Hook
The number that matters in SR-KALSHIEX-2026-02 is not the hundred-billion-dollar market cap floor. It is fifteen and a half percent.
Fifteen point five. Against a statutory minimum of fifteen. Half a percentage point of deliberate, unnecessary self-harm, written into a rule filing with the U.S. Securities and Exchange Commission, for a product that does not exist in any licensed American venue: a perpetual security futures contract on individual stocks. No expiry. Roughly six and a half turns of leverage. Cleared through Kalshi Klear LLC. Scheduled to go effective on the second of November, 2026.
Around that margin sit three more numbers. A minimum market capitalization of one hundred billion dollars. A minimum average daily traded volume of four hundred and fifty million. And a funding rate mechanism that moves cash between longs and shorts every day until the contract price converges on the spot price of the stock underneath it.
I have traded this machine. Not this filing โ the machine. The funding print at fixed intervals. The mark that drifts away from the last trade. The liquidation cascade at three in the morning when a spot venue halts and the anchor simply vanishes. I have watched a perpetual lose its mooring, and I have watched competent people discover that they were never holding a position at all. They were holding a promise about an oracle.
What Kalshi is asking for is permission to run that machine on the most crowded, most gap-prone equities in the world, inside a clearing entity with no operating history, under two federal regulators who have not yet agreed on what the machine legally is.
Everything else is noise. This is the trade.
Context
A perpetual swap is a derivative with no expiry. No settlement date, no delivery, no roll. You hold it until you close it or until the funding payments drain you. The structure was popularized offshore in 2016 by BitMEX and then industrialized by Binance, Bybit and OKX. In crypto it long ago stopped being a derivative on the price. It became the price. Perpetual open interest on the major offshore venues routinely exceeds spot volume by a wide margin, which means the reference market for most liquid crypto assets is a contract that never settles.
The mechanism that keeps a perpetual tethered to reality is the funding rate. Every funding interval, the venue computes a premium index โ the difference between the perpetual's mark and a spot reference price โ adds an interest rate component, clamps the result, and transfers that amount from the crowded side to the other side. If the perpetual trades above spot, longs pay shorts. The payment creates an incentive to short the perpetual and buy spot, which pushes the contract back toward the anchor. Do that enough times and you have a synthetic version of cost of carry, paid in installments instead of crystallized at delivery.
Two things follow. The contract can trade forever, because somebody is always paying to keep it honest. And the entire edifice depends on a single input: a spot price that is actually true when you read it.
That is the machine. Now the venue.
Kalshi operates KalshiEX LLC, a designated contract market under the Commodity Futures Trading Commission. Its existing business is event contracts โ binary outcomes on economic and political questions. In 2023 the CFTC refused to allow its congressional control contracts. Kalshi sued, and in 2024 the courts let the contracts stand. The pattern is established: file through the formal process, absorb the pushback, litigate the boundary. That is not a criticism. In a jurisdiction where the rulebook lags the product by a decade, filing first and arguing later is the only rational sequence available.
Now the company wants to list perpetual futures on single stocks. It filed the rule change with the SEC under SR-KALSHIEX-2026-02, with a parallel path through the CFTC. It proposes Kalshi Klear LLC as the clearing entity โ an affiliate, which makes the structure vertically integrated in the way CME and ICE are. It sets the margin at 15.5 percent rather than the 15 percent floor. It restricts the underlying universe with a market capitalization floor of one hundred billion dollars and a liquidity floor of four hundred and fifty million in average daily traded volume.
Read against the historical record, the ambition is enormous and the precedent is grim. Single stock futures have been legally permissible in the United States since the Commodity Futures Modernization Act of 2000 resolved the Shad-Johnson accord. They launched in 2002 on two venues. NQLX was dead by 2004. OneChicago lasted until 2020 and closed without ever achieving meaningful volume. The category has been tried, licensed, and abandoned.
So the question is not whether the structure is legal. It is whether the third attempt, wearing a funding rate and a crypto-native vocabulary, can answer the question the first two could not.
