By Douglas Gillison
WASHINGTON, June 8 (Reuters) – Prediction market watchdogs may struggle to police betting on the U.S. midterm elections, with thousands of races offering ever more ways for insiders to make a quick buck on a rising number of platforms, according to experts and new data.
A surge in suspicious trades on Kalshi and Polymarket, the two largest players, has stoked fears that these fast-growing markets are creating new avenues for insider trading.
Kalshi in April suspended three congressional candidates for betting on their own races, and regulators are investigating whether former congressman George Santos engaged in potential insider trading on the platform.
That may prove to be the tip of the iceberg, said insider trading experts. With at least 6,590 state and federal legislative seats up for election this year, an expected betting bonanza will test safeguards in a novel marketplace where insider trading remains a legally murky concept and its regulator is stretched, the people said.
“We may see a slow response or we may see no response if and when insider trading happens in the midterms,” said Ilya Beylin, a professor at Seton Hall Law School who has studied prediction market oversight. Suspicious trades would undermine Americans’ faith in democracy at a time when polls show many already worry the system is in danger, he added.
Kalshi, Polymarket and the Commodity Futures Trading Commission, the commodity derivatives regulator which is pushing for jurisdiction over prediction markets, say they are equipped for the challenge with various monitoring tools and other resources.
The platforms are also bolstering controls, with Kalshi blocking election trades by politicians and campaign workers and Polymarket cracking down on trading on private information. In April, the U.S. Senate banned members and staff from prediction market betting.
Combined monthly global trading volumes on Kalshi and Polymarket surged nearly fivefold from September to reach about $24 billion in April, compared to roughly $14 billion a month wagered through legal sportsbooks in the United States last year on average, according to Pew Research Center.
“As innovation expands the reach and complexity of our markets, we are continuing to grow alongside it, both operationally and technologically,” a CFTC spokesperson said, adding the agency will “enforce the law aggressively.”
A lawyer for Santos declined to comment.
GROWING INFORMATION ASYMMETRY
Prediction markets allow traders to buy and sell binary “yes” or “no” contracts on the outcome of an array of events.
Traders can bet, for example, on which party will control Congress or a state legislature, as well as on individual races, and events around those contests and their candidates.
In addition to roughly 470 congressional races, 6,122 state and territorial legislative seats are up for election this year, according to the National Conference of State Legislatures, on top of local races for district attorney, mayor, the judiciary and other offices. Each candidate generates insiders – campaign workers, pollsters, fundraisers, donors, friends and family.
While insider trading is banned in commodity derivatives markets, there have been relatively few cases, said legal experts. And when it comes to elections, there is a smorgasbord of potentially relevant nonpublic information, from unpublished polling data to a brewing scandal, some of which may not be well-understood by regulators, said Beylin.
“They will need to learn, and that learning process often involves trial and error,” he added.
That pool of potential inside information is expanding as election bets become more esoteric. During 2024, a major election year, Polymarket listed 1,293 related markets, with $7.26 billion in trading volume, according to the Anti-Corruption Data Collective, a nonprofit research group.
With few major contests last year, the number of U.S. election markets on Polymarket fell, but the ratio of markets to races rose seven-fold to 17.4, indicating election markets are becoming more granular, according to Michelle Kendler-Kretsch, an ACDC researcher.
Increasingly, they focus less on the winners and losers, and more on the variables of a race, such as voter turnout, margin of victory and when candidates may drop out, according to ACDC’s analysis, which has not previously been reported.
Last year, for example, Polymarket listed multiple markets on the share of “inactive” ballots, or those ranking candidates who have since been eliminated, in the Democratic primary for New York City mayor.
That trend “creates a more significant information asymmetry, while the number of potential insiders grows and the risk of insider trading increases,” said Kendler-Kretsch.
BOLSTERING CONTROLS
Besides Kalshi and Polymarket, there are at least four other U.S.-authorized platforms offering election contracts, a handful of brokers providing access to the products, and several other players hoping to launch. As with traditional markets, the companies are the first line of defense.
Polymarket recently launched a U.S. operation, but its main exchange is not U.S.-regulated and has not generally mandated “Know Your Customer (KYC)” identity checks. While it bars U.S. residents, authorities have flagged concerns that those controls can be easily bypassed.
“We maintain a comprehensive market integrity framework” and the company is focused on transparency, a spokesperson said.
Thanks to internal surveillance, Polymarket has referred nearly 100 user wallets to law enforcement, including one allegedly used by a U.S. soldier to place an inside bet on the removal of Venezuela’s Nicolas Maduro, the spokesperson said.
Kalshi defines an insider as anyone in a position to directly influence a contract’s outcome, said Robert DeNault, the company’s head of enforcement. Aside from KYC checks, Kalshi uses public records to identify federal politicians and campaign staffers before they trade, and plans to do the same for local elections where data is available. It also monitors trades for anomalies.
It can then home in on a subset of suspicious activity, said DeNault. “You can leverage the tools … to gather a lot of information,” he said. Although the workload can be large, it is manageable, he added.
As powerful as such tools are, each lead has to be investigated by humans, said Aitan Goelman, who was CFTC enforcement director from 2014 to 2017. The cases Kalshi flagged in April, for example, showed humans had to conclusively identify the users.
With CFTC enforcement staffing at 105 positions, which budget data shows is its lowest level for at least 20 years, and many experienced investigators having left, the agency may not have the manpower to probe a lot of referrals, said Goelman.
The agency spokesperson said it was relying on experienced personnel and had been hiring continuously since December.
Goelman, though, argued its resources were “not even close” to meeting the agency’s needs.
(Reporting by Douglas Gillison in Washington; editing by Michelle Price and Nick Zieminski)