Hyperliquid Weekly Data Report — Week of September 14–20, 2026
Every metric below was fetched from a public API — the Hyperliquid official endpoint, DefiLlama, published venue tallies — on September 21, 2026, and the fetch is archived with a timestamp. This report prints only what a public source can verify: where a commonly cited metric has no public API (staking totals, weekly liquidations, per-address activity), the table says so rather than estimating. The week in one line: HYPE closed $93.75, +21.3%, setting a new all-time high at $94.60; protocol fees rose for the first time in five weeks to $19.25M; and platform open interest reached $16.87B.
The verified metrics
| # | Metric | Reading | Source | Note |
|---|---|---|---|---|
| 1 | HYPE price (Sep 20 close) | $93.75 | Hyperliquid candles | +21.3% WoW ($77.26 → $93.75) |
| 2 | Weekly high / low | $94.60 / $75.15 | Hyperliquid candles | High printed Sat Sep 19, low Tue Sep 15 |
| 3 | All-time high (intraday) | $94.60 | Hyperliquid candles | Sep 19; the close sits −0.9% below it |
| 4 | Spot price (Sep 21, ~03:15 UTC) | $93.09 | Hyperliquid candles | −0.7% off the Sunday close; +15.9% on 7 days |
| 5 | Market cap | $20.7B | Derived (222.446M × $93.09) | $20.85B at the $93.75 weekly close |
| 6 | Circulating supply | 222.4M | CoinGecko / venue tallies | Of 955.3M total; 1B max supply |
| 7 | Token volume (24h) | $1.72B | CoinGecko | Sep 19 print; HYPE spot + perp token volume |
| 8 | Perp OI — platform total | $16.87B | Hyperliquid API | Core $12.90B + builder markets $3.97B |
| 9 | Perp OI — core order book | $12.90B | Hyperliquid API | 178 live markets of 234 listed |
| 10 | Perp OI — HYPE book | $2.02B | Hyperliquid API | 21.72M HYPE × mark $93.09; +21% WoW |
| 11 | Perp OI — HIP-3 builders | $3.97B | Hyperliquid API | 23.5% of platform OI (was 28.5%) |
| 12 | Live perp markets | 178 + 147 | Hyperliquid API | Core + builder markets with seeded books |
| 13 | Perp volume (24h) | $6.76B | Hyperliquid API | Core $5.78B + builders $0.98B |
| 14 | HYPE funding (8h) | 0.039% | Hyperliquid API | 0.0000494/hr; flipped from ~0.010% — longs now pay |
| 15 | Protocol fees (7d) | $19.25M | DefiLlama | +49.8% WoW — first rise in five weeks |
| 16 | Protocol fees (30d) | $75.8M | DefiLlama | 7d run-rate annualizes to ~$1.00B |
| 17 | Implied buyback capacity (7d) | ~$19.1M | Derived | ~99% of revenue to AF (third-party estimate) × fees |
| 18 | TVL — Hyperliquid L1 | $6.50B | DefiLlama | Up 3.6% from $6.279B on Sep 14 |
| 19 | TVL — Arbitrum bridge | $0.60B | DefiLlama | Up from $0.43B on Sep 14 |
| 20 | Total TVL (protocol) | $7.10B | DefiLlama | Within 0.1% of the $7.095B published platform-wide figure |
| 21 | L1 chain TVL (on-chain DeFi) | ~$1.39B | Chain data (third-party) | Rank #7, level with Arbitrum — a different construction to row 18 |
| 22 | HYPE burned (cumulative) | ~48.7M | Press tally | ~4.9% of max supply; no single-API ledger |
| 23 | 2026 YTD revenue | ~$429M | Press tally | First among crypto projects; narrower than gross fees (row 15) |
Not verifiable via public API (printed as unavailable, not estimated): staked HYPE and staking APY, HLP vault TVL and return, Assistance Fund weekly USD purchases and cumulative ledger, weekly liquidations, 7-day perp volume, spot 7-day volume, active addresses, transaction count, HIP-3 weekly volume share as a single published series. Rows 22 and 23 are third-party tallies, labeled as such: no public endpoint publishes the burn ledger or a year-to-date revenue series, and they are not interchangeable with row 15.