Core
Start with the anchor, because everything else is downstream of it.
A perpetual on a single stock needs a spot reference price to compute funding. Which price? A consolidated tape print? The primary listing exchange's official close? A volume-weighted index across venues? Each choice is a governance decision with direct P&L consequences, and the filing does not resolve it in public view.
For crypto this problem is chronic and mostly tolerated. Exchanges go down. Order books thin out. A wick on a shallow venue produces an index print that no one else saw, and the funding calculation inherits the ghost. I have been on the wrong side of exactly this. In 2022 I held a perpetual position in a mid-capitalization asset when its primary spot venue halted for maintenance. The perpetual kept trading. The mark did not move because the mark was waiting on a price that did not exist. For roughly twenty-two hours, accounts were valued against a stale number, funding clamped to its limit, and everyone who assumed the mark was real got liquidated on a print that never cleared anywhere else.
Equities do not have this problem occasionally. They have it structurally. Listed stocks have opening auctions and closing auctions, limit up and limit down bands, volatility halts, circuit breakers, and after-hours sessions that are thin enough to be moved by a single order. A perpetual on Apple has to price continuously through all of it. At 4:05 in the afternoon, which print does the funding rate anchor to? The official close? The last after-hours trade? The consolidated mid?
Whichever one Kalshi chooses, someone will be liquidated by that choice. That is not a bug in the design. It is the design.
Now the margin.
Fifteen and a half percent is approximately six and a half times leverage. For context, Regulation T requires fifty percent initial margin for long equity positions and one hundred and fifty percent for shorts. Portfolio margin accounts can reach fifteen percent, but only for hedged, diversified books, and the broker retains the right to raise requirements intraday. Single stock futures at established venues are margined through SPAN-style models that typically require twenty percent or more for the largest capitalized names, and considerably more for volatile ones.
The Kalshi number is not absurd on an ordinary Tuesday. The problem is that single stocks do not gap like indices. They gap like individual companies. Meta fell roughly twenty-six percent in a single session in February 2022. Netflix fell roughly thirty-five percent in a single session in April 2022. Post-earnings moves in the largest technology names have repeatedly exceeded ten percent in one session, and merger breaks, fraud disclosures and guidance withdrawals produce moves that no margin model calibrated on daily volatility two years earlier will capture.
Run the arithmetic. A customer posts fifteen and a half percent. The stock gaps twenty-five percent against them overnight. The account is at negative nine and a half percent of notional. That deficit does not vanish. It goes to the clearing member, and from the clearing member to the clearing house.
The second-order problem is the daily settlement. Marking to market once a day means the loss crystallizes at the close, but the gap already happened during the session. Between the last mark and the settlement mark, the account is impaired but not yet called. Without intraday variation margin, without an automatic deleveraging waterfall, without a live risk engine re-valuing positions on every print, the exchange is exposed for the entire duration of the move. The filing does not disclose whether any of that machinery exists.
I do not say this from a model. I say it from a book. During the 2022 drawdown I was holding Curve and Lido positions that I believed were diversified because they had different names. When the market broke, they behaved like one position wearing two labels. I did not use an algorithm to fix it. Over two weeks I manually cut leverage by forty percent, not because I had a forecast, but because I could not answer a single question: what is my loss if the mark is wrong for one day? If you cannot answer that question, your leverage is a rumor. The same question applies to every account Kalshi intends to onboard, multiplied by six and a half, aggregated into a clearing entity that has never cleared anything.
That clearing entity is the third piece.
Kalshi Klear LLC is an affiliate of the exchange. Vertical integration between a DCM and its clearing house is standard practice and not inherently a red flag. What is unusual is a brand-new clearing entity taking single-name gap risk as its first product line. Clearing is a capital business before it is a technology business. The clearing house is the counterparty to every trade. It needs a default fund, a waterfall that specifies who absorbs losses in what order, sufficient skin from the clearing members, and a governance structure that can act independently of the trading venue's commercial incentives.
