The Filing
On the CFTC's comment docket this month, a trade association representing the largest derivatives venues on earth asked a regulator to prohibit a product its own members do not list. The Commodity Markets Council โ the Washington vehicle for CME Group, ICE, and the commodity trading houses that clear through them โ told the Commodity Futures Trading Commission that perpetual futures pose significant risks to market integrity. It urged the agency not to approve the perpetual crude oil contracts proposed by Kalshi, a CFTC-regulated designated contract market. And it asked the Commission to close regulatory loopholes that let retail traders reach offshore platforms.
Bloomberg reported the submission. Three asks, one docket, and roughly eleven years of unresolved market-structure argument compressed into a few pages of regulatory comment.
Strip the language and the request is narrower than it sounds. This is not a letter about oil. It is a letter about jurisdiction โ about which venue gets to publish the number that prices a barrel, and which balance sheet absorbs the loss when that number is wrong. The comment period is a formality. The substance is a boundary dispute that has been building since 2016, when an exchange nobody in Chicago had heard of invented a derivative with no expiry date and discovered that retail traders preferred it to everything else on the board.
Who Is Actually Speaking
The Commodity Markets Council is not a household name and is not meant to be. Its members sit on the boards of the venues that clear the world's agricultural, energy, and metals risk, and on the trading desks of the firms that move the physical barrels. When it files, it files for the incumbency: the clearinghouses, the futures commission merchants, the merchandisers whose hedging programs are calibrated to the term structure of NYMEX WTI.
That origin matters for how the argument should be read. A trade group defending its members' franchise will always find a market-integrity rationale, and the rationale will always be partially true. The analytical task is to separate the part that is structurally correct from the part that is rent preservation. Both are present here. The first is worth the Commission's attention. The second is worth naming.
What Kalshi Is, and Why Its Charter Changes the Argument
Kalshi is a designated contract market. That phrase carries weight and is routinely misread.
A DCM is not a broker, not a swap execution facility, not a foreign board of trade operating under a registration that the Commission rarely inspects. It is a licensed exchange under the Commodity Exchange Act, bound by core principles covering financial integrity, market surveillance, position limits, and the affirmative obligation to prevent manipulation on its own book. When Kalshi proposes a contract, it proposes it inside the perimeter โ not outside it.
Kalshi spent 2024 litigating against its own regulator for the right to list election contracts, won at the district level, and became the most consequential market-structure story of the cycle. That litigation established a precedent the industry has been quietly exploiting ever since: a CFTC-registered venue with enough legal capital can force the Commission to articulate why a product is unlawful rather than merely unwelcome.
Perpetual crude is the second act, and it is a materially harder case. Elections are a pure cash-settled event with a discrete outcome and no underlying physical market to disturb. Crude oil has a spot market, a storage network, a freight curve, and a forty-year-old futures complex whose front-month print is embedded in physical supply contracts from Rotterdam to Singapore. Kalshi is not proposing to add a product to a market. It is proposing to add a second price to a market that already has one โ and the second price would have no delivery obligation attached to it.
That is the whole case. Everything else is commentary.
What a Perpetual Actually Is
A perpetual futures contract is a derivative with no expiry date. That single design choice deletes the mechanism that makes a futures contract a hedging instrument.
In a standard future, the exchange fixes a settlement date. As that date approaches, arbitrage pulls the contract price toward spot, because a holder who reaches expiry must either deliver the underlying or cash-settle against a published index. Convergence is what makes the instrument usable for hedging. A refiner that will buy crude in March sells a March future today and knows the two legs offset at expiry. Basis risk is bounded by the delivery mechanism. That bound is the product.
A perpetual has no expiry, therefore no delivery, therefore no convergence. Instead the venue imposes a periodic cash flow between longs and shorts โ the funding rate โ engineered to pull the perpetual's mark toward an index reference. The formula is mechanical: a premium index measuring the perpetual's deviation from spot, clamped by a rule, sometimes with a fixed interest component. On the large crypto venues the interest component sits at 0.01% per eight-hour epoch, which annualizes to roughly 10.95% before the premium term moves it, and the clamp typically binds at ยฑ0.05% per epoch in stress.
