Perpetual Futures and the 24/7 Oil Market

Chapter 18 (Futures and Swaps) covers the contracts that have priced oil for forty years: fixed-expiry futures that converge to spot through physical delivery, and the over-the-counter swaps built on top of them. This appendix covers a newer and stranger instrument that reached oil in 2025 and 2026, the perpetual future, or perp. It began in crypto, has no expiry date, and keeps itself pegged to the underlying price not by delivery but by a continuous cash payment between traders. Whether it becomes a serious part of the oil market or stays a speculative sideshow is still an open question, but the largest exchange operators are now moving into it, so it is worth understanding precisely.

What a Perpetual Is

The perpetual swap was invented by the crypto exchange BitMEX, which launched the first one, on bitcoin, in May 2016. Its defining trick solves a problem: without an expiry date, nothing forces the contract price back to the spot price. A conventional future converges because it expires, so on the last day it must equal the thing it settles into. A perp never expires, so instead it uses a funding rate, a small payment exchanged directly between long and short holders at fixed intervals, commonly every eight hours. When the perp trades above the underlying, longs pay shorts, which discourages buying and nudges the price down; when it trades below, shorts pay longs. The payment scales with the gap, so the funding rate acts as a financial spring constantly pulling the contract back toward an external reference.

Two prices run side by side. The mark price is where the contract actually trades on the exchange. The oracle or index price is the external reference, an aggregate of real spot or futures prices, and it is what liquidations and the funding calculation are measured against, so that a brief spike in the contract's own order book cannot wrongly liquidate traders. When the two are close, funding sits near zero and the perp is tightly pegged. Perpetuals are now the dominant form of crypto derivative by volume, and the perpetual swap is arguably the one genuinely new market-structure idea crypto has produced, which is why it is now being ported onto traditional assets, including oil.

How Oil Perps Reached the Market

Oil perps reached scale through Hyperliquid, a decentralized perpetuals exchange that runs on its own blockchain and became the dominant on-chain venue in 2024 and 2025. In October 2025 Hyperliquid activated a framework it calls HIP-3, which lets outside parties permissionlessly launch their own perpetual markets, including on real-world assets like equities and commodities, in exchange for staking 500,000 of Hyperliquid's HYPE tokens, worth very roughly 25 to 30 million dollars depending on the token price, and winning a listing auction. A pseudonymous builder who goes by Shoku deployed a venue styled trade[XYZ] that lists a WTI crude perpetual under the ticker xyz:CL, margined in the dollar stablecoin USDC, and it quickly came to handle around 90 percent of the real-world-asset volume on the network.

The centralized exchanges followed. Binance launched gold and silver perpetuals in January 2026 and WTI, Brent, and natural-gas perpetuals on April 1, 2026, with leverage advertised up to 100 times. The headline activity grew quickly: during the oil volatility of early 2026, the WTI perp on Hyperliquid traded as much as roughly 1.5 billion dollars in a single day, and the whole trade[XYZ] book approached 4 billion. But the depth behind those numbers is thin. One microstructure study found that within two-hundredths of one percent of the mid-price, the CME WTI contract offered roughly 19 million dollars of executable depth against roughly 150,000 dollars on the on-chain venue, more than a hundredfold difference, and that on-chain crude depth was under one percent of the CME's. Much of the reported volume is high-frequency churn, so a commercial hedger still cannot move a real block without heavy slippage.

Why the Clock Is the Real Attraction

The genuine draw for oil is the trading calendar. The CME's energy markets close from roughly Friday afternoon to Sunday evening, and oil's biggest shocks, OPEC decisions and Middle East conflict, do not respect the weekend. After the US and Israeli strikes on Iran in early 2026, with traditional floors shut, oil perps did the price discovery: Hyperliquid's crude volume spiked to nearly 1.7 billion dollars a day, many times its normal level, and the contract had already moved sharply by the time the CME reopened on Monday. On-chain venues have become the marginal price-discovery point for oil on weekends and holidays, not the primary market during normal hours, where they remain a sliver of CME liquidity.

The Institutional Response

The established exchange operators are now moving in, which is what lifts this above a crypto curiosity. Coinbase listed the first US-regulated perpetual-style futures, on bitcoin and ether, in July 2025, cleared through a CFTC-registered exchange with no monthly expiry, and reported more than 211 billion dollars of notional volume in them by mid-2026. In late May 2026 the CFTC went further, approving the first true US perpetual futures, on cryptocurrencies, at the prediction-market platform Kalshi, granting Coinbase’s US customers access to its global perps, and publishing a framework for registered US platforms to launch their own. Kalshi’s crypto perps cleared 8.5 billion dollars of volume within weeks. Crucially, the approval covered crypto only; no oil perpetual is approved for US trading. And Intercontinental Exchange, the owner of the ICE Brent benchmark, took a minority stake in the crypto exchange OKX in early 2026 and, in May 2026, announced ICE Brent and ICE WTI perpetual futures to be operated by OKX under license to ICE’s benchmark prices. ICE’s chairman Jeffrey Sprecher, who had built the regulated bitcoin venue Bakkt back in 2018, told investors he had met the Hyperliquid team several times, called the venue bigger than NASDAQ, and pressed regulators with the question of why established exchanges should be barred from offering perpetuals when the activity is already happening offshore.

