The director of the longest-running election betting market in modern American politics used to work in a relatively sleepy field, presiding over what he called a “connoisseur’s market” where bets were capped at $500 (£388) and no one took. a lot of money.

But the world in which Thomas Gruca, a marketing professor at the University of Iowa’s Tippie College of Business, now operates has changed dramatically.

In recent weeks, as the U.S. heads toward a presidential election that most pundits and polls say is too close to call, a new crop of companies have burst onto the scene.

They are attracting hundreds of millions of dollars in bets on the outcome of the race — and attention from both campaigns and the media, as predictions on many of the biggest sites have tilted decisively in Republican candidate Donald Trump’s favor.

The frenzy was sparked in part in September, when a federal judge rejected arguments from U.S. market regulators that offering election trading to Americans violated state gambling laws and was not in the public interest.

The decision is being appealed. But since the ruling, more than $100 million has been wagered on Kalshi, the company that sued regulators, according to its website.

Other companies, including Interactive Brokers and popular stock trading platform Robinhood, have also entered the fray, joining sites that have long served customers outside the U.S. in countries like Britain, where betting on American politics is fair game.

At rallies, Trump has taken note of the sudden increase in election betting — and the increasing odds of his victory.

His most high-profile supporter, technology billionaire Elon Musk, has also drawn attention to the phenomenon. pointing out calls to his followers on social media earlier this month on the election betting markets, arguing that they were “more accurate than opinion polls because there is real money at stake”.

Prof. Gruca’s Iowa Electronic Markets has little in common with its new competitors.

The exchange is not wrestling with US regulators in court over its legality.

Under rules agreed with the government, it can accept bets for research purposes, but the bets are capped and the exchange is not allowed to advertise or make money from the activity.

It also monitors a small portion of the trade, mainly from Americans: a pool totaling less than $30,000, said Prof. Gruca.

There is another big difference.

Unlike larger platforms like Kalshi, Polymarket, Betfair and PredictIt, where the odds favor Trump by about 60% or more, traders in the Iowa market currently have their money on Kamala Harris.

Professor Gruca is proud of his stock’s track record: On average, through nine elections, the Iowa Electronic Markets has predicted the popular vote outcome within nearly a percentage point, proving to be a more accurate indicator than the polls.

So he’s surprised by the numbers on some of the bigger sites.

“The numbers are extreme – it’s a 50-50 race,” he told the BBC. “We’ve done this for 60 to 40 races, we’ve seen that before. This is as close or closer than anything we’ve ever seen.

The betting markets, which have a long history outside the US, have been wrong before: for example, they heavily discounted the odds of a Trump victory in 2016 and thought Republicans would do better in the 2022 midterm elections. would do than they did.

But academics say markets are often useful forecasting tools.

Still, Professor Gruca said the public should be skeptical of some of the platforms, noting that they have no track record and could be subject to manipulation given the huge sums of money involved and the risk that the pool of participants may not is deep. enough to match all bets.

“If you don’t have limits, it’s the deep pockets that influence prices,” he said. “Your opinion is weighed by the size of your checkbook.”

On Polymarket, big bets from four accounts controlled by a French trader earlier this fall helped tilt the odds for Trump.

Researchers have also told reporters that they have seen signs of wash trading at Polymarket – that the same people are buying and selling repeatedly, giving the illusion of more activity than there actually is.

Polymarket – which allows users to bet against each other on a specific future outcome and unusually works using crypto – did not respond to a request for comment from the BBC.

Its CEO, Shayne Coplan, has previously called the platform a “reality check” and “much-needed alternative data source,” noting that it correctly predicted President Joe Biden would drop out of the race.

Others have a darker view.

After the court’s ruling in September, the watchdog group Better Markets warned that allowing such bets would “compromise the integrity of our elections, cause market manipulation and victimize countless investors.”

Professor Gruca said he thought election betting was an “afterthought” when it came to threats to democracy.

But the debate on this issue is likely to continue.

In the US, states regulate betting and some have banned election betting entirely.

Prediction markets — which differ from gambling because they rely on trading rather than a centralized company overseeing the odds — fall under the jurisdiction of federal market regulators, which have long taken a dim view of such activities.

But gambling standards have relaxed in the US, especially after the 2018 Supreme Court ruling that paved the way for sports gambling.

In response to an increase in applications from companies seeking to offer trading on events, the Commodities Futures Trading Commission (CFTC) proposed a rule in May that would explicitly prohibit trading on political contests.

“Contracts involving political events ultimately commodify and degrade the integrity of the unique American experience of participating in the democratic election process,” CFTC Chairman Rostin Behnam said at the time.

He warned that allowing such trading would push the CFTC beyond its mandate and expertise, and into the role of “electoral agent.”

Ultimately, resolving the question could be another election bet.