The mechanics look elegant in a whitepaper. In production they are a liquidation engine with a calendar attached.
A perpetual does not converge to spot; it converges to a funding formula. Every property that follows โ the manipulation surface, the liquidation cascade dynamics, the market maker's hedging cost, the shape of the order book at 3 a.m. โ is a downstream consequence of that one substitution. If you accept the substitution, you accept its arithmetic.
Why Crypto Perpetuals Worked and Oil Breaks Them
Crypto perpetuals scaled because the underlying has no cost of carry. Bitcoin cannot be stored in a tank. There is no convenience yield to a coin. The only carry is the cost of borrowing the quote currency, which is a single, observable, mean-reverting number set by the money market. Funding therefore lives in a narrow band and behaves like an interest-rate instrument. The whole architecture โ thin maintenance margins, high leverage, an insurance fund, auto-deleveraging โ was calibrated against a funding series with low variance.
Crude oil has the opposite profile. Its fair-value relationship is a cost-of-carry model with three components that move independently: physical storage cost, financing cost, and convenience yield. Storage at Cushing runs in the range of tens of cents per barrel per month in calm markets and reprices violently when inventories tighten. Convenience yield is unobservable, non-stationary, and spikes precisely when the market is most stressed โ which is exactly when a derivatives venue most needs its price to be stable.
Run the arithmetic. With WTI at $70 and a six-month contango of $2, the implied annualized carry is roughly 5.7%. That is before a risk premium, before the inventory shock, before the geopolitical gap. An oil perpetual would need a funding band several hundred basis points wide to track its own fair value โ and a funding band that wide is a liquidation engine that fires on inventory prints.
This is where the CMC's technical case is strongest and where the letter, at least as reported, does not go. The problem is not that retail traders would lose money on a perpetual oil contract. Retail traders lose money on every leveraged product ever listed. The problem is that the funding mechanism is mathematically incapable of representing a physical carry curve whose dominant input is an unobservable convenience yield that reprices on a weekly government inventory release. You would be pricing a barrel with an instrument that has no storage in it.
The Negative Print a Perpetual Cannot Make
On 20 April 2020, the front-month WTI contract settled at negative $37.63. The price went below zero because delivery was mandatory, Cushing was full, and holders with no tank space paid to be rid of the obligation. It was an ugly, clarifying event. It told the entire physical complex exactly how much the marginal barrel was worth when the marginal storage bid disappeared.
A perpetual oil contract would not have printed that number. It would have printed a funding rate. The mark would have been dragged toward an index that was itself derived from a futures curve that had already gone negative, and the funding mechanism would have smoothed, clamped, or deferred the signal. The venue would have reported a functioning market while the physical market was telling participants to pay someone to take their exposure.
That is the integrity problem, stated precisely. Convergence is a legal obligation, not a market outcome. A contract with no delivery date has no obligation to be right about the physical world, and in the one historical episode where being right about the physical world mattered more than anything else, the perpetual structure would have produced a cleaner-looking chart and a worse signal.
I have spent enough time inside settlement design to know how this gets argued away. The response is always that the perpetual's index can be constructed from the deliverable contract, so the information is preserved. It is not. Index construction is a smoothing operation, and smoothing operations are how venues manufacture the appearance of liquidity during the exact intervals when liquidity is the scarce input. The information survives in the futures curve. It does not survive in the perpetual's mark.
Basis Fragmentation and the Reference Price
The CMC's market-integrity claim is usually read as a retail-protection claim. Read it as a liquidity claim instead and it becomes far more serious.
Hedgers use WTI futures because the contract converges. A producer shorting the front month is not expressing a view on crude; it is extinguishing basis risk against a physical obligation. That hedge works because the contract's expiry forces the exchange-traded price and the physical price to meet. Remove the expiry and you remove the hedge's terminal condition. You now have a venue publishing a crude-linked price that never has to reconcile with a barrel.
If that venue accumulates volume โ and retail flow follows leverage, not convergence โ you get two things simultaneously. First, liquidity fragments across two price-discovery venues, one of which has no delivery obligation. Second, the reference price for physical supply contracts becomes contestable, because there is now a second screen to point at in a negotiation. The front-month WTI print is not merely a market price. It is an input into freight contracts, refinery margins, airline fuel hedges, and sovereign budget assumptions in petro-states. When you add a second price to a reference asset, you do not add information. You add an arbitration.
