On October 9, the Commodity Futures Trading Commission filed a rule classifying event contracts as swaps under the Commodity Exchange Act. Five days later, Kalshi CEO Tarek Mansour took the DC Fintech Week stage. The five-day gap is the story. The rule opens a 30-day public comment window, and Mansour used his slot to reframe the debate before that window shuts. That is not a product update. That is a lobbying position delivered as a keynote.
The more revealing line came later. Pressed on persistent claims that Kalshi's trading volume is inflated, Mansour attributed the discrepancy to market makers submitting fixed-size orders. Read that carefully. The chief executive of a CFTC-regulated exchange just conceded that the number the industry quotes as his market share may be double-counted at the order-book level. No one in the room asked for the audit.
Check the code, not the hype.
Prediction markets let users buy contracts on event outcomes — elections, sports, weather, central-bank decisions. Price becomes probability. The concept is decades old. The regulatory clarity is not.
Two platforms define the landscape. Kalshi operates as a Designated Contract Market, a CFTC-licensed venue inside the United States. Polymarket operates on-chain, globally, with a hybrid order-book and AMM model, and has historically been fenced out of the US market. Same asset class, opposite structural bets. Kalshi wagers on compliance. Polymarket wagers on reach.
The conflict underneath is constitutional, not commercial. The CFTC argues event contracts are swaps, granting it exclusive federal jurisdiction. State regulators — New Jersey most aggressively — argue sports-related event contracts are gambling under state gaming law. The federal appellate courts have split: one panel sided with Kalshi, two sided with the states. New Jersey has petitioned the Supreme Court. The NFL has aligned with the states, which tells you where the lobbying capital sits.
Mansour's framing is clean: it is in-shore versus offshore. Either the United States builds a regulated domestic venue, or users route to unregulated platforms beyond its reach. It is a good line. It is also a rhetorical device that converts a messy jurisdiction fight into a binary choice, and binaries win public comment periods.
There is a second, quieter story that the framing buries. Kalshi runs two businesses, not one. Event contracts are the regulated franchise. BTC and ETH perpetual futures are a separate product line that drops the venue directly into the most crowded red ocean in crypto derivatives. The company is not only fighting for a new category. It is also competing in an old one, against venues with deeper books and no jurisdictional overhang.
The macro backdrop matters more than the venue-level detail. Prediction markets are arriving at the exact moment institutional capital is being absorbed into crypto through spot ETFs. That flow creates something the category needs and cannot manufacture on its own: a stable, regulated liquidity base. The same institutions buying BTC exposure through a wrapper are the natural counterparties for event contracts on rate decisions and macro prints. Kalshi's real bet is not on retail gamblers. It is on the desk that wants to hedge a CPI print without touching an offshore book. That is a defensible thesis — and it is also why the jurisdiction question, not the product, is the whole trade.
Start with what the coverage does not contain. There is no matching-engine spec, no settlement architecture, no oracle disclosure, no throughput figure. Kalshi's technical stack is absent from the conversation entirely. That absence is itself data.
Kalshi's defensive moat is institutional, not technical. It is a CFTC license plus a run of court precedent. The matching engine, the resolution logic, the data feeds that determine who gets paid — all undisclosed. When I audited the dependency chains of three mid-cap DeFi protocols during the Terra collapse in 2022, I found hardcoded integration expirations that had already lapsed while the protocols kept operating without an emergency pause. The lesson was not that the code was bad. The lesson was that the entire position rested on a dependency no one had mapped. Kalshi has the same shape. Its dependency is not a stablecoin. It is a court docket, and dockets do not publish a changelog.
Now the volume question, because that one is auditable. Mansour blames fixed-size market-maker orders for inflating the tape. If accurate, that is a reporting convention, not fraud. Market makers post standing quotes; when a venue sums gross order flow rather than net executed volume, the count inflates. This is a known distortion in order-book venues. In 2020, I scraped historical TVL and borrow-rate data across Aave and Compound to test whether the super-yield pools were real. They were not. Most were arbitrage traps that decayed within weeks, and the divergence only appeared once the data was reconciled across sources. The method applies here: do not trust a venue's self-reported number. Trust the reconciled number. Kalshi has not published one.
