Public records contradict Solana Research Institute’s $18B claim

Public exchange and blockchain data produce totals that differ from an $18 billion liquidation estimate tied to the Oct. 10, 2025 crypto market crash.

The Solana Research Institute renewed a July letter in an Aug. 14 post that reported roughly $18 billion in liquidations over a 14‑hour period and cited a $3.21 billion spike in one minute. A six‑exchange analysis from data provider Amberdata identified the same $3.21 billion one‑minute peak at 21:15 UTC and attributed 93.5% of that minute’s activity to forced selling. Amberdata’s total for the 14‑hour window was $9.89 billion, with $6.93 billion concentrated in the 40 minutes from 20:50 to 21:30 UTC. A European regulator separately referenced market estimates near $19 billion for the full day; those figures reflect different scopes and timeframes.

The Solana Research Institute letter does not list a common set of trading venues or an aggregation method that would reconcile its $18 billion figure with venue‑level samples. Exchange reports, multi‑exchange samples and day‑wide market estimates each cover different pools of trading activity and time ranges.

Binance published a postmortem that describes operational effects during the event. The exchange reported that its spot and futures matching engines and API trading kept running, while some internal modules malfunctioned after 21:18 UTC. Internal transfers and Earn redemptions experienced delays, and local collateral prices for assets such as USDe, BNSOL and WBETH diverged after 21:36 UTC. Binance identified two compensation batches for users affected by those price dislocations that together totaled about $283 million. A European regulator stated that Binance’s use of internal collateral pricing contributed to local price depegs and amplified forced selling.

On‑chain venues recorded distinct events. A public reconstruction of Hyperliquid’s activity found roughly $2.10 billion in auto‑deleveraging across 34,983 ADL executions in about 12 minutes. The reconstruction is not peer reviewed and uses publicly available ledger data; it documents large‑scale ADL on an on‑chain derivatives platform, a mechanism that reduces profitable traders’ positions when liquidations and risk buffers are insufficient.

Lending markets also logged stress. Protocol records and a Chaos Labs report showed some Aave markets experienced five‑block delays in price updates. Reported outcomes for Aave included about $180 million liquidated and roughly $500,000 in bad debt or expected deficit. Chaos Labs estimated that liquidation fees and SVR revenue left the protocol approximately $1.5 million net positive after those deficits.

Taken together, venue disclosures and on‑chain records provide itemized figures for specific mechanisms: ADL on Hyperliquid, module delays and collateral pricing dislocations on Binance, and oracle delays and lending losses on Aave. Centralized exchanges have so far supplied limited event‑specific ADL totals, and some reconstructions remain the only source for certain on‑chain outcomes.

Regulatory steps since the crash include a UK cryptoasset framework finalized in June 2026. The framework requires qualifying UK trading platforms and principal dealers to publish post‑trade information as close to real time as possible and no later than one minute, and it imposes pre‑trade transparency rules on larger operators. The rules apply to DeFi activity when there is a clear controlling person that carries out regulated activity. The framework does not mandate standardized cross‑venue reporting of liquidation volumes, ADL use or backstop losses.

Available documents show differences in scope, timing and reporting methods across sources. The Solana Research Institute’s letter references discussions with the UK regulator and the Solana Foundation, but the material released with the letter does not include independent confirmation from the regulator. The Oct. 10 event remains a case where venue reports and public data offer multiple, nonidentical measures of liquidation and loss allocation.