The reaction showed how seriously the incumbents take the threat. In the weeks after the CFTC’s approval, shares of the big exchange operators sold off hard: Cboe fell about 26 percent, CME about 12 percent, ICE about 11 percent, and Nasdaq about 9 percent. CME sued the CFTC in June 2026, alleging that letting Kalshi list perps violated federal law and inflicted “textbook competitive injury,” while Kalshi, Coinbase, and the CFTC cast the suit as an attack on competition. The rest hedged their bets: Cboe weighed converting its long-dated bitcoin and ether futures into true perpetuals, Nasdaq said it was waiting for clarity, and ICE’s operating chief noted that “not a single customer has asked us for perps” but that the firm would offer them if institutional demand appeared. The leverage on offer, though, splits sharply by venue: the offshore platforms advertise more than 100 times, while the regulated US versions are far tamer, Kalshi around six times, comparable to a traditional future, and Coinbase up to twenty.

Why Oil Has Not Gone Perpetual, and the Skeptic's View

WTI on the CME has stayed a fixed-expiry, physically delivered contract for a reason. Its monthly expiries, its delivery point at Cushing, and its position limits are exactly what force the paper price to track real barrels: a trader who holds to expiry must make or take delivery, and that obligation anchors the future to the cash market. A perpetual has no expiry and no delivery, so it cannot perform that physical convergence at all.

For all the momentum, the perpetual remains, for now, a speculative instrument rather than a hedging one, and several cautions follow from its design. It settles in a stablecoin, not in barrels, so it cannot anchor to oil the way a deliverable contract does. Its funding rate is an incentive, not an arbitrage, and under one-sided pressure it can blow out without dragging the price back into line. Its weekend pricing leans on the venue's own order book rather than an outside feed, which is a manipulation surface that a delivery-backed contract does not have. Its headline volume is not the same as executable depth, much of which is high-frequency churn. And its high leverage is enforced by auto-liquidation: rather than a broker phoning for more collateral, a perpetual position is closed out automatically the instant it can no longer cover its margin, which protects the venue but turns a sharp move into a cascade of forced selling. When President Trump surprised the market with 100 percent tariffs on China in October 2025, the resulting crypto selloff auto-liquidated more than 19 billion dollars of leveraged positions in a day.

It is worth asking why these venues liquidate at all rather than simply call for more margin, the way a futures broker does. The answer is that a margin call needs three things a perpetual venue does not have. It needs a known, creditworthy counterparty: a broker can let a position run underwater and phone you for a wire because it did the paperwork, knows who you are, and can pursue your other assets if you do not pay, whereas a perp trader is often pseudonymous, global, and has posted nothing but the collateral in one account, so the only enforceable margin is what is already on deposit. It needs time: margin calls run on a daily settlement cycle with hours to wire funds, but a perpetual trades around the clock and a position at fifty or a hundred times leverage can gap through its whole margin in seconds, far faster than any human could answer a call. And it needs a clearinghouse to absorb a failure, the mutualized capital that stands behind every regulated futures trade, where a perp venue has only its own balance sheet and an insurance fund to protect. Lacking all three, the venue substitutes an automated engine that closes the position against its posted collateral the instant it falls short, then covers any shortfall from the insurance fund and, if that is exhausted, by auto-deleveraging, forcibly closing the most profitable traders on the other side. It is the price of a market that promises you can never lose more than you put down, a promise regulated futures have never made.

The open question for the oil market is whether the regulated push, led by the ICE and OKX deal and a more permissive CFTC, turns the perpetual into a genuine 24-hour complement to the CME, or whether it stays a high-leverage, crypto-native bet that prices oil on weekends and little else.

The Borrowed Price

The deepest objection is that no commodity perpetual has yet solved its own price. A crude future on the CME knows what a barrel is worth because a trader who holds it to expiry must make or take delivery at Cushing; the price is anchored to physical oil. A perpetual has no such anchor, so it has to import an oil price from somewhere. The legitimate way is to license a regulated benchmark, which is exactly what the May 2026 ICE and OKX deal does: it runs perpetuals explicitly under license to ICE’s Brent and WTI benchmark prices. The crypto-native venues disclose no such license. The largest on-chain crude market instead runs an oracle that tracks the nearby CME futures price through a feed its pseudonymous builder supplies, rolling off the front month to the second contract before expiry to avoid the delivery period, so it is often marking against the second-nearby future rather than the front month. And when the CME is shut over a weekend that oracle has nothing live to follow and is reported to drift along a moving average until New York reopens. The price is borrowed, and when the lender is closed it is improvised.

