The same five questions every week, so the answers are comparable. This week the market structure question has a clear answer: leverage was added into a drawdown, not removed. Global derivatives open interest rose 16.14% in 24 hours to $479.65bn on 16 September, and the market then rallied +5.6%.
The defining structural fact of the week is the direction of open interest against the direction of price. On 16 September, with bitcoin at $74,944 and the Senate having just failed to advance the CLARITY Act, global derivatives open interest reached $479.65bn - up 16.14% over 24 hours. Bitcoin derivatives alone accounted for about $51.96bn. Traders did not cut exposure into the sell-off; they added to it.
That is a materially different structure from de-risking. When open interest falls with price, the leverage has been removed and the market is lighter afterwards. When it rises, the book is adding into a losing side - which produces the sharp liquidation cascades that define the 16 September flush ($224.81m, 84.6% longs) and, symmetrically, the squeeze that follows when price reverses.
The composition matters as much as the direction. Bitcoin derivatives alone accounted for about $51.96bn of the $479.65bn global book. In a spot-led selloff, open interest would shrink as margin calls force closures; instead the book grew, which means new positions were opened to absorb selling and existing longs did not liquidate voluntarily. That is the mechanical reason the CLARITY-day flush reached $224.81m with 84.6% longs: the market was already carrying more leverage than the spot order book could clear at the prior price.
| Window | Total | Composition | Detail | Reported by |
|---|---|---|---|---|
| 24h to 16 September (CLARITY flush) | $224.8m | $190.10m longs (84.6% of the window) | bitcoin-only reading: $143.2m liquidated | Coinglass via BingX; CoinsProbe |
| Week to 20 September, bitcoin perps | $193m | longs only | both sides forced: a midweek flush, then a post-Fed rally | The Assets |
The same structure two weeks in a row would read very differently. Two weeks ago the market de-risked into month end: open interest fell as price softened, which left the book lighter and the recovery sharper. Last week the book grew into weakness and then kept growing. That persistence is what makes this week's $479.65bn reading a structural event rather than a one-day anomaly. The market is carrying more leverage at a higher price than it was at the prior week's low.
Bitcoin dominance helps explain the composition. At 58.8%, bitcoin is more than half of total crypto market value, and its derivatives book is the largest single source of cross-margin risk. When the global OI number rises, it is largely a bitcoin story; when altcoin OI rises faster than bitcoin OI, it is an appetite story. This week was the former, which is why the $224.81m liquidation was so concentrated in the bitcoin book.
Funding is the week's most reassuring number. Bitcoin perp funding sat at +0.0044% per 8 hours on 16 September, about 4.82% annualised, against a neutral anchor of 0.01%. Ether funding moved from +0.004% per 8 hours midweek to +0.063% by 17 September. For a week in which bitcoin rose +5.6%, funding never approached the levels that would indicate crowded leveraged long demand.
That combination - price up, funding flat - points to spot-led demand. A rally carried by leverage shows up as funding paying longs to stay long; a rally carried by spot shows up as funding sitting at the anchor while price climbs. This week was the second kind, which is why the move did not come with the usual funding blow-off.
Ether funding followed a different path. It opened the week at +0.004% per 8 hours and climbed to +0.063% by 17 September, even as bitcoin funding stayed close to its anchor. The divergence is consistent with ether's larger weekly gain (+6.3%) and with the re-staking and ETF-flow narratives that absorbed attention after the Fed decision. For now the premium is small enough that it does not flag an over-levered long book, but it is the series to watch if the rally extends.
| Horizon | Cost | As % of notional |
|---|---|---|
| 1 day | $13.20 | 0.01% |
| 1 week | $92.40 | 0.09% |
| 1 month | $396.00 | 0.40% |
| 1 quarter | $1,204.50 | 1.20% |
The term structure of funding is as important as the spot level. A flat 8-hour rate through the week means the marginal long is not paying a growing premium to hold the position overnight. Compare that to the typical late-stage rally, where funding climbs each session and the annualised rate moves toward double digits. The absence of that term structure is what keeps the desk from reading this as a leverage-driven move, even though OI itself is high.
Stablecoin supply is the other side of the same read. Total stablecoin supply held at $303.0bn this week, with USDT at $183.3bn and USDC at $74.2bn. A leveraged rally often shows up as stablecoin supply shrinking as cash is deployed into margin; a spot-led rally often shows up as supply flat or rising as new capital enters. The stable read +0.02% supports the funding read: cash is neither being rushed into margin nor fleeing the system.
The venue question changed shape this week. Payward confirmed it will deploy on-chain perpetual futures markets for US clients on Hyperliquid's HIP-3 builder-deployed markets, with Bitnomial, Payward's CFTC-regulated exchange as the deployer and NinjaTrader Clearing carrying accounts, restricted to allowlisted users onboarded through a Payward broker. That is the first concrete regulated-onchain perpetual structure to move from talks to a named plan.
Two numbers frame the trade-off. The wrapper obliges a 500,000 HYPE stake - $46m at the market price - and Payward brings 6.6 million funded accounts. The trading engine stays on-chain; the regulatory, clearing and custody layers sit on top. Whether that combination wins share from offshore venues is the question the next two quarters will answer.
