Posted June 1, 2026 — three days after the CFTC's coordinated perpetual-contracts policy package.

The big change is small on paper and large in practice. On May 29, 2026, the CFTC took four coordinated actions that together open a domestic path for perpetual derivatives in the United States: it issued a policy statement on the listing of perpetual contracts (case-by-case review under Reg 40.3 for asset classes outside the existing order); approved KalshiEX's BTCPERP contract — the first CFTC-regulated bitcoin perpetual future on a Designated Contract Market; released a staff advisory on 24/7 trading and clearing operations; and issued an interpretive letter and no-action position for Coinbase Financial Markets recognizing certain Deribit perpetuals as foreign futures under Reg 30.1 — operationalizing Coinbase's $2.9B acquisition of Deribit for US institutional clients. Same week, CME flipped its full crypto futures complex to 24/7 trading — including the Spot-Quoted Futures product CME has quietly run since June 2025, which is the closest economic analog to a perp anyone offers under a CFTC-regulated DCM wrapper. The infrastructure that has run tens of billions in daily volume offshore for half a decade now has multiple US-sanctioned mirror images.

The interesting question isn't whether perps will trade onshore. That's decided. The interesting question is what the design of a perp actually is, where the three regulated/permissionless paradigms converge — and what each one structurally implies for the rest of US derivatives. That's the piece below.

1. What a perp actually is

A perpetual future is a derivative that mimics a spot exposure without ever expiring. It has three load-bearing parts:

• A reference price — an oracle or index that says what the underlying is worth right now.
• A funding mechanism — periodic cashflows between longs and shorts that pay whichever side is on the wrong side of the contract-vs-reference gap. This is the economic tether.
• A margin and liquidation system — collateral, maintenance requirements, and forced unwinds that keep the system solvent when prices move.

The first two are what make a perp a perp. The third is what a dated futures exchange already does. So the contract-design conversation collapses to: how is the reference price constructed, and how does funding actually work? Everything else — leverage, fees, hours, custody — falls out of those choices and the regulatory regime around them.

2. The contract-design knobs

A small number of parameters do most of the work. Designers turn each one against the same trade-off — economic accuracy vs. user and platform safety.

• Oracle source & cadence. What feed defines the reference price, and how often it's updated. Continuous, redundant institutional feeds (Pyth, Chainlink, CF Benchmarks) are the modern default; the GSR build I wrote up for Hyperliquid uses a 3-second oracle cadence with median-of-marks blending. CME uses the BRR/BRTI; Kalshi uses CF Benchmarks' BTC Real-Time Index; Deribit uses its own composite index.
• Funding interval & formula. Hyperliquid accrues continuously and settles hourly with a 4%/hr absolute cap; Kalshi and Deribit both pay funding every 8 hours, the offshore default. The funding multiplier and any interest-rate floor or clamp determine how aggressively the contract is pulled back to the index.
• Mark-price safety clamps. The mark used for margin and liquidation can't be allowed to jump arbitrarily on a single bad print. Hyperliquid clamps mark moves to 1% per oracle update and oracle prices to 10× start-of-day; DCMs use circuit breakers and price-banding to similar effect.
• Leverage & margin schedule. Maximum leverage, initial vs. maintenance margin, isolated vs. cross. DCMs deliver these through a clearing house; permissionless venues do it through on-chain margin tables and protocol-level liquidation; the foreign-futures path runs through an FCM into the foreign exchange's margin system.
• Contract size & tick. Kalshi's BTCPERP is 1/10,000 BTC per contract — deliberately retail-sized. CME's standard is 5 BTC (0.1 for the Micro). Deribit uses inverse contracts with fixed-USD face value ($10 BTC, $1 ETH). Hyperliquid HIP-3 builders set szDecimals at registration.
• Listing and surveillance. Who can list a market, who supervises it, what gets surveilled. This is the dimension that splits the universe in three — and where the CFTC's May 29 actions land.

