	$1.8B / 16h	8 days
012	BitMEX Flash Crash Sep 25, 2019	Detected	~52 ~65th	Cat A	Range	$600M / 2h	Hours
013	Bear Market Capitulation Nov 14–25, 2018	Detected	~55 ~70th · est.	Cat A	Bear Mid	$750M / 24h	24 months
008	Silent ATH Reversal Nov 10 – Dec 4, 2021	Detected	~74 ~92th · persist	Cat A	Bull Top	Multi-$B / weeks	28 months
004	Multi-Trigger Wipeout May 12–19, 2021	Detected	~88 ~98th · persist	Cat A	Bull Top	~$9B / 7 days	6 months
007	The Great Deleveraging Oct 10–11, 2025	Detected	~91 ~99th · persist	Cat A	Bull ATH	$19.3B / 24h ← all-time record	Partial
006	Yen Carry Flash Crash Aug 2–5, 2024	Amplified	~58 ~71st	Cat B	Bull Mid	$1.05B / 18h	3 days
009	DeepSeek Flash Crash Jan 27–28, 2025	Amplified	~63 ~78th	Cat B	Bull Mid	$855M / 24h	~1 week
010	Post-October Bleed Nov 19–24, 2025	Amplified	~54 ~66th	Cat B	Bear Early	$1.9B / 4h	Ongoing
011	2026 Bear Continuation Jan 28 – Feb 11, 2026	Amplified	~38 ~29th	Cat B	Bear Mid	$5B+ / 14 days	Ongoing
001	LUNA Death Spiral May 7–13, 2022	Out of Scope	~22 ~14th	Cat C	Bull Top	~$1B deriv / $400B cap	Never
002	FTX Exchange Collapse Nov 2–11, 2022	Out of Scope	~24 ~18th	Cat C	Bear Mid	$700–800M acute	Never
005	CeFi Contagion Jun 12–27, 2022	Out of Scope	~28 ~21st	Cat C	Bear Mid	$1T+ cap destroyed	18 months
014	Thin Market Non-Event Mar 23, 2026	Detected	34–39 Building · low pct	Diagnostic	Range	No cascade	Contained

Detailed Post-Mortems

Category A — Derivatives-Driven

Events originating in accumulated leverage. LSI is the primary early-warning tool. Derivatives stress was structurally elevated before each documented event in this category.

Event 003

Black Thursday

Mar 12–13, 2020BTCUSDT$10,500 → $3,782

LSI: High Cat A Bull Mid

Liquidations$1.8B in 16 hours

BitMEX share~$1.4B (77% of total)

Trigger typeMacro / External — COVID-19 global risk-off

OI at T-24hAll-time high at the time

Funding ratePersistently positive for weeks prior

Market biasExtreme long skew — halving narrative dominant

Cascade depth3 distinct liquidation waves

InfrastructureBlockchain congestion + exchange outages mid-event

Collateral loopBTC-margined positions: collateral fell with positions — self-liquidating

Price low$3,782

Recovery8 days

Post-eventFull deleveraging. Clean baseline. Bull market resumed.

Key Lessons

The reference event for LSI validation. Strongly positive funding + ATH OI + extreme long-skew + concentrated halving narrative = maximum structural fragility before any external trigger arrives.
BTC-collateral creates a self-liquidating feedback loop: falling collateral price forces additional liquidations, which push price lower, which forces more liquidations.
Infrastructure failure — blockchain congestion, exchange outage — is part of the event mechanics, not incidental. Market fragility and platform fragility manifest simultaneously under peak stress.

LSI Signal HIGH reading reflected correctly. Cascade magnitude was proportionate to the pre-event structural condition. This event is the primary validation case for the LSI framework.

Event 012

BitMEX Flash Crash

Sep 25, 2019BTCUSDT$9,300 support break

LSI: Medium Cat A Range

Liquidations$600M in 2 hours

Volume$2.5B traded in 2-hour window

Trigger typeSupport break — $9,300 level held 3 months

Long composition90%+ of liquidations were long positions

Exchange dominanceBitMEX controlled price discovery — its liquidation engine was the market

RecoveryHours

Structural damageNone. Mechanical flush only.

