U.S. Election Betting: Kalshi’s Polymarket Rival Quickly Gets Traction

This week in prediction markets:

• Presidential betting passes $30 million on Kalshi.

• Trump name-drops Polymarket – or tried to, anyway.

• Arb opportunity? There’s a discrepancy between Polymarket’s individual swing state markets and a contract about all of them.

This U.S. election cycle will be over in two weeks. For some, it will be a sigh of relief. But prediction market platforms are having the time of their lives.

The market gives Republican nominee Donald Trump a 14 percentage point lead over his Democratic rival Kamala Harris. These odds must come solely from U.S. nationals (and permanent residents), because Kalshi’s terms and conditions prohibit foreign nationals from trading on the platform.

By contrast, the industry leader Polymarket prohibits U.S traders from the platform but does not require a know-your-customer process to validate who is who. Polymarket’s bets are denominated in cryptocurrency, whereas Kalshi’s are settled in regular dollars.

Initially, some might have thought that Kalshi was coming to market with a handicap, because it’s regulated. But this has allowed the platform to escape the allegations of market manipulation by shadowy, possibly foreign, dark money that have dogged Polymarket.

Instead, Kalshi created a rebuttal to such accusations for its much larger competitor.

In a thread on X (formerly Twitter), Tarek Mansour, Kalshi’s co-founder, outlined how market manipulation falls flat on his platform, where, as on Polymarket, Trump leads with a commanding margin.

“The median bet size on Harris is LARGER than the median bet size on Donald Trump,” he wrote. “Trump’s odds are not the result of a few people pushing the odds up. It’s the opposite.”

Looking through Dune, a crypto analytics portal (Kalshi wouldn’t be on Dune as it’s not a crypto platform), the bet size numbers aren’t too far off on Polymarket.

Data shows that overall on the site the median betting size is between $10 and $50.

Now, for Polymarket’s presidential election contract, median data isn’t immediately available on Dune. The next best thing is the average bet size (remember the difference between average and median) which comes in at $210 for Trump and $113 for Harris.

Going back to the overall Polymarket platform, the number of users who have placed bets above $100,000 is only 27 – while there are over 220,000 users on the platform.

Some of those 27 users are placing $20 million-plus bets that are bound to put pressure on markets. But the fact that Polymarket’s numbers are in line with Kalshi’s on Trump suggests that these whales aren’t having the effect that critics of prediction markets think they are.

‘Polypoll’ and swing states

Sometime between launching a crypto token that’s off to a rocky start and working a shift at McDonald’s (the employer of last resort for failed crypto bros) Trump name-dropped Polymarket at a rally, calling it “Polypoll.”

This isn’t the first time the Republican candidate has acknowledged Polymarket, as his team posted favorable odds on his Truth Social page during his first surge this summer.

While Polymarket’s election contract has Trump well ahead of his Democratic rival at 62-37, some of the numbers aren’t adding up.

Individual state-level contracts have Trump winning every swing state, even Nevada, with a healthy margin.

But a contract asking if he’ll win every swing state is coming in at just 32% on the ”yes“ side.

The largest holder on the “no” side, who goes by the handle “AntiMaga,” has $5 million in positions, largely taking the “no” side on Trump or Republican-related contracts.

Meanwhile, the largest holder on the other side of the bet, Yeueu, is holding much less in terms of positions overall, but is up 88% on the swing state contract.

All of this is a reminder that markets aren’t perfect, and inefficiencies exist – or arbitrage opportunities, depending on your perspective.