None of that history exists yet. There is no loss experience, no calibrated model validated against real single-stock events, and no evidence in the filing about whether the board of the clearing entity is truly separate from the board of the exchange. If the trading venue and the clearing venue share a failure mode, they are not two institutions. They are one institution with two names and two ways to fail at once.
Then the definitional question, which is the only question that actually matters.
The Exchange Act's definition of a security future contemplates a contract for future delivery of a single security or a narrow-based security index. The architecture of that definition presumes a settlement date. There is a date of determination, a delivery obligation, a terminal event. A contract with no expiry does not obviously fit inside that architecture. And if it does not fit as a security future, the natural alternative classification is a security-based swap on a single security.
That distinction is not academic. It is binary, and it decides whether the product exists. A security-based swap must be executed on a registered security-based swap execution facility and cleared through a registered clearing agency under a separate regulatory regime. Kalshi's CFTC designation as a contract market does not confer SEC registration of any kind. If the SEC or the CFTC concludes that a no-expiry contract is properly a security-based swap, the DCM path does not merely become difficult. It ceases to exist.
So the entire product rests on one clause. Not on technology. Not on liquidity. Not on demand. On whether a regulator reads a sentence and concludes that a contract without an end is still a contract of future delivery.
And the likely outcome is not a clean answer. Under the rule change process, the Commission can approve, disapprove, or approve with conditions. It can also suspend and institute proceedings, which is a way of saying nothing while saying it slowly. The realistic scenarios are an approval order with footnotes constraining the no-expiry structure, or a reproposal that pushes the effective date past November 2.
Now the threshold design, which is where the hidden risk sits.
One hundred billion dollars of market capitalization and four hundred and fifty million of average daily volume. In the United States, that is a universe of roughly one hundred to one hundred and fifty names, and after liquidity screening, fewer. Apple, Microsoft, Nvidia, Amazon, Alphabet, Meta, Tesla, Broadcom, JPMorgan, Eli Lilly and their immediate cousins. That is the pool.
The design intent is obvious and, on its own terms, defensible. Restrict the universe to names where the spot anchor is deep and reliable, and the funding mechanism is less likely to be gamed. Concentrate the liquidity so the order book is not a vacuum.
But the same restriction creates the tail. Mega-capitalization equities are not independent bets. In a systemic drawdown they are one bet on one factor, expressed through a dozen tickers. In 2022 essentially the entire complex drew down together, on the same days, for the same reasons. Add six and a half times leverage and a daily funding transfer and you get correlated liquidation across the entire product line on the same afternoon, in a clearing house with no default history. Diversification does not exist in a universe where every name is a proxy for the same variable.
The counterintuitive point is that the best feature of the design is the thing most people cannot see.
The funding rate solves cold start. That is genuinely the hardest problem in a new listed product, and it is precisely what killed the previous attempts. OneChicago had no self-correcting convergence mechanism. It relied on market makers quoting both sides of a contract that nobody needed to hold, and when the flow did not arrive, the spread widened and the venue died.
A perpetual is different. The funding rate is a perpetual arbitrage invitation. If the contract drifts from spot, the carry trade pays you to close the gap. You do not need a designated market maker to inventory risk indefinitely. You need arbitrageurs with spot hedging access, and the mechanism recruits them automatically. That is elegant, and I will say so plainly. It is the only reason a no-expiry structure has a chance of accumulating open interest in its first year.
But elegance is not the same as fairness, and this is where I stop admiring the architecture.
The funding rate is not a market signal. It is a transfer. On offshore perpetual venues, the recurring pattern over time is that the crowd sits net long and pays the basis to a much smaller set of makers and funds. The perpetual does not create return. It transports it, from the many to the few, on a fixed interval, forever. Moving that mechanism inside a regulated venue does not eliminate the transfer. It invoices it. The funding becomes a cost of carry with a KYC file attached, plus a clearing fee, plus an execution spread, plus whatever the intermediary chain takes on the way through.