Historical notePreview of Black Thursday mechanism — same exchange, same dynamics, 6 months earlier

Key Lessons

Single-exchange dominance directly amplifies cascade severity. When one venue controls price discovery, its liquidation engine becomes a self-referential cascade driver with no external circuit breaker.
Support level breaks are the tripwire, not the cause. The months of accumulated long positions at $9,300 were the loaded pressure. The break only released what was already there.
Fast recovery confirmed the LSI read — moderate leverage, no structural damage underneath. The event was a clean mechanical flush.

LSI Signal MEDIUM reading. Crash magnitude and instant recovery were proportionate to a moderate pre-event structural condition.

Event 013

Bear Market Capitulation

Nov 14–25, 2018BTCUSDT$6,300 → $3,128

LSI: Medium (est.) Cat A Bear Mid

Liquidations$750M / 24h — BitMEX record at the time

Trigger typeHash war (BCH fork) + $6,300 support break

Support duration6 months — accumulating stop-losses throughout

Price low$3,128 (Dec 15, 2018)

Hash war effectBCH miners sold BTC treasury to fund war — direct sustained spot pressure

Recovery24 months

Structural damageMedium — bear market continued, miner capitulation extended the bottom

Key Lessons

The longer a support floor holds, the larger the liquidation event when it breaks. Six months of accumulated stop-losses at the same level is six months of compressed energy waiting for a single release.
External sell pressure from miner treasury liquidation (hash war funding) added sustained spot selling that ran for weeks after the initial cascade — extending the damage well beyond the liquidation event itself.
Bear market capitulation events carry longer recovery timelines because the underlying cycle has already turned, not because the derivatives cascade was uniquely severe.

LSI Signal MEDIUM estimated. Historical data from this period is less granular. Cascade magnitude is consistent with moderate pre-event leverage accumulation at a long-established support level.

Event 008

Silent ATH Reversal

Nov 10 – Dec 4, 2021BTCUSDT$69,076 ATH → 28-month bear market

LSI: High Cat A Bull Top

ATH$69,076 — Nov 10, 2021

Trigger typeNarrative exhaustion — no single catalyst

Failed retestsMultiple — each failure confirmed buyer exhaustion

BITO ETF effectAnticipated catalyst exhausted demand immediately on launch — the event became the top

LiquidationsMulti-billion over several weeks

Bear market low$15,480

Recovery28 months — longest in this archive

Bear length12 months of declining structure

Key Lessons

Slow tops produce longer bear markets than sharp crashes. No single trigger means no cathartic liquidation event — the unwinding is gradual, recovery is extended accordingly.
High LSI + failed ATH retest is the maximum risk configuration. The structural condition is fully loaded; repeated failure to advance is confirmation that new buyers are exhausted.
Anticipated catalysts that become tops are structurally more dangerous than surprise triggers — the crowd has already positioned for the catalyst, already leveraged, already the exit liquidity when it arrives.

LSI Signal HIGH reading correctly reflected maximum structural fragility at a cycle top. The absence of a sharp initial catalyst is consistent with narrative exhaustion rather than external shock.

Event 004

Multi-Trigger Wipeout

May 12–19, 2021BTCUSDT$58,000 → $29,300

LSI: Extreme Cat A Bull Top

Liquidations~$9B across 7 days

Triggers3 separate events in 5 days — Musk/Tesla reversal, China mining ban, US regulatory signals

Crypto-margined OI70% at peak → 38% after — permanent structural shift

Dip-buy attemptsEach attempt invalidated by the next headline within days

Double leverageBTC-margined positions: collateral fell simultaneously with positions

Mining impactChina ban → sustained spot selling from relocating miners for months after

Price low$29,300 (initial)

Recovery6 months to reclaim prior levels

Key Lessons

EXTREME LSI entering a multi-trigger cluster is the most destructive combination in this archive. Each trigger that would have been survivable in isolation became catastrophic because structural fragility was already at maximum.
Crypto-margined OI creates double leverage — when the collateral falls, liquidation thresholds are reached faster, accelerating the cascade non-linearly.
The permanent reset of crypto-margined OI from 70% to 38% shows that structural damage can outlast the price recovery. The market rebuilt, but on a different — healthier — foundation.

LSI Signal EXTREME reading. The largest pre-2025 liquidation event arrived at the archive's highest pre-event LSI classification. Signal-to-outcome correspondence was direct and unambiguous.