3. Four listing paradigms, side by side

The clearest way to see what the CFTC just changed is to put four venues alongside one another: the permissionless HIP-3 perp (Hyperliquid), the freshly-approved true DCM-listed perp (Kalshi's BTCPERP), the Coinbase/Deribit foreign-futures pathway, and CME's Spot-Quoted Futures — the perp-equivalent product CME has already run for a year under a standard DCM wrapper.

The CME spot-quoted column is worth flagging up front because it tends to get omitted from "US perps" conversations. Mechanically it's a series of monthly futures that quote at the spot price (not the basis premium); a daily financing adjustmentamortizes the lead-month future's basis to expiry, paid by the long when basis is positive and the short when it's negative. Functionally that is a perp settled daily — same economic tether, different legal wrapper, full §1256 tax treatment, $500–$6,000 retail-sized notional. CME launched it for BTC/ETH in June 2025 and extended to XRP and SOL in December 2025; it now goes 24/7 alongside the dated complex. If you wanted to know what "regulated US perps" would look like before Kalshi's May 29 approval — this was already the answer.

A few things stand out. All four converge on the same economic mechanism — periodic cashflows that pull the contract toward an index — but the implementation diverges in instructive ways. Hyperliquid encodes the caps as protocol constants enforced on-chain (4%/hr, 1% per mark update, 10× oracle clamp); Kalshi achieves analogous economic effects through DCM apparatus (margin schedule, circuit breakers, position limits); Coinbase/Deribit inherits Deribit's 8-hour mechanism unchanged and bolts on FCM intermediation; CME spot-quoted does the same economic work via a once-per-day financing adjustment derived from the lead-month future's basis. The differences in cadence (continuous → 8h → daily) and the differences in cap surface (protocol constant → DCM rulebook → exchange methodology → daily price limits) are the only meaningful design distinctions left. Same product, four wrappers.

4. The caps and floors, interactively

The most useful way to feel what the caps actually do is to put a premium on the contract and watch the funding rate respond. The dashboard below lets you do that — pick a venue, drag the premium, and see which clamps bind, which don't, and what the position costs to hold over one funding interval.

The economically interesting case is the small, persistent premium. At a 5 bp premium on a $50k position, an 8-hour Kalshi or Deribit interval pays the funded side $25; the equivalent Hyperliquid hourly accrual is about $3. Both rates are small, and both caps are far from binding — they exist for the regime that hits during a crisis, not the regime you trade in. The hardest design choice isn't the cap level; it's how aggressive to make the funding multiplier so that a 5 bp premium decays to zero before opportunistic arb desks have to step in.

5. The third path: foreign futures via FCM

The Coinbase/Deribit interpretation is the most operationally consequential of the May 29 actions, and the easiest to underweight. Here's the structure:

• Reg 30.1 categorization. The CFTC's Market Participants Division confirmed that perpetuals on Deribit FZE may be categorized as "foreign futures" under Regulation 30.1. That category exists explicitly so US persons can access non-domestic derivatives through a CFTC-registered Futures Commission Merchant — which is what Coinbase Financial Markets (CFM) is. The interpretation is conditional: it covers perps structurally similar to the Deribit Perpetuals and based on digital commodities with deep, active, and continuous spot market trading. That language is the gating criterion for any future product expansion through this channel.
• The margin no-action. Equally important — and less covered — is the parallel no-action letter on margin transfers. Under standard FCM segregation rules (CEA §4d, CFTC Reg 1.20–1.26), FCMs can't post customer assets with foreign affiliates without specific relief, and crypto-as-margin doesn't fit cleanly under the traditional money-and-securities framework. The MPD said it will not pursue enforcement if CFM posts customer crypto assets and stablecoins with its foreign broker affiliate — even where those assets may be reused by the broker. That last clause is what makes this operative at scale: Deribit can run a normal, capital- efficient prime brokerage relationship with US institutional customer collateral.
• What it operationalizes. Coinbase paid $2.9B for Deribit in August 2025. Deribit is the largest crypto-options venue by a wide margin — $60B in open interest, more than $1T traded in 2025, a record $266B notional in October 2025. Coinbase has owned that book for nine months but has been unable to offer it to US institutions through a regulated channel. The May 29 actions are the regulatory clearance the acquisition was effectively structured around. Practically, this puts deep crypto-derivatives liquidity — for BTC and ETH first, others to follow — inside the reach of US institutional balance sheets in a way that's never existed before.
• What it does not do. The interpretation is specific. It applies to Deribit FZE (regulated by Dubai's VARA), perps with deep-spot underlyings, and intermediation through Coinbase's FCM. It is not a blanket recognition of offshore venues. A permissionless venue like Hyperliquid does not become accessible through this route — it has no foreign regulator, no foreign- board-of-trade structure, no FCM intermediation. Listings on esoteric underlyings (the GSR perp, an FX-pair perp, a niche commodity perp) likely fail the "deep, active, continuous spot market" test even if the venue qualifies.