The economics of the wrapper turn on two numbers that rarely move together: the stake and the account base. If only a small slice of the 6.6 million funded accounts trades the product, the stake is a sunk cost. If a large slice does, Payward becomes a meaningful share of Hyperliquid's permissioned volume and a recurring fee source. The validator-revocation clause means the protocol can punish poor risk management, which aligns the deployer's incentives with the chain's rather than with short-term revenue.
| Reading | Value | Note |
|---|---|---|
| On-chain perp volume, 16 September | $25.80bn | the weekly low, the day the CLARITY vote failed |
| On-chain perp volume, 19 September | $26.90bn | the weekly high, into the weekend |
| Weekly range | $147bn - $152bn | DeFiLlama's on-chain perps cohort, seven days |
The competitive question is whether this wrapper grows the pie or shifts the slices. Offshore perpetual volume was estimated at $90 trillion in 2025, up from $28 trillion in 2023. If the HIP-3 structure captures even a single-digit share of that stock it becomes meaningful to Hyperliquid's fee economics; if it merely moves existing offshore traders behind a regulated front, the protocol gains revenue but the market gains no new risk takers. The first scenario is bullish for HYPE demand; the second is a redistribution story.
The regulatory asymmetry is what gives the wrapper a chance. While Congress stalled on the CLARITY Act, the SEC issued a five-year Innovation Exemption and the CFTC published a no-action position for passive trading software. That combination lets Payward operate in a defined safe harbour without waiting for a statute. The risk is that the safe harbour expires or is challenged before the product reaches scale.
Both sides of the book took damage inside one week, and the order matters. The CLARITY flush on 16 September removed 84.6% longs from the book at $224.81m. The recovery into the weekend then squeezed anyone short the $82,000 level: bitcoin ran from $74,944 to a weekly high of $81,911.
The risk into the next week is the mirror of the opportunity. Open interest is high, funding is neutral, and price sits at the top of its range. That is a market with room to run on a squeeze - and with a large, freshly added long cohort that would face forced selling on any failed retest of the breakout.
The weekly liquidation total was $193m of longs, but the CLARITY-day event accounted for $190.10m of that. A single session absorbed roughly three-quarters of the week's long liquidations, which is why the recovery was so sharp: the weak hands had already been removed. The same structure works in reverse at the top of the range, which is why any failed retest of $82,000 is the most important short-term trigger.
Open interest quality matters more than its absolute level. A book that added into the 16 September low is different from a book that chased the weekend breakout. The first cohort has a better cost basis and more cushion; the second is more sensitive to any reversal. Because the global book closed the week at $479.65bn, the second cohort is the one most exposed into the next window. The desk treats the weekend high as a breakout to validate, not a level to chase.
The altcoin liquidation picture was quieter, which is itself a signal. When bitcoin falls and altcoin OI rises, it usually means traders are rotating leverage into higher-beta tokens. That rotation did not show up this week: the altcoin book was stable relative to bitcoin, and the funding divergence between BTC and ETH was modest. The absence of an altcoin leverage surge is another reason the desk reads the move as spot-led rather than leverage-led.
All three revolve around a single question: whether the rally is built on spot absorption or on leverage. Spot absorption shows up as rising price, flat funding and modest ETF inflows. Leverage shows up as rising open interest, funding premiums and large ETF-driven moves. This week matched the first profile, which is why the desk treats any failed retest of $82,000 as the main risk into the next window.
The base case is that the breakout holds but does not accelerate. Funding is flat, stablecoin supply is steady, and ETF flows are positive but not extraordinary. Those are the conditions for a slow grind higher, not a vertical extension. If funding starts to climb, or if ETF inflows spike above the $6.21m weekly pace, the read would shift toward a leverage-driven leg - and the liquidation risk would rise with it.
It means leverage was added into a losing side rather than removed from the market. Global derivatives open interest reached $479.65bn on 16 September, up 16.14% over 24 hours, while bitcoin printed $74,944. That structure produces sharp liquidation cascades in both directions, and the market delivered the first of them the same day.
No. Bitcoin perp funding sat at +0.0044% per 8 hours - about 4.82% annualised - against a 0.01% neutral anchor. Price rose +5.6% with funding at the anchor, which points to spot-led demand rather than crowded leveraged long positions.
Payward will deploy on-chain perpetual futures markets for US clients on Hyperliquid's HIP-3 rails. Bitnomial, Payward's CFTC-regulated exchange creates, owns and administers the markets and settles contracts; NinjaTrader Clearing carries client accounts. Only allowlisted accounts opened through a Payward-registered broker can trade. HIP-3 requires a 500,000 HYPE stake, about $46m at the market price.
$224.81m across bitcoin derivatives, with 84.6% of it longs, and a bitcoin-only reading of $143.2m. Bitcoin fell to $74,944, a three-week low, before reversing (Coinglass via BingX; CoinsProbe).
A high open-interest book sitting at the top of its range. Price broke $82,000 late in the week on modest ETF support of $6.21m. A failed retest of that level would force the freshly added long cohort from the 16 September flush to sell into a market that no longer has the same bid.