Event 007

The Great Deleveraging

Oct 10–11, 2025BTCUSDT$122,574 → $104,782

LSI: Extreme Cat A Bull ATH

Liquidations$19.3B in 24 hours — all-time record (9× prior record)

TriggerTrump 100% China tariff via Truth Social, after equity market close

Total OI pre-event$235.9B — all-time high

Funding rate~30% annualized

Flash point$3.21B liquidated in 60 seconds at 21:15 UTC

Accounts liquidated1.62 million

ADL activationAuto-deleveraging broke market maker neutrality — liquidity thinnest since 2022 persisted through Q4

USDe depeg$0.65 on Binance during peak cascade

Altcoin damageATOM near-zero wicks. Major alts −40% to −70%.

RecoveryPartial. ADL reform became legislative priority Q4 2025.

Key Lessons

$235.9B total OI at 30% annualized funding, after-hours trigger, and ADL activation combined into a single 24-hour event 9× larger than any prior record. The magnitude was structurally predetermined — the trigger was incidental.
ADL activation breaks market maker delta-neutrality mid-cascade, removing the primary liquidity source at the exact moment it is most needed. This is a systemic amplifier that existing exchange architecture had not addressed.
Stablecoin depegs (USDe to $0.65) during peak liquidation indicate that collateral infrastructure stress now occurs simultaneously with market stress — a structural feature of the post-ETF institutional era.

LSI Signal EXTREME reading on $235.9B notional OI. The largest liquidation event in history arrived in the exact structural configuration the LSI framework was designed to identify.

Category B — External Trigger + Derivatives Amplification

LSI detects the loaded weapon. Root cause is external. Severity gap explained by the amplifier.

Event 006

Yen Carry Flash Crash

Aug 2–5, 2024BTCUSDTBTC −11% USD / −15% JPY

LSI: Medium Cat B Bull Mid

Liquidations$1.05B in 18 hours

External triggerBOJ unexpected 0.25% rate hike — USD/JPY appreciated ~10% in 3 weeks

Carry trade scale¥40 trillion (~$250B) estimated global unwind

JPY amplifierBTC −15% JPY vs −11% USD — Japanese retail margin call channel quantified

Preview signalJul 24 mini-shock was a 10-day warning — same mechanism at reduced scale

Recovery3 days — fastest in archive

Structural damageNone. FX channel cleared, derivatives layer intact.

Key Lessons

MEDIUM LSI plus a $250B cross-currency carry unwind produced Category A-scale liquidations. When an external amplifier is this large, it multiplies whatever fragility LSI is measuring regardless of the raw score.
The JPY-denominated price differential (−15% vs −11% USD) directly quantifies the Japanese retail margin call channel — a cross-asset transmission mechanism invisible to derivatives-only monitoring.
Preview events at reduced scale are forward-looking signals when the underlying condition persists. The July 24 mini-shock used the identical mechanism 10 days before the full event.

LSI Signal MEDIUM reading correctly sized the derivatives layer. The gap between signal and severity is explained entirely by the $250B FX carry unwind — an external amplifier outside the scope of any derivatives model.

Event 009

DeepSeek Flash Crash

Jan 27–28, 2025BTCUSDT$105,000 → $98,000

LSI: Medium-High Cat B Bull Mid

Liquidations$855M in 24 hours

External triggerDeepSeek R1 release — punctured AI infrastructure narrative

Equity damageNvidia −$600B market cap in a single session

AI tokensAI-linked tokens −20% to −70%

Correlation channelInstitutional allocators held BTC and tech simultaneously — correlated buyer base became correlated seller base

Recovery~1 week

Post-event trajectoryBTC set new ATHs above $124K by July 2025

Key Lessons

The correlated institutional buyer base is also a correlated seller base. When the same allocators hold BTC as part of a technology/AI portfolio, a narrative shock to that thesis drives selling across the entire portfolio simultaneously.
V-shaped recovery confirmed LSI correctly read moderate derivatives fragility. The event cleared quickly because selling was driven by narrative re-rating, not structural insolvency.
Technology narrative shocks are a distinct trigger class in the post-ETF institutional era — not a feature of earlier cycles where crypto retail dominated positioning.

LSI Signal MEDIUM-HIGH reading. Fast V-shape recovery confirmed the derivatives layer was moderately loaded. LSI correctly sized the event within its scope.