The five-week chain
| Week | Close | High | Low | Fees (7d) | TVL L1 (close) |
|---|---|---|---|---|---|
| Aug 17–23 | $82.37 | $83.46 | $57.02 | $24.6M | $6.16B |
| Aug 24–30 | $79.99 | $86.73 | $76.59 | $21.2M | $6.43B |
| Aug 31–Sep 6 | $87.94 | $89.60 | $79.52 | $16.5M | $6.41B |
| Sep 7–13 | $77.26 | $88.31 | $76.59 | $12.9M | $6.33B |
| Sep 14–20 | $93.75 | $94.60 | $75.15 | $19.25M | $6.50B |
The shape: a +44% breakout week in late August, two weeks consolidating under the $89.60 record, one full down week at −12.1%, and now a +21.3% week that reclaims the record outright. The fee series traced the same arc down a week behind price and has now turned. The five-week fee total is $94.5M; the five-week price range is $57.02–$94.60. Note the range column: this week printed both the highest high and the lowest low of any week since August — a $19.45 wide week, wider than the −12.1% drawdown week that preceded it.
Inside the week: the daily path
| Date | Close | Δ | Protocol fees (day) |
|---|---|---|---|
| Mon Sep 14 | $80.33 | +4.0% | $2.92M |
| Tue Sep 15 | $76.91 | −4.3% | $3.03M |
| Wed Sep 16 | $78.23 | +1.7% | $2.83M |
| Thu Sep 17 | $85.12 | +8.8% | $2.50M |
| Fri Sep 18 | $92.61 | +8.8% | $3.96M |
| Sat Sep 19 | $92.22 | −0.4% | $1.90M |
| Sun Sep 20 | $93.75 | +1.7% | $2.13M |
Two structural notes on the daily pattern. First, the two +8.8% sessions on Thursday and Friday are 87% of the week's net move, and the week's low ($75.15) printed on Tuesday — so the entire round trip happened inside seven sessions, not as a drift. Second, fee prints did not track price direction: Thursday's +8.8% printed the week's lowest weekday fee ($2.50M) while Friday's equal-sized move printed the week's highest ($3.96M). The fee line responds to how much volume a move generates, not to which way it goes.
The same seven days for BTC and ETH
| Date | HYPE Δ | BTC close | BTC Δ | ETH close | ETH Δ |
|---|---|---|---|---|---|
| Sep 14 | +4.0% | $78,182 | +1.8% | $2,515 | +1.6% |
| Sep 15 | −4.3% | $75,608 | −3.3% | $2,397 | −4.7% |
| Sep 16 | +1.7% | $76,180 | +0.8% | $2,418 | +0.9% |
| Sep 17 | +8.8% | $76,396 | +0.3% | $2,446 | +1.2% |
| Sep 18 | +8.8% | $80,916 | +5.9% | $2,613 | +6.8% |
| Sep 19 | −0.4% | $81,292 | +0.5% | $2,633 | +0.8% |
| Sep 20 | +1.7% | $81,221 | −0.1% | $2,646 | +0.5% |
Read across, the +21.3% week decomposes into one HYPE-specific session and one market-wide one. Thursday Sep 17 is the idiosyncratic day: HYPE +8.8% against BTC +0.3% and ETH +1.2% — roughly thirty times and seven times the majors' moves on the same candles. Friday Sep 18 is the shared day: all three rose hard, though HYPE's +8.8% still outran BTC's +5.9% and ETH's +6.8%. The week's worst HYPE session, Tuesday's −4.3%, was actually milder than ETH's −4.7%. Weekly totals on the Sunday-close clock: HYPE +21.3%, BTC +5.7% ($76,810 → $81,221), ETH +6.9% ($2,476 → $2,646). BTC remains roughly 36% below its October 2025 record, which is the frame for reading this week as a token event inside a milder market week.
The HIP-3 builder census, measured
Builder-deployed markets are measured directly from the API's perpDexs and metaAndAssetCtxs endpoints, summed across every registered dex. This week's census, run Sep 21:
| Venue | Live markets | Open interest | 24h volume |
|---|---|---|---|
| xyz (Trade[XYZ]) | 108 | $3.877B | $0.950B |
| para (Paragon) | 27 | $0.02B | ~$0.00B |
| io (EntropyIO) | 8 | $0.06B | $0.02B |
| mkts (Markets by Kinetiq) | 4 | $0.01B | $0.01B |
| flx, vntl, hyna, km, abcd, cash | 0 | — | — |
| Builder total | 147 | $3.97B | $0.98B |
Ten builder DEXs are registered; five have any market with a seeded book, and one of those five carries ~98% of builder open interest. The top xyz books by open interest: SP500 $429.4M, SKHX $340.0M, GOLD $298.0M, CL $202.3M, SKHY $199.1M, XYZ100 $195.3M, SILVER $177.3M, SPCX $157.8M, BRENTOIL $149.3M, NVDA $141.0M, MU $135.0M, INTC $111.7M. Six venues carry markets with zero open interest — registered, listed, unseeded.