And here is the part that the compliance framing obscures. Aave's interest rate curve is a governance parameter chosen by token holders and adjusted by vote. It has almost nothing to do with real market supply and demand; it is an administrative price wearing the costume of a market price. Kalshi's funding rate is at least attempting discovery. But both instruments are administrative, both are set by a small number of people, and only one of them admits it. When you trade a perpetual, you are trading someone's parameterization of reality.
Contrarian
Everyone is watching the headline. The approval. The vocabulary. The effective date. The question nobody is pricing is the only one that generates P&L: who is on the other side of the funding payment, and who absorbs the loss when the anchor fails.
A perpetual is a zero-sum transport mechanism with fees attached. For every account that is up on a funded contract, another account is down by the same amount, minus the venue's cut. That is true whether the venue sits in Seychelles or New Jersey. The difference is not whether value is extracted. It is who is allowed to extract it and under what disclosure.
Onshore, the makers become registered dealers. The clearing member becomes a supervised institution. The transfer continues, but now it runs through a chain of regulated intermediaries who each take a slice and each bear a documented obligation. The retail account at the end of that chain is paying more, not less, for the same exposure it could have obtained offshore. What it buys in return is enforceability โ a counterparty that cannot vanish, a dispute process, a balance sheet behind the contract.
That is a real product. It is not a better product. It is a more expensive product with a legal identity. Anyone modelling this as a growth story rather than a substitution story is misreading the mechanism.
The second blind spot concerns the moat. Kalshi's advantage here is a regulatory time difference, not a technology. There is no proprietary algorithm in a funding rate; the formula has been public since 2016. There is no network effect in a matching engine. If this product is approved and works, CME has the clearing capital, the surveillance infrastructure, the customer base and the existing clearing members to replicate it in a single quarter. Kalshi is not building a business that compounds. It is buying an option on a rulebook, and the option expires in roughly twelve to twenty-four months.
There is a third thing, and it is the reason I keep circling the thresholds rather than the product. A hundred-billion-dollar market cap floor and a four-hundred-fifty-million liquidity floor are not neutral parameters. They are filters. They codify that only the largest issuers may have a regulated perpetual. The smaller, more interesting, more genuinely useful applications of the structure are excluded by arithmetic before anyone evaluates them.
I spent part of 2025 inside a compliance drafting exercise in London, translating regulatory language into rules a trading desk could actually follow. What I learned is that compliance costs are regressive. They are trivial for institutions with legal departments and fatal for everyone else. MiCA gave Europe the appearance of clarity and, in the same motion, set reserve requirements that functionally filtered small stablecoin issuers out of existence. Clarity that only the large can afford is not clarity. It is a licensing filter wearing the word transparency. The Kalshi thresholds are the same instrument in a different jurisdiction.
And a fourth point, smaller but stubborn. The word perpetual is doing emotional work on whoever reads it. No expiry sounds like no deadline, no roll, no decay. But the funding payment is the decay. It is simply paid in installments instead of crystallized at settlement. I have held a perpetual through a sustained funding regime that cost more over six weeks than the round-trip commission on a futures roll would have. Nobody who has actually carried that position believes the contract is free to hold.
Takeaway
Watch the vocabulary, not the press release. If the order and the accompanying releases describe the product as a security futures product, the contract market path survives and November 2 is a real date. If the word swap appears anywhere in the disposition, the path closes and no amount of liquidity planning matters.
Watch the clearing disclosure. Before launch, Kalshi Klear either publishes intraday variation margin, an automatic deleveraging waterfall, and default fund sizing, or it does not. Silence on those three items means the product is running an unhedged structural exposure, and the first single-name gap settles the argument in public.
Watch open interest at ninety days, not volume. A market with genuine hedging demand accumulates open interest that persists through quiet sessions. A market that is pure retail speculation prints enormous volume and negligible open interest, because nobody holds a position through the funding interval if they can avoid it. High volume with low open interest is the signature of a casino, not a market.
Holding the line when the world screams to sell is a discipline. Holding the line when the mark itself is a rumor is something else entirely. The mark is the only thing that is real. The filing has a date printed on it, and that date is the most honest sentence in the document.