This week offered a rare, telescoped view of an industry in transition. Inside the markets, quant firms are deploying bots that have largely pushed retail arbitrageurs out of the most profitable strategies — making prediction markets look increasingly like every other corner of Wall Street. Outside the markets, the regulatory war is intensifying on every front: new proposed rules from the CFTC, a new legal challenge from a former CFTC chair, a discriminatory tax in Kentucky, and an estimated $10 to $27 billion in illegal American trades flowing offshore anyway.

We lead with the arbitrage story not because it's the biggest headline, but because it's the best lens. When sophisticated quant firms start treating prediction markets as hunting grounds for algorithmic profit, the category has crossed a threshold. The debate over whether prediction markets are "real" financial instruments is over. The new debate — and this week's five stories are all chapters in it — is over who controls them.

The answer is genuinely unsettled. Regulators, states, courts, and former regulators are all pulling in different directions. The market itself is not waiting for a resolution.

The Free Money Is Almost Gone

A New York Times investigation this week profiled a 25-year-old former actuary in Des Moines who has made more than $1 million betting on prediction markets. Ryan Noel doesn't care about sports at all — he cares about math. His strategy is arbitrage: when Kalshi and Polymarket post different odds on the same event, you can buy both sides of the bet and guarantee a win regardless of outcome. If the prices diverge enough to cover fees, you profit either way.

The strategy works because markets aren't perfectly synchronized. Casual users moving a market slightly out of line create the opportunity; arb traders pounce and even things out. But the window is closing fast. Wall Street quant firms like Susquehanna International Group are now deploying automated bots that spot and close these gaps in seconds. What used to be 30-second opportunities at 8% margins have compressed to two-to-five seconds at 4-5%.

"You're not betting against Joe Schmo anymore. You're betting against a quant firm with infinitely better technology than you."

Alex Monahan, founder of OddsJam

A Wall Street Journal analysis found that one-tenth of the top one percent of Polymarket accounts rake in more than two-thirds of all profits. The free money is almost gone — and when it's completely gone, prediction markets will look like every other liquid financial market: efficient, institutional, and very hard to beat.

Selig Draws the Lines

On Wednesday, the CFTC published a 267-page proposed rulemaking on event contracts — the agency's most significant formal step yet toward defining what prediction markets can and can't offer. Then Semafor published a wide-ranging interview with Chair Mike Selig that read like a victory lap.

As reported by Axios, the proposed rule draws clear lines on sports contracts:

Allowed:

  • Final scores, point differentials, win-loss results, tournament advancement

  • Season-long performance metrics and individual player stats

  • Skill-influenced games like poker tournaments

Not allowed:

  • Single plays (one pitch, one shot, one foul)

  • Injuries, officiating decisions, pre-collegiate sports

  • Casino-style games of pure chance (roulette)

  • Contracts on terrorism, assassination, or war

On states that have tried to regulate or ban prediction markets outright, Selig was unambiguous. The CFTC is already suing Minnesota; he's watching others. He'd welcome a Supreme Court ruling on jurisdiction, calling the agency's authority "black-letter statutory text." Congress is free to change the statute — but he doesn't think they need to. People raising concerns, he added, "are the same people that failed to take action under the last administration."

Kentucky Tries the Tax Play

While most states have responded to prediction markets by attempting to ban them, Kentucky took a different approach: tax them into submission. In April, the state enacted a 14.25% excise tax on prediction market operators' transaction fees — the first state-specific tax of its kind on federally regulated derivatives.

The Coalition for Fair Markets — Kalshi, Polymarket, and Crypto.com — filed suit Friday, calling the law discriminatory, unconstitutional, and preempted by federal law. The comparison is pointed: Kentucky taxes horse track wagers at 9.75%. Federally regulated prediction markets are being taxed at nearly 50% more.

"No State currently levies a State-specific excise tax of any kind on derivatives transactions that take place on a federally designated exchange, let alone the sort of specifically targeted and discriminatory tax that Kentucky has imposed here."

Coalition for Fair Markets lawsuit filing

Attorney General Russell Coleman, leaning hard into the gambling framing, vowed to fight back. The outcome matters beyond Kentucky: even if states can't ban prediction markets outright, a sustained tax lever could be just as effective.

The Guy Who Wrote Dodd-Frank Disagrees

Gary Gensler ran the CFTC from 2009 to 2014 and oversaw implementation of Dodd-Frank — the same law the current CFTC is using to claim jurisdiction over prediction markets. This week, he filed an amicus brief in the Sixth Circuit Court of Appeals arguing that Dodd-Frank never gave the CFTC authority over sports wagering, directly contradicting Chair Selig and Kalshi's own legal position.

"If Dodd-Frank had preempted the states on sports betting, it would have been one of the biggest stories about Dodd-Frank at the time. But nobody ever mentioned it."

Gary Gensler, amicus brief

The case stems from Kalshi's lawsuit against Ohio after the state ordered the platform to stop offering sports contracts. The CFTC has backed Kalshi in that fight. Gensler is now telling the appellate court that both are wrong — and that the CFTC lacks the staff and experience to regulate sports betting even if the statutory authority existed. A former agency chair filing against his own former agency's legal position is unusual enough to matter. The Sixth Circuit will have to decide who's reading Dodd-Frank correctly.

$10 to $27 Billion Going Offshore

While regulators fight over who controls prediction markets, a new study suggests Americans have already voted with their wallets — by going around the rules entirely. Rutgers statistician Harry Crane estimated this week that approximately 30% of all trading volume on Polymarket's crypto-based offshore platform comes from US-based users, people accessing a platform they're legally banned from, typically via VPN.

The scale is striking:

  • $10.6 to $26.7 billion in estimated US trading volume on Polymarket's offshore platform from May 2025 to April 2026

  • ~50% of activity in Polymarket's sports markets estimated to be US-based

  • $133 billion — projected US offshore trading volume by 2030 if current trends hold

Polymarket has operated a separate, licensed US app since December 2025 — but that platform's April volume was $1.6 billion, compared to $9 billion on the offshore version. Polymarket declined to comment. The CFTC said it would exercise extraterritorial jurisdiction for bad actors on a case-by-case basis, but isn't chasing every VPN user.

The regulatory fight over prediction markets is happening against the backdrop of a market that's already enormous — and mostly invisible.

If someone forwarded this to you, subscribe at PredictionMarkets.media

Forward to a colleague who tracks markets, policy, or financial innovation.

Sources:

Keep Reading