6. The GSR build as a worked example

The Gold/Silver Ratio Perp spec I wrote earlier in May (Hyperliquid HIP-3 builder-deployed) is useful here precisely because it's the same question, in a non-crypto underlying, on the permissionless side of the line:

• Oracle as the entire job. The deployer publishes the XAU/XAG ratio every ~3 seconds; if the oracle stops, marks fall back to local within 10 seconds and the 500,000-HYPE deployer stake is slashable.
• Funding does the tethering. Funding multiplier 0–10; interest component zeroed (no built-in drift); 4%/hr absolute cap inherited from the protocol.
• Protocol clamps protect everyone from the deployer. Mark moves clamped to 1% per update, oracle clamped to 10× SoD. These exist precisely because the deployer can't be trusted with unbounded write access to the reference.
• Non-24/7 underlying matters. Metals close on weekends; the perp doesn't. HIP-3 imposes additional risk controls for non-24/7 asset classes; a DCM listing a similar product would do this with explicit holiday-margin schedules.

The GSR spec is essentially a Hyperliquid-flavored answer to every question Kalshi just answered in its BTCPERP submission. The surprising thing — three days into the new regime — is how much the answers overlap. The vocabulary is different (deployer stake vs. clearing house; setOracle vs. settlement procedures), but the economic skeleton is identical.

7. Implications for onshore derivatives

Eight things change for US derivatives markets, in roughly the order they'll show up in price and product flow:

• DCMs face structurally different competition. Kalshi is the first regulated US perp by name, but CME already had the economic equivalent running since June 2025 in spot-quoted futures, and the Coinbase/Deribit pathway gives institutional traders access to deeper offshore liquidity through a US-regulated FCM. The competitive matrix isn't "regulated vs. unregulated" — it's four regulatory wrappers around the same economic instrument, each optimized for a different customer segment (retail-sized DCM perp at Kalshi, retail-sized financing-adjusted future at CME, institutional foreign-futures at Coinbase/Deribit, self-custodied permissionless at Hyperliquid). The 24/7-trading move at CME is the protective layer beneath its spot-quoted product — not a hedge, an alignment.
• §1256 tax treatment becomes a real differentiator. DCM-listed perps (Kalshi BTCPERP, CME Spot-Quoted Futures) qualify for §1256 mark-to-market and 60/40 long-term/short-term capital gains — a meaningful effective-rate advantage for active US traders. Foreign-futures-categorized perps may qualify under §1256(g) depending on facts. Offshore unregistered access (Hyperliquid, Bybit) yields ordinary-income treatment plus all the AML/sanctions exposure. The pre-tax-vs-after-tax gap pushes flow toward the regulated paths quietly but persistently — and CME spot-quoted has had this advantage for a year already without most of the market noticing.
• Reference indices become quasi-regulated infrastructure. The CFTC's foreign-futures interpretation is contract-structure-conditional: it requires "deep, active, continuous spot market trading." That makes the BTC and ETH spot indices (BRR, CF Benchmarks RTI) precondition-level infrastructure for product approval. Index providers become quasi-regulated utilities — the analog of LIBOR's old role but for digital commodities. Expect index consolidation and increased scrutiny of methodology, publisher composition, and outage-handling.
• Margin innovation got unblocked. The no-action on customer-asset transfers is operationally large. Stablecoin margin, crypto-as-collateral, and explicit permission for affiliate asset reuse at foreign brokers — these are market-structure changes wrapped in a procedural release. Expect the same template to spread to other FCM–foreign-affiliate pairs (CME group with non-US affiliates; ICE with foreign clearing services).
• Cross-venue basis becomes a real market. With BTC perps trading on Kalshi (DCM), Deribit-via-Coinbase (Reg 30.1), eventually CME, and offshore on Hyperliquid, the basis between venues is structurally funded — different funding cadences (1hr vs 8hr), different reference indices, different margin pools, different liquidation behavior. Funding-rate divergence between Kalshi and Deribit on the same underlying is a trade.
• DeFi venues face new pricing pressure but keep their own advantages. Hyperliquid and Drift now compete with regulated alternatives for the same flow. But permissionless listings — any underlying, in days, no Reg 40.3 review — remain a structural advantage for esoteric underlyings. Expect bifurcation: deep, mainstream underlyings flow onshore for tax-and-custody reasons; long-tail and novel exposures stay permissionless. The GSR perp is a textbook example of the latter — no DCM is listing XAU/XAG anytime soon.
• Onshore listings will accelerate. Reg 40.3 case-by-case review is the only door for non-BTC perps to come onshore via the DCM path. The first ETH approval is the bellwether; once that template exists, SOL, basket products, and traditional commodity perps follow quickly. Six-month outlook: 5–10 regulated perp products live across DCMs, plus ETH likely added to the Coinbase/Deribit foreign-futures interpretation.
• The clearing-and-settlement stack has to modernize. The 24/7 staff advisory is the most underappreciated of the four actions. US clearing houses run on business-day cycles — variation margin moves daily, settlement is T+1 or T+0, holiday calendars matter. CME's spot-quoted product chose daily financing precisely because that's what its clearing stack can natively support; the offshore venues run hourly or 8-hourly because their stacks were built for it. Whichever US clearing house ships a sub-daily variation-margin pipeline first — and ICE, CME, and OCC are all positioned to do it — owns the next round of product launches, and the gap between "regulated perp" and "regulated true perp" closes.

This site exists to demonstrate what AI now makes possible in financial engineering. The perp regime moving onshore is the most consequential change in US-regulated derivatives in fifteen years, and it's happening in a window where one person, one oracle, and one weekend can prototype a working market. That's the story.

Sources

• CFTC press release 9242-26 — Policy Statement on the Listing of Perpetual Contracts (May 29, 2026).
• CFTC press release 9240-26 — Approval of KalshiEX BTCPERP.
• CFTC press release 9241-26 — Categorization of Certain Crypto Asset Perpetuals as Foreign Futures and FCM Customer-Margin No-Action Letter.
• Coinbase — Deribit Joins Coinbase: Unlocking the Future of Global Crypto Derivatives (acquisition close, August 2025).
• CoinDesk — U.S. CFTC opens crypto 'perp' door with first approval at regulated firm.
• Lowenstein Sandler — CFTC Approves U.S. Bitcoin Perpetual Futures Contract and Issues Related Guidance.
• CME Group — 24/7 cryptocurrency futures and options trading from May 29.
• CME Group — Spot-Quoted Futures product page and FAQ.
• CoinDesk — CME Group Expands Crypto Derivatives With Spot-Quoted XRP and Solana Futures (Dec 2025).
• Hyperliquid docs — Funding mechanics and oracle / mark price.
• Internal — GSR Perp Contract Spec & Launch Runbook (Daniel Kaufman, May 2026).