One week-over-week comparison is worth stating precisely, because it has been misread elsewhere: builder open interest fell from $4.12B to $3.97B (−3.6%) while the core book rose from $10.32B to $12.90B (+25.0%), so builder share of platform open interest fell from 28.5% to 23.5%. The share fell because the denominator grew four times faster than the numerator shrank.
Valuation arithmetic (not a forecast)
From verified inputs only. A $20.85B market cap (at the $93.75 weekly close) against the annualized 7-day fee run-rate of ~$1.00B gives a ~20.8x price-to-gross-fees multiple on circulating supply. The 30-day fee base of $75.8M annualizes to ~$0.91B, or ~22.9x. On fully diluted supply of 955.3M, the same market cap is ~$89.5B and the multiple is ~89x.
The comparison with last week is the interesting part: a week ago the market cap was $17.7B against a $0.67B run-rate — 26.4x. This week the market cap rose 17.8% while the run-rate rose 49.3%, so the multiple compressed from 26.4x to 20.8x even as the token made a new all-time high. That is a rare configuration and it is the arithmetic expression of the week's fee story. These are restatements of the table above, not price targets — the multiple is set by the market, the inputs are set by the APIs.
Watching next week
- The September 29 unlock: 14,175,778 HYPE — ~1.4% of max supply, roughly $1.33B at the week's close. Two prior unlocks (Aug 29 and Sep 6) were absorbed without sustained downside, though both arrived into a market with falling fees rather than rising ones.
- October 3, AQAv2's first payment: the date the estimated reserve-yield contribution to the Assistance Fund stops being a projection. Any disclosure of the actual amount resets the blended run-rate.
- Whether fees hold above $16.5M: the W3 print is the reference line. Holding above it frames the last month as a volatility cycle rather than a downtrend.
- Funding: 0.039% per 8h means longs are now paying a real rate to stay long into a record. If funding keeps climbing while price stalls, that is the crowding signature that preceded last week's drawdown.
- Lending utilization: the $269M day-one draw is a launch number. The first read on whether it persists comes from utilization-driven rates once the novelty passes.
Frequently asked questions
Why are some metrics marked "not verifiable" instead of estimated?
Because an estimate without a source is just a number. Staking totals, HLP returns, AF purchases and address counts matter, but no public API publishes them as of this writing. Rather than print plausible-looking values, this report leaves them blank and names them, so readers know exactly what is and isn't known.
Why are there two different TVL figures for Hyperliquid?
They measure different things. Row 18–20 measure the Hyperliquid protocol: liquidity in its own venue plus the Arbitrum bridge, which DeFiLlama tracks on the protocol page and which this site has used every week. Row 21 measures DeFi protocols deployed on the Hyperliquid L1 chain, which is the figure behind this week's "level with Arbitrum, rank #7" claim. Both are DeFiLlama-tracked and both are correct; they differ by roughly five-fold because they count different pools of capital.
Does the +49.8% fee rise change the revenue picture?
It changes the run-rate, not the definitional caveat. Gross fees (row 15) and protocol revenue (row 23's basis) are different lines: DefiLlama separates them, and press figures quoting ~$13.5M of weekly revenue use the narrower measure. At $19.25M of gross fees for the week, the annualized gross run-rate is ~$1.00B; the annualized revenue run-rate is materially lower and is not published as a single series we can verify weekly.
How is the ~20.8x multiple calculated?
Market cap ($20.85B at the weekly close) divided by the trailing-7-day fee run-rate annualized ($19.25M × 52 ≈ $1.00B). It is a price-to-gross-fees multiple, not an earnings multiple, and not a price target. Last week the same construction produced 26.4x; the compression comes from the denominator growing faster than the numerator.
Why did HYPE perp open interest grow 21% in one week?
Because more capital was committed to HYPE-denominated positions as the token rallied — 21.72M HYPE of notional open interest at a $93.09 mark. Open interest is a stock: it measures positions held at a moment, not volume traded during the week. The fact that it grew roughly in line with price means the book scaled rather than levered: position count rose, but not far faster than the underlying.
Is this financial advice?
No. This is data analysis for informational purposes. Nothing here is a recommendation to buy, sell or hold HYPE or any other asset.