Then the structural tell that the insider-trading cases expose. The CFTC has brought enforcement against two Kalshi users — Santos and Perez — with fines, disgorgement, and three-year bans. Kalshi cites these cases as proof of its compliance capability. Read them the other way. In the Santos matter, the platform reportedly received roughly 100 complaints within minutes. Two facts fall out of that single number. First, Kalshi has a real, engaged user base — that is the strongest bullish signal in the entire story, and it is buried under the volume dispute. Second, outcome resolution depends on a central operator's judgment, not a verifiable feed. Every contract on the venue settles through the same centralized determination that made the manipulation possible in the first place. The security model and the settlement model are the same object.
That object has a cost curve. Manipulation detection scales with user count, not against it. A venue with a hundred complaints in minutes is a venue where the monitoring burden grows faster than revenue, unless the controls automate. Kalshi has not disclosed whether they do. The compliance story it tells investors is, at the code level, an unverified claim.
There is a liquidity problem hiding under all of this, and it is the one prediction-market bulls never price. Every event contract is a separate market. Election-night liquidity does not carry over to a weather contract or a rate decision. That fragmentation means the category cannot inherit the network effects of a single deep book. Each new market restarts the cold-start problem. Polymarket solved a version of this with tokenized liquidity incentives during its growth phase; Kalshi, as a licensed venue with no token, cannot. Its answer has to be market makers, which loops straight back to the fixed-size order issue inflating the tape. The two problems are the same problem.
One more layer deserves scrutiny. Event contracts scale only if settlement sources are trustworthy, and trustworthiness at scale means infrastructure — oracles, resolution feeds, dispute mechanisms. The insider-trading cases exist precisely because the resolution step is where value concentrates. If the domestic category legalizes, the demand for verifiable settlement infrastructure rises with it, and that demand lands on oracle providers. I have been skeptical of oracle decentralization for years; a network that solves trust by appointing a committee of nodes is not solving trust, it is relocating it. But relocated trust with an audit trail beats a centralized judgment call with none, and the prediction-market boom will make that gap visible. Watch the settlement layer, not the venue.
I track narratives the way I tracked NFT collections in 2021 — with a decay rate, not a vibe. The prediction-market narrative is currently in an acceleration phase, driven entirely by regulatory milestones: the CFTC rule, the court split, the Supreme Court petition. Strip those out and what remains is a platform with an undisclosed stack, an unaudited volume figure, and a business that depends on a ruling it does not control. Acceleration phases end. The question is what the narrative decays into — a regulated category or a footnote — and that answer is being written in a courtroom, not a codebase.
Data over drama. Always.
The market is pricing this as a technology story or a product story. It is neither. It is a jurisdiction story, and the jurisdiction story has a counter-intuitive edge the bullish coverage misses.
Consider the CFTC's swap designation. Kalshi treats it as a win because swaps route under the Commodity Exchange Act and sidestep the SEC's securities test. The Howey analysis mostly supports that — no common enterprise, no profits from a promoter's efforts. But mostly is doing heavy lifting. Swaps classification sits adjacent to securities law, and an aggressive SEC or a state attorney general can contest the boundary. A win that depends on the other side declining to litigate is not a moat. It is a temporary truce.
Then there is the enforcement record. Kalshi markets the Santos and Perez cases as evidence of self-policing. But a venue that needs a hundred-complaint firehose to surface manipulation does not have a security moat. It has a monitoring cost, and monitoring costs scale with volume. The very growth that makes the venue valuable makes the control problem harder.
There is also a blind spot in the comparison everyone draws. Kalshi versus Polymarket is not a competition between two platforms. It is a competition between two regulatory regimes, and the offshore side is not standing still. Polymarket has already moved on insider-trading controls of its own. If the US category legalizes and Polymarket gains domestic access, the compliance premium Kalshi banks on evaporates, because the premium was never a feature — it was a function of exclusion. Kalshi's differentiation is a function of who is banned, not of what it built.
And the bear-market angle deserves its own line. Prediction markets look like a haven because they do not depend on crypto prices. True. They depend on something more fragile: a Supreme Court docket and a 30-day comment window. In a bear market, traders reach for assets with stable cash flows. A venue whose terminal value is a legal ruling has no stable cash flow. It has a conditional one.
Audit trails do not lie. Narratives do.
Three signals will resolve this, and none of them are on a chart. Watch the CFTC comment period for whether the swap classification survives intact. Watch the Supreme Court docket for whether New Jersey's petition is accepted — that is the binary event that either unlocks or closes the domestic category. And watch for any third-party reconciliation of Kalshi's volume, because the number everyone cites has never been audited.
The prediction market does not sell probability. It sells jurisdiction. Until someone reconciles the tape and the courts finish their work, the only honest position is to size the trade to the ruling, not the narrative.