I watched this exact failure mode in miniature during my 2020 audit of Uniswap V2 yield mechanics. The pool published a price. The price was arbitrageable against the deeper venue. Retail LPs read the pool price as a valuation signal and supplied liquidity against it. My model projected roughly 40% principal erosion for naive LPs within six months on stablecoin pairs, driven almost entirely by the gap between the published price and the executable one. The lesson generalizes: a thin venue publishing a price next to a deep venue publishing a price is not price discovery. It is a subsidy transfer from the participants who believe the thin venue to the participants who arbitrage it.
Kalshi's perpetual oil, at launch volume, would be the thin venue. The CMC knows this. That is not a cynical observation; it is the actual mechanism of the harm they are describing, and it is more persuasive than the retail-protection framing they led with.
The Offshore Loophole Is Not an Exchange Problem
The third ask in the letter is the one that will be remembered, because it is the one the Commission cannot satisfy. The CMC wants the CFTC to close regulatory loopholes preventing retail investors from accessing offshore trading platforms.
The loophole, stated precisely, is the foreign board of trade registration regime and its interaction with the Commission's ability to reach the customer rather than the venue. A venue outside US jurisdiction lists a product. A US retail trader reaches it through a website, a virtual private network, and a stablecoin on-ramp. The Commission's enforcement authority attaches to the intermediary that touches US persons โ the FCM, the introducing broker, the money transmitter โ not to the offshore order book.
The arithmetic of that gap is not theoretical. Offshore crypto perpetual venues clear notional volumes in the trillions annually, and they have done so for years while the Commission has been fully aware. No enforcement action has materially reduced that flow, because there is nothing to enforce against at the point where the trade happens. The trade happens in code on a server in a jurisdiction that has decided not to care.
This is where the structural analysis has to be honest about what regulation can and cannot do. Code enforces; policy dictates. The Commission can dictate that a US-facing intermediary may not onboard a customer to an offshore perpetual venue. It cannot dictate that the venue stop existing. Every attempt to close an offshore loophole by letter rather than by touching the settlement rail has produced the same outcome: the flow migrates to the venue with the weakest compliance posture and the settlement asset with the fewest intermediaries.
The settlement rail is the actual chokepoint. Stablecoin issuance, banking access for the issuers, and the correspondent relationships that let dollar-denominated value move at all โ that is where jurisdiction still lives. I spent 2023 running a retail CBDC pilot with a team of five engineers against a $500,000 budget, optimizing a permissioned ledger to 10,000 transactions per second with privacy features intact. What that project demonstrated to me was not that state ledgers are faster than public chains. It was that a permissioned ledger's real advantage is not throughput; it is the ability to freeze a balance at the settlement layer without asking anyone's permission.
That capability is the answer to the offshore loophole, and it is a capability the Commission does not currently possess over the instruments in question. Asking for it in a comment letter is asking the wrong agency for the wrong tool. The CMC is asking the CFTC to do something that only the Treasury, the banking regulators, and the stablecoin issuers can actually do. That does not make the ask wrong. It makes it misaddressed, which in regulatory strategy is the same as ineffective.
Sequencing Risk and the Cost of Compliance
There is a technical reason regulated venues struggle to replicate perpetuals, and it has nothing to do with risk appetite. It is sequencing.
A perpetual's utility comes from atomicity. Margin check, funding accrual, mark update, liquidation trigger, and insurance fund draw all execute in one indivisible sequence against a shared state. On a public chain this is free โ the protocol enforces the order of operations, and the enforcement is the product. On a regulated venue, atomicity must be replicated by a central counterparty that absorbs the sequencing risk onto its own balance sheet, because a licensed exchange cannot let one leg settle while another fails. That is a capital cost. It is a real, quantifiable, permanent capital cost.