Event 010

Post-October Bleed

Nov 19–24, 2025BTCUSDT$95,000 → $81,600

LSI: Medium Cat B Bear Early

Liquidations$1.9B in 4 hours

Institutional triggerBlackRock IBIT single-day outflow $523M — largest on record at that date

BTC/Nasdaq correlation~0.80 — equity selloff transmitted directly

ETF feedback loopOutflows → reduced liquidity → lower price → further outflows

Structural contextOct 10 ADL event had broken market maker neutrality — Nov 21 hit a structurally weakened market

OI changeDropped 35% from October peak before this event

RecoveryOngoing as of archive date

Key Lessons

The ETF-era institutional feedback loop operates differently from retail panic — outflows are systematic (risk model and rebalancing driven), not emotional, and can persist for weeks rather than clearing in hours.
Structural damage from a prior event (Oct 10 ADL activation) leaves the market more vulnerable to the next shock, even when LSI reads MEDIUM rather than EXTREME.
Sequenced events — where Event N weakens the structure that Event N+1 then strikes — require cross-event analysis. Each event in isolation appears smaller than the combined structural damage.

LSI Signal MEDIUM reading reflected the derivatives layer post-October reset. Severity was amplified by institutional ETF outflow mechanics and residual structural damage from Oct 10 — both outside derivatives measurement scope.

Event 011

2026 Bear Continuation

Jan 28 – Feb 11, 2026BTCUSDT$90,000 → $60,033

LSI: Low Cat B Bear Mid

Liquidations$5B+ across 14 days

Macro catalysts7 distinct catalysts in 14 days: FOMC hold, Warsh nomination, Microsoft CAPEX reduction, sustained ETF outflows, BTC below 365-day MA

ETF outflows$6.18B cumulative since November 2025

BTC/VIX correlation0.88 — highest on record

On-chain lossEntity-adjusted realized loss $3.2B on Feb 5 — record

Drawdown from ATH−52% from $126K peak

RecoveryOngoing as of archive date

Bear phase patternLOW LSI confirmed: bear-phase risk travels through institutional channels

Key Lessons

LOW LSI in a bear market does not mean the market is safe — it means derivatives leverage is compressed while risk is concentrated in the institutional/macro layer. The most damaging ongoing bear markets in this archive all carry LOW LSI readings.
BTC/VIX correlation at 0.88 confirms that in the institutional era, crypto drawdowns are increasingly synchronized with broad risk-off events rather than being driven by internal crypto dynamics.
Bear markets attract cascading negative catalysts because each price decline triggers further institutional rebalancing, which creates conditions for the next catalyst to amplify the damage. The sequence is self-reinforcing.

LSI Signal LOW reading was accurate for the derivatives layer, which was genuinely compressed. The $5B+ in damage came through institutional ETF outflows, macro repricing, and VIX correlation — channels that derivatives monitoring does not cover.

Category C — Off-Chain Structural Failure

LSI blind spot by design. Risk was in hidden credit, balance sheet insolvency, or protocol mechanics. LSI correctly read the derivatives layer.

Event 001

LUNA Death Spiral

May 7–13, 2022LUNA / UST$119 → ~$0 / UST depeg

LSI: Low Cat C Bull Top

Liquidations~$1B derivatives / ~$400B market cap destroyed

Root causeAlgorithmic stablecoin design failure — mint-to-defend mechanism was self-defeating at scale

Anchor Protocol19.5% APY sustained by $6M/day subsidy — not yield, a timer

Early signalCurve pool UST imbalance visible on-chain days before public collapse

Information dynamicOn-chain transparency: sophisticated holders who monitored Curve exited before retail

RecoveryNever — LUNA ceased to exist as a functional asset

ContagionDirect trigger for CeFi Contagion (Event 005) — stETH illiquidity and 3AC exposure both downstream

Key Lessons

A 19.5% APY sustained by $6M/day in subsidies is not a yield — it is a countdown. The protocol was structurally insolvent from inception; only the scale of retail inflow determined how large the eventual collapse would be.
On-chain transparency creates information asymmetry in reverse — those monitoring the Curve pool imbalance and on-chain wallet activity exited before the public announcement. The collapse was not sudden for all participants.
LSI correctly read LOW because the derivatives layer was not the source of risk. The failure was entirely within the protocol architecture and economic design.

LSI Signal LOW reading was accurate for the derivatives layer. The risk was in DeFi protocol mechanics and Anchor Protocol's economic structure — outside the scope of any derivatives monitoring framework.