A natural defense is that a daily settlement price is a public fact, the kind of number the financial press prints every morning, and US law agrees: facts are not copyrightable, so anyone may quote today’s WTI close. But that principle is narrower than it looks. Quoting one daily number as news is not the same as continuously tracking a benchmark and settling leveraged positions against it; that systematic, product-building use is what exchanges control through license contracts and misappropriation law, regardless of copyright, and the CME’s own derived-data license names settlement prices and contracts-for-difference explicitly. The free-fact umbrella also reaches only the prices actually published as facts. Exchange settlement prices for the liquid futures, WTI, Brent, COMEX gold, the LME official prices, are quoted freely in the press. The price-reporting-agency assessments that underpin most physical oil, Platts Dated Brent, Argus, and OPIS, are not: they are proprietary, subscription-only assessed values, fenced by license and database right, and cannot be rebroadcast at all. So a perp can lean on a quotable futures settlement, but it cannot freely build itself on a Platts assessment, and even the futures number is free to quote, not free to systematically redistribute and settle against around the clock.

That design raises questions a delivery-backed contract never has to answer. There is, as yet, no commodity perpetual at scale running on a self-contained oracle that has been battle-tested and approved in the open. That a pseudonymous party sits in the price-setting seat for an oil benchmark is itself uneasy, and the natural inference is that a named, regulated price-setter would invite legal tests this structure may not withstand. Oracle manipulation is a recognized attack class in decentralized finance with real precedents, and the structural worry writes itself: what stops whoever controls a thin, deployer-run oracle from quietly building a position and then moving the feed against everyone else? Hyperliquid layers on guardrails, a stake-weighted median of several exchange feeds for its core markets, a cap on how far each update can move, and a review that can slash a deployer whose price jumps too far, but these constrain the risk rather than remove it, and the builder-deployed model still concentrates trust in one party. Finally, the headline volume itself would not survive the scrutiny a regulated venue faces. Spoofing has been a federal crime on US futures since the Dodd-Frank Act, enforced with penalties like JPMorgan’s 920 million dollar settlement in 2020, while studies of unregulated crypto venues have estimated that a large majority of reported volume is wash trading. Much of the perpetual’s high-frequency churn would be unlikely to clear that bar, which is precisely why commercial oil hedgers, who answer to auditors and regulators, still treat these venues with caution. It is no accident that the same incumbents who sued and lobbied over the CFTC’s approval, the CME and ICE, also pressed regulators to scrutinize the offshore venues for manipulation risk.

There is a tell, finally, in who is and is not in this market. The active participants are crypto-native: algorithmic market-making vaults, proprietary trading bots, and retail and non-crypto speculators drawn by the weekend access, which is how JPMorgan’s analysts described the oil-perp flow during the 2026 volatility. No bank commodity desk, oil major, or large physical trading house such as Vitol, Trafigura, Glencore, Gunvor, or Mercuria has disclosed taking positions on these venues, and the institutional-flavored moves so far, the ICE and OKX deal and Coinbase’s regulated perpetuals, are announced products aimed at retail or limited to crypto underlyings like bitcoin, not realized commercial oil-hedging flow; there is still no US-regulated onshore oil perpetual at all. Part of the reason is plumbing, custody, counterparty risk, and the fact that Hyperliquid blocks US users, but part is integrity, the very benchmark and manipulation concerns above. When the firms that actually hedge oil for a living stay on the sidelines, that is itself a verdict on whether the instrument is ready to be one.

Sources include BitMEX, Hyperliquid and trade[XYZ] documentation (oracle and HIP-3 design), Coinbase, the CFTC, Businesswire and Ledger Insights (the ICE-OKX licensing announcement), CoinDesk (Sprecher's May 2026 remarks and the CME and ICE manipulation complaints), the CastleLabs Hyperliquid oil microstructure study, the CME Derived Data License and Dow Jones v. Board of Trade (benchmark IP), OWASP and CertiK (oracle manipulation), CFTC PR 8260-20 (the JPMorgan spoofing settlement), and the NBER crypto wash-trading study, and The Wall Street Journal (June 2026, the CFTC crypto-perp approval, Kalshi and Coinbase volumes, the exchange-share selloff, the CME lawsuit, and the October 2025 auto-liquidation cascade). Figures from the youngest, on-chain part of this market are approximate and move quickly.