Liquidation of positions on crypto exchanges
Monitoring forced position closures in futures markets in real time. These closures show where leveraged traders have succumbed to pressure—and where the market may accelerate.
Calculation completed: 2026-10-21 21:00 UTC. Start of the last 5-minute interval with position liquidations in this calculation: 2026-10-21 21:00 UTC. Windows from 1 to 24 hours are sliding and shift every 5 minutes. Data collected from 13 Cryptocurrency exchanges, a list of which is available in the exchange table below.
Live feed of liquidations
Liquidations by assets (24h)
| # | Assets | 1h | 4h | 12h | 24h | Preponderance | |
|---|---|---|---|---|---|---|---|
| 1 |  BTCBitcoin | $586.2K | $3.4M | $74.4M | $87.2M | 94% S | |
| 2 |  ETHEthereum | $218.7K | $5.2M | $50.2M | $56.2M | 83% S | |
| 3 |  ZECZcash | $19.5K | $2.5M | $10.0M | $15.6M | 93% S | |
| 4 |  DOODDoodles | $4.2K | $190.8K | $6.7M | $8.6M | 68% S | |
| 5 |  IOSTIOST | $165.6K | $1.9M | $6.0M | $8.2M | 61% S | |
| 6 |  SOLSolana | $8.6K | $934.7K | $6.2M | $7.9M | 79% S | |
| 7 | S SNDK.STOCKSNDK.STOCK | $93.8K | $217.0K | $4.3M | $6.4M | 55% L | |
| 8 |  NEARNEAR Protocol | $5.4K | $373.9K | $2.7M | $5.3M | 72% S | |
| 9 | S SPCX.STOCKSPCX.STOCK | $207 | $692.7K | $3.1M | $4.2M | 60% S | |
| 10 |  XRPXRP | $2.3K | $823.5K | $3.2M | $4.1M | 92% S | |
| 11 |  JCTJanction | $1.3K | $49.5K | $696.3K | $2.6M | 70% S | |
| 12 |  ENAEthena | $77.4K | $275.6K | $2.3M | $2.5M | 89% S | |
| 13 | C CL.COMMODITYCL.COMMODITY | $1.1K | $588.0K | $1.4M | $2.5M | 77% S | |
| 14 | S SOXL.STOCKSOXL.STOCK | $150.8K | $257.7K | $851.0K | $2.1M | 63% L | |
| 15 |  PEPEPepe | $1.8K | $116.3K | $1.5M | $1.9M | 75% S | |
| 16 |  PUMPPump.fun | $197.7K | $638.9K | $1.1M | $1.5M | 83% S | |
| 17 |  WLDWorldcoin | $10.7K | $165.0K | $1.2M | $1.5M | 57% L | |
| 18 | S SKHYNIX.STOCKSKHYNIX.STOCK | $0 | $14.8K | $447.7K | $1.4M | 53% L | |
| 19 |  STRKStarknet | $61.9K | $479.1K | $1.2M | $1.4M | 59% S | |
| 20 | K KORU.STOCKKORU.STOCK | $40.2K | $50.6K | $483.0K | $1.2M | 53% S |
Liquidations by exchange (24h)
| Cryptocurrency exchange | Total | Longs | Shorts | Share | Preponderance |
|---|---|---|---|---|---|
Binance USD-M2026-10-21 21:00 UTC | $120.2M | $30.5M | $89.7M | 47.1% | 74.7% S |
OKX2026-10-21 21:00 UTC | $36.1M | $11.1M | $25.0M | 14.1% | 69.2% S |
Hyperliquid2026-10-21 20:40 UTC | $30.8M | $15.3K | $30.8M | 12.1% | 100.0% S |
Bybit2026-10-21 21:00 UTC | $27.4M | $6.3M | $21.0M | 10.7% | 77.0% S |
Gate.io2026-10-21 21:00 UTC | $19.4M | $3.0M | $16.5M | 7.6% | 84.7% S |
Bitget2026-10-21 20:55 UTC | $11.0M | $1.7M | $9.3M | 4.3% | 84.3% S |
HTX2026-10-21 20:40 UTC | $5.5M | $297.4K | $5.2M | 2.1% | 94.6% S |
Aster2026-10-21 20:40 UTC | $2.2M | $1.0M | $1.2M | 0.9% | 54.2% S |
Kraken2026-10-21 20:50 UTC | $2.0M | $659.1K | $1.3M | 0.8% | 66.7% S |
Lighter2026-10-21 20:50 UTC | $410.2K | $23.1K | $387.1K | 0.2% | 94.4% S |
Binance Coin-M2026-10-21 19:00 UTC | $334.4K | $55.2K | $279.2K | 0.1% | 83.5% S |
dYdX2026-10-21 16:35 UTC | $34.2K | $0 | $34.2K | 0.0% | 100.0% S |
Bitfinex2026-10-21 18:25 UTC | $21.9K | $210 | $21.7K | 0.0% | 99.0% S |
The start of the most recent 5-minute interval with position liquidations (UTC) is indicated below the exchange name. All totals on the page are for the exchanges in this table only, not for the entire market.
What are position liquidations and why are they important for Bitcoin-traders?
Liquidation is the forced closure of a position by the exchange when a trader's margin falls below the required level. This happens all the time in the crypto market: a trader opens a long position with 10x leverage, the price drops by 8-10%, and the exchange automatically closes the position. The volume of the liquidated position is added to the order book as a market order, further amplifying the movement.
Forced position closures often occur in waves, one after another. In January 2024, exchanges forcibly closed over $800 million in long positions in a single hour after a false breakout of $49,000. The first wave of forced closures knocked out stops, and the second wave pulled out the last remaining leveraged positions. This is why the market often reverses after large forced position closures: all the weak capital has already been squeezed out.
For a trader, monitoring position liquidations offers two advantages: (1) assessing current pressure—a lot of position liquidations means an overheated market; (2) identifying entry points—after a cascade, the market often forms a local bottom or top.
Interpreting position liquidation data
Mass liquidations of long positions
The sharp decline wiped out leveraged long positions. The cascading effect accelerates the decline. But after the full squeeze, a short-term bottom often forms—weak hands are already out of the market.
Mass liquidation of short positions
A short squeeze is when a price rise forces short position holders to close their positions, accelerating the rally. This is a classic scenario after positive news or a breakout of a key resistance level.
Few position liquidations + high OI
The market is calm, but many positions have accumulated. "The calm before the storm"—open interest could sooner or later turn into a cascade of position liquidations. Be careful with leverage.
Equal liquidations of long and short positions
The market is ranging, with both sides taking hits. A sideways trend with high volatility is a typical consolidation before a directional move.
Data Methodology
We collect forced closes from public derivatives exchange feeds and aggregate them by coin, exchange, and party in dollar terms. Binance, OKX, Bybit, Bitget, Gate.io, and most other exchanges are connected directly. Hyperliquid is fed through CoinGlass, and Lighter is fed through Coinank. Number of crypto exchanges included in the calculation: 13. The full list is in the crypto exchange table above. Exchanges not listed are not included in the totals.
An important disclaimer: exchanges do not publish the full order feed. Binance and Aster only publish the most recent forced close for each symbol within 1000 ms, without the order ID. If multiple forced closes occur for a symbol within a second, the remaining ones are not included in the stream. Without an ID, it is also impossible to guarantee that two snapshots of the same partially executed order are not double-counted. Therefore, our figures are estimates based on the snapshots received, not the full execution log. How to read this data during cascades is explained in analysis of the collapse of October 10, 2025.