Event 002

FTX Exchange Collapse

Nov 2–11, 2022FTT / BTC / Broad marketFTT $25 → ~$0

LSI: Low Cat C Bear Mid

Liquidations$700–800M acute derivatives liquidation

Root causeExchange insolvency — customer funds misappropriated via Alameda Research

Collateral structureExchange-issued token (FTT) as primary collateral — circular self-collateralization

Trigger typeInformation event — CoinDesk balance sheet leak Nov 2, not a price event

Binance withdrawalCZ announced FTT liquidation Nov 6 — accelerated the bank run

Customer funds$8B+ customer deposits missing — never recovered

RecoveryNever — FTX ceased operations Nov 11, 2022

ContagionBlockFi, Voyager — both rescued by FTX earlier in 2022, both collapsed with it

Key Lessons

Exchange-issued tokens as primary collateral create a catastrophic circular structure. When the loop breaks, the collateral and the exchange fail simultaneously — there is no independent collateral to absorb the loss.
The trigger was information, not price. Calm derivatives markets in early November 2022 were not a clean bill of health — they reflected only the derivatives layer, with no visibility into the off-chain balance sheet.
Entities rescued by a distressed counterparty inherit its distress. The FTX rescues of BlockFi and Voyager transferred concentrated exposure, not capital strength.

LSI Signal LOW reading was accurate — the derivatives layer was not overextended. The risk was in an opaque off-chain balance sheet that had never been audited. No derivatives model could have provided earlier warning.

Event 005

CeFi Contagion

Jun 12–27, 2022Broad marketCelsius → 3AC → Voyager → BlockFi → FTX

LSI: Low Cat C Bear Mid

Capital destroyed$1T+ total market cap across the contagion chain

Root causeInstitutional credit layer collapse — hidden bilateral loan exposure across CeFi entities

Celsius$12B AUM frozen Jun 12 — stETH booked at par, traded at discount

Three Arrows CapitalGhosted margin calls — $10B+ in undisclosed borrowing

Contagion chainLUNA → stETH depeg → Celsius → 3AC → Voyager/BlockFi → FTX (Nov 2022)

stETH problemBooked at par, illiquid at scale — illusory solvency for all entities holding it as primary asset

Recovery18 months for broader market

Entity recoveryCelsius, 3AC, Voyager — never

Key Lessons

Contagion in the CeFi credit layer travels through hidden bilateral loan exposure, not through price. Each entity's balance sheet was opaque. The connections only became visible when the chain began to break.
Illiquid assets booked at par (stETH, locked staking positions) are solvency illusions. When redemptions are demanded, the par-value accounting collapses instantly — not gradually.
LOW LSI in a bear market with known CeFi stress is the highest-risk configuration in this archive for off-chain structural failure. Calm derivatives while the credit layer is collapsing is the signature of a Category C event in progress.

LSI Signal LOW reading was accurate for the derivatives layer. The $1T+ in capital destruction originated in re-hypothecated credit exposure and illiquid asset mismarking — structures invisible to derivatives-based monitoring.

Diagnostic Cases — Correctly Contained

Events where LSI identified an unusual structural configuration and correctly determined no amplification mechanism was present. No cascade occurred. Signal worked as designed.

Event 014

Thin Market Non-Event

Mar 23, 2026BTCUSDT−1.8% then 8h range

LSI: Building Diagnostic Range

LSI at event34–39 · Building zone · low percentile

Conviction86–100 — 10 consecutive hours at extreme levels

Volatility92–97 — extreme reactivity throughout

Crowd alignment14–21 — genuinely thin participation

Pressure0–10 — nobody under stress

Price outcome−1.8% then 8-hour range — contained move

LiquidationsNone significant — no cascade

RecoveryImmediate — within the same session

ClassificationCorrectly detected — thin market, no amplification mechanism

Key Lessons

Extreme conviction and extreme volatility without crowd participation is not a cascade setup — it is a thin market reacting to a small number of large positions. The absence of crowd alignment removes the amplification mechanism that turns OI momentum into forced liquidations.
Low pressure (0–10) with high conviction means positioned participants are comfortable. There is no forced exit risk. The move resolves when the positioning runs out of counterparties, not through liquidation.
LSI in the Building zone with low percentile correctly signalled that structural fragility was absent. The unusual dimension combination — high conviction, high volatility, thin crowd — was the diagnostic signal, not a pre-cascade warning.

LSI Signal Building zone, low percentile. The platform correctly identified the structural configuration as thin-market positioning without an amplification mechanism. No cascade occurred. This case validates the platform’s ability to distinguish between alarming-looking dimension combinations and genuine structural fragility.