The CMC's argument, properly reconstructed, is that this capital cost should be mandatory. That is a prudential argument, not a market-integrity argument, and it is a defensible one. A clearinghouse that stands behind every leg of a non-expiring contract has to post capital against a liability with no terminal date. The margin model for such a contract is a different mathematical object than the margin model for a dated future, and the difference is not a rounding error.
But the argument cuts both ways. If the compliance cost is genuinely high, then a compliant venue that nevertheless chooses to list the product has priced that cost and accepted it. The Commission's role is to verify the margin model, not to decide whether the product should exist. Refusing to approve a contract because the incumbent's cost structure is uncompetitive is not market integrity. It is industrial policy conducted through the approval process, and it will be litigated exactly as such.
The Liquidity Illusion, Imported
Here is the part the letter does not say, and the part that matters most for anyone holding risk in this market.
The market maker on the other side of a perpetual is structurally short volatility and short the basis. In crypto, the basis is a borrowing rate and the variance is tolerable, which is why the large venues can run tight spreads with thin insurance funds. In oil, the basis is a physical carry whose dominant driver is a weekly inventory print, and the variance is not tolerable at retail leverage.
Run the position through a standard stochastic treatment. The market maker's P&L is a function of funding accrual minus the hedging cost of the offsetting futures position, and the hedging cost is dominated by roll slippage and inventory-driven basis jumps. Add a retail-facing leverage cap of 20x โ typical for perpetual products โ and you have constructed an instrument that transfers inventory risk from the participants who can warehouse it to the participants who cannot. The transfer is not a bug. It is the product's economic function, whether or not anyone says so.
I have modeled this before in a different venue. The impermanent loss result from my Uniswap V2 work was, mathematically, the same object: a liquidity provider selling optionality to arbitrageurs and being compensated by fees that were calibrated to a variance regime that did not hold. The fees looked adequate in the calm sample. They were catastrophically inadequate in the stress sample. Perpetual funding works identically. It looks adequate against historical basis variance. It is not adequate against the tail, and the tail in crude oil arrives on a Wednesday morning with a government press release attached.
Liquidity that only exists in the regime where you do not need it is not liquidity. It is a marketing number, and it is the number that will be quoted in every deck arguing for the product's approval.
What the Flow Data Suggests
My 2024 work on ETF inflows gives me a prior here, and the prior is not comfortable for either side.
I built an algorithm that tracked daily institutional inflows against retail outflows across fifteen venues and correlated the net series with S&P 500 volatility indices. The finding that mattered was not the direction of flows. It was the concentration dynamic: capital consolidated into the deepest instrument, and liquidity drained from everything adjacent. I forecast a 15% correction on that basis, presented it to a private investment club in Warsaw against a $2 million allocation, and the forecast held.
Import the same lens to this docket. If a retail-accessible perpetual oil contract lists and gains traction, the flow that reaches it comes disproportionately from the front-month futures complex, because that is where the adjacent liquidity sits. Open interest migrates. The front month thins. The roll becomes more expensive for the physical hedger who has no alternative venue and no ability to use a non-converging instrument. The hedger pays for the retail product's existence, and the hedger passes the cost into the physical supply chain, where it becomes a line item in the price of everything that moves.
That is the second-order effect nobody puts in a comment letter, and it is the strongest argument the incumbents have. They will not make it, because making it requires admitting that the front-month complex is a public good rather than a franchise.
The Agent Economy Does Not Roll Contracts
Set the docket aside for a moment and look at the direction of the instrument itself, because this is where the policy decision runs into something larger than a trade association.
Perpetuals are the native derivative of machine finance. An autonomous agent executing a hedging or market-making strategy cannot manage a roll schedule without a calendar-aware execution layer, a funding account, and a delivery obligation it has no physical capacity to satisfy. A contract with no expiry removes an entire class of operational failure from the agent's design space. That is why perpetuals dominate on-chain derivatives volume and why every serious attempt to build machine-to-machine settlement converges on a non-expiring, continuously-funded instrument.
In 2025 I designed a tokenomic protocol for autonomous agents trading compute resources under a $1.2 million consortium grant. The hardest design constraint was not consensus. It was that the agents needed to hold and settle positions continuously, without a settlement calendar, and without a human in the loop to authorize a roll. Every dated instrument we modeled introduced a failure mode at the boundary. We ended up building around a continuous funding primitive because nothing else survived the operational constraints.
