Kalshi's New Sector Markets Step Out of the SEC's Lane, and Into a Data Vendor's

Editor's note: TrueEdge builds odds and prediction-market tools and earns affiliate commissions from licensed sportsbooks. We have argued in labelled commentary against criminalising federally regulated exchanges. This piece argues a position on prediction-market regulation; weigh it with that in mind.

Kalshi on Friday launched what it calls sector KPI markets: event contracts on industry-wide credit-card spending, point-of-sale data, foot traffic and app downloads. They settle on data from Carbon Arc, a company founded in 2021 that describes itself as a "neutral data layer for the AI economy." Kalshi's example is total downloads across the AI app category rather than downloads of any one app. Its suggested customers are a retailer hedging a slowdown in consumer spending, a restaurant operator hedging foot traffic, and a developer hedging weaker adoption across a category.

"Atomizing equities into specific events and KPIs is one of Kalshi's best use cases," Mansour said in the announcement. Carbon Arc's Kirk McKeown said his company was built to be "the pipes, plumbing, permissions and payment" for that data.

The more interesting thing about this launch is when it happened and how it is built. It also shows the exchange trading one regulatory problem for a different one.

The SEC got there first on single companies

On October 13 the Securities and Exchange Commission approved Cboe's plan to list binary options on the key performance indicators of 23 public companies, up to 100 contracts in its original filing, with Robinhood as the first retail broker. The examples include Nvidia's edge-computing revenue, Coinbase's transaction volume and Tesla's vehicle deliveries. Trading cannot start until Cboe's clearinghouse is registered, and the SEC says standard securities law applies to the contracts.

Kalshi had objected on August 5, arguing that approval "would prejudge the answers to several questions expressly posed in the Joint Product RFC," the request for comment the SEC and the CFTC published in June on where swaps end and security-based swaps begin. The SEC approved anyway. Chairman Paul Atkins' statement said the order "does not address — and does not foreclose — CFTC jurisdiction" over similar contracts on the same metrics.

That sentence is a polite truce. It does not settle anything. Kalshi already lists company-level KPI markets through the Public Companies Hub it launched in August ("Earnings are the corporate Super Bowl," its product lead said at the time). Those are the contracts with the clearest overlap with Cboe's.

Where the statute draws its line

The Securities Exchange Act defines a security-based swap to include a swap based on "the occurrence, nonoccurrence, or extent of the occurrence of an event relating to a single issuer of a security or the issuers of securities in a narrow-based security index, provided that such event directly affects the financial statements, financial condition, or financial obligations of the issuer." That is 15 U.S.C. § 78c(a)(68)(A)(ii)(III), and it is the SEC's statutory hook into event contracts.

A contract on Tesla's quarterly deliveries is very close to that language: one issuer, and a number that ends up in its revenue. Seward & Kissel's August note on the question names customer metrics and production milestones as issuer-specific events that potentially implicate the framework, while noting that whether they "directly affect" an issuer's financials is exactly what the agencies are asking.

A contract on credit-card spending across a whole retail sector, measured by a third party's data, is much further away on every element. It is not about a single issuer. Many of the businesses in an "AI app" category do not issue public securities at all, so it is hard to call the basket an index of securities, narrow or otherwise. And the settlement figure is not an entry in anyone's financial statements. It is Carbon Arc's estimate of activity.

Kalshi has not said that jurisdiction drove the design, and we are not claiming it did. But a company that formally objected to the SEC moving into KPI contracts in August has launched, ten days after losing that argument, the version of the product that the security-based-swap language reaches least well. A company whose sports revenue share has fallen to about 65 per cent from about 95 per cent nine months ago, per Mansour this week, needs the non-sports business to be clearly within the CFTC's jurisdiction.

The strongest case: this is just the better product

The best objection is that the sector design is about the customer, not the regulator.

A regional grocery chain does not care whether Walmart beats its comparable-sales estimate. It cares whether consumers in its markets are cutting back. Sector-wide spending is the exposure an operating business actually has. Single-company KPI contracts are mainly a way to trade around earnings, and they are the ones with the obvious insider-trading problem, because the company's own employees know the number before anyone else. In a sector aggregate no single company's staff can see the whole number. On this view Kalshi has built the more useful hedge and the less abusable contract, and the SEC question is a side effect.

We think most of that is correct, especially the hedging point. Nothing in the CFTC's swap definition fits better than a contract a restaurant operator could use against a fall in foot traffic. If any event contract has "a potential financial, economic, or commercial consequence," this one does.

The insider point is where the argument changes rather than ends. A sector contract removes one set of insiders and adds another: the people who build the settlement number. Carbon Arc's description, as Kalshi's announcement gives it, is that it sources hard-to-get company data and clears it through a compliance layer. That means the data passes through a supplier chain before it is published, and the people along that chain see it early. Kalshi's announcement does not say when Carbon Arc's figures are fixed, who sees them before settlement, or whether anyone at Carbon Arc or its suppliers is barred from trading. Kalshi has the tools for the job. Since June it has required employment information from traders in high-risk markets and bars presumptive insiders, and it told the AP that the screening had stopped at least 100 potential insider trades. The open question is whether a data vendor's staff count as insiders under that screening, and the launch post does not say.

Who checks the number

The second change is about verification. When a Cboe contract on Tesla deliveries settles, the number comes from a company filing that is public, audited and carries legal liability. When a Kalshi sector contract settles, the number comes from a private dataset that most traders cannot see in full and cannot reproduce.

The CFTC has already written down why that matters. Its September 22 advisory on mention markets warned about settlement that "may be neither independently generated nor externally verifiable." The phrase was aimed at contracts on what a person says, but the concern is general. Exchanges have always used private benchmarks, and a commercial data source is not suspect for that reason alone. But the CFTC's concern is a single source that nobody outside the vendor can check, and a sector KPI contract settles on exactly that.

The design does move Kalshi's corporate-metrics business toward the CFTC's side of the statutory line. It also stakes that business on the integrity of Carbon Arc's data. That trade could be worth making. But a contract whose result only one vendor can check needs more disclosure than a press release about methodology, timing and who is barred from trading. The CFTC, which is working toward a final event-contract rule, should ask for it before the markets get large.