This week on Hyperliquid: 2026 week 38
In week 38, 2026 (14 Sept 2026 to 20 Sept 2026), $414M of Hyperliquid positions were liquidated in 46,938 liquidations, 40% of it longs; the largest day was 15 Sept 2026 with $116M, and the largest single liquidation $18.4M on BTC.
Liquidated per day
- Longs
- Shorts
Largest hour: 2026-09-18 13:00 UTC with $54.8M.
Most concentrated markets
The top 5 markets took 76.7% of the liquidated USD of the week.
Largest single liquidation
$18.4M on BTC short at 2026-09-16 18:08 UTC.
Funding extremes
| Market | Per hour | Annualised |
|---|---|---|
| PURR | +0.0036% | +31.2% |
| CASHCAT | +0.0035% | +31.1% |
| PONS | +0.0034% | +30.2% |
| TRX | −0.0011% | −9.8% |
| ADA | −0.0001% | −1.1% |
| TRUMP | 0.0000% | −0.2% |
Open interest change
| Market | Start → end | Change |
|---|---|---|
| kPEPE | $16.9M → $49.4M | +191.4% |
| ARB | $14.6M → $37.3M | +156.4% |
| ENA | $42.9M → $96.9M | +125.7% |
| AVAX | $16.2M → $36.5M | +125.7% |
| NEAR | $143M → $316M | +121.3% |
← 2026 week 37 · 2026 week 39 → · The same week across every venue
Method
- Week: ISO 2026-W38, 14 Sept 2026 to 20 Sept 2026, Monday 00:00 to Sunday 24:00 UTC, Hyperliquid only.
- Liquidations: hourly sums per coin and side from the stored liquidations (every account, from the Hyperliquid fill record), single events for the largest one.
- Funding: average of the stored hourly rates over the week, markets with at least $10M average open interest. Open interest: earliest and latest stored sample of the week per market; the change table lists markets with at least $20M at either end.
- Cluster hit rate: clusters the scoreboard recorded this week; one counts once it is 24 hours old, as reached when the mark touched it within 24 hours.
Cite this data
Source: Blockmetric (blockmetric.io), Hyperliquid data note, 2026-W38, retrieved 22 Nov 2026, 01:28 UTC, https://blockmetric.io/research/weekly/2026-w38
Free to use with attribution and a link to the page. How to cite
Data as of 2026-11-22 01:28 UTC. Source: Blockmetric Research. · Report incorrect data