Macro trends crush micro-protocols. The trend here is toward continuous, non-expiring, 24/7 instruments settled against a funding formula. A DCM approval decision is a micro-event against that trend. The Commission can decline to approve Kalshi's contract. It cannot decline to have the demand exist. The demand will be met, at a venue with a weaker margin model, settled in an asset with fewer intermediaries, and the US will have exported the market along with the regulatory oversight it was trying to preserve.
This is the same error I watched the DA layer market make. An entire category of infrastructure was built for demand that did not exist at the scale claimed โ most rollups never generated enough data to justify a dedicated availability layer, and the capital went into specialized capacity for a workload that stayed small. Here the error is inverted. The demand is real, the instrument is native to the next cycle, and the policy is being written as if the demand were speculative. Building dedicated infrastructure for demand that does not exist wastes capital. Refusing to license infrastructure for demand that does exist exports the market.
The Contrarian Read: The Letter Is the Best Case for Onshore Perpetuals
The obvious reading of the CMC filing is that incumbents are protecting turf, and that reading is correct as far as it goes. But it is not the most useful reading, and the most useful reading is uncomfortable for everyone involved.
Consider what the letter actually concedes. It concedes that perpetual futures are a distinct product class with a distinct risk profile that retail traders want badly enough to reach offshore for. It concedes that offshore venues are currently serving that demand at scale. It concedes that the Commission's existing tools are inadequate to stop it. And it asks the Commission to use its approval authority over a compliant domestic venue as the instrument of control.
That is an argument for onshore licensing, not against it. If the demand is real, persistent, and currently being served by venues with no US margin model and no US surveillance, then the rational regulatory response is to license a compliant venue with a conservative margin framework, position limits, and surveillance obligations โ and let it compete. Prohibition does not reduce the risk. It relocates the risk to the jurisdiction with the weakest supervision, which is the opposite of what a prudential regulator should want.
This is the same structural mistake that intent-based architectures make in DeFi. Intent systems are marketed as an improvement over order-book DEXs, but the MEV does not disappear โ it moves off-chain into solver networks where the ordering is set by a handful of searchers with no transparency obligation. The risk is not eliminated. It is relocated to a venue with worse observability. A CFTC prohibition on perpetual oil would do exactly that: relocate the funding-rate risk from a DCM with position limits to a server in a jurisdiction with none.
So the honest position is uncomfortable. The incumbents' technical objection is correct โ the funding mechanism cannot represent physical carry, and a non-converging crude price is a genuine systemic problem. The incumbents' proposed remedy is wrong, because prohibition exports the problem rather than solving it. What the Commission should do is neither approve the contract as filed nor reject it. It should require a margin and funding framework calibrated to inventory-driven basis variance, with position limits that cap the retail leverage that makes the transfer of inventory risk possible. That is a harder decision than a yes or a no, which is why it will not be the one taken.
What to Watch
Three signals will determine how this resolves, and none of them is the comment period.
Watch whether CME files for its own crypto perpetual. The group's members have been exploring perpetual-style structures on digital assets for two years, and the internal contradiction of opposing the instrument class in energy while developing it in crypto is not sustainable under public scrutiny. If CME files, the market-integrity argument collapses into a venue argument overnight, and the Commission will have to rule on the instrument rather than on the applicant.
Watch the Kalshi docket for the margin model. The Commission's approval authority lives or dies on the terms and conditions, not on the press release. A filing with a funding clamp calibrated to crypto variance and a 20x retail leverage cap will tell you more about the product's actual risk than any comment letter.
Watch the stablecoin settlement rails. Every offshore perpetual is settled in a dollar-denominated token whose issuance sits inside someone's banking relationship. That is the only chokepoint with real reach, and it is being tightened slowly and mostly by accident.
The bear market makes this easier to see. In a cycle where survival dominates returns and half the listed protocols are bleeding liquidity, the instruments that endure are the ones with a delivery obligation or a sovereign backstop. A perpetual has neither. It has a formula, a funding epoch, and a liquidation engine.
That is enough to run a market. It has never been enough to price a barrel.