How ‘amateur gamblers’ are outperforming Wall Street experts

Economists at major banks and investment firms, who are paid high salaries to predict the direction of the economy, had expected the latest jobs report released on Wednesday to show the creation of about 68,000 jobs last month.

A group of anonymous online bettors, who place bets on forecasting website Kalshi, expected to see 54,000 new jobs.

The report ended up showing that the US economy added 130,000 jobs at the start of the year. Both groups got it wrong by a wide margin — and to similar degrees.

In Kalshi’s five years of existence, its thousands of bettors have proven to be as accurate, on average, in predicting certain economic indicators as highly trained experts, a working paper published last month by the National Bureau of Economic Research revealed. The group is also quite good at predicting the Federal Reserve’s interest rate decisions, and even better than professionals at forecasting the inflation rate.

“Getting information from a large group of people can be a remarkably good form of forecasting,” said Jonathan Wright, an economics professor at Johns Hopkins University and co-author of the paper.

Thomas Simons, an American economist at Jefferies, noticed when Kevin Warsh was leading the prediction markets to be Donald Trump’s nominee for chairman of the Federal Reserve. Simons had ruled out that possibility because of Warsh’s previous advocacy for higher interest rates rather than the lower rates that Trump prefers.

“‘How could he be in front of this? It doesn’t make sense,'” Simons recalled.

But the markets were right, and he decided he shouldn’t ignore the odds. Bettors, he realized, have an advantage: They don’t need to make a prediction if they’re not very confident they’re right. Professional experts have no choice; Even when faced with confusing data, and without much conviction in the number, they venture a guess.

“You have to predict these numbers every month, even when you’re not really sure,” Simons said. “So it makes me think that, if I go back to my assumptions, the people who are most sure are the ones who will participate.”

Another working paper, from economists at London Business School and Yale University, found that Polymarket bettors, overall, forecast corporate profits more accurately than analysts paid to advise investors on whether to buy or sell.

Theis Jensen, a Yale professor who worked on the paper, believes that the relatively good performance of the thousands of amateurs is due to incentives. Professional analysts may have conflicts of interest, such as their firms’ trading commissions may increase in response to optimistic forecasts. Analysts can also avoid publishing nonstandard earnings forecasts, which can cause more embarrassment than following the crowd.

“The nice thing about prediction markets is that you have to put your money where your mouth is,” Jensen said, “and it really encourages declaring your true beliefs.”

Of course, this has been true for decades. The first online prediction markets emerged in the early 2000s. Sites like Intrade focused primarily on elections and the likelihood of other world events, and were generally considered to be quite accurate. In the 2010s, American regulators cracked down on these sites, declaring them illegal gambling platforms.

But some platforms have continued to operate in Europe, where political and economic contracts are a complement to huge sports betting volumes. The same still goes for Kalshi, which won a lawsuit that allowed it to operate legally in 2024, and for Polymarket, which is only sporadically accessible in the US as lawsuits have blocked trading in several states.

Still, the volume of betting, even on non-sports issues, has grown at such a rapid pace that experts and analysts are paying attention. On any given day, more than $60 million is at stake across platforms on political and economic issues — far more than the first sites achieved.

Edward Ridgely runs Stand, a company that allows punters to trade simultaneously on Kalshi and Polymarket and follow other top traders. He said many of his highest-volume clients work in the same areas they bet on. A user in Hong Kong buys and sells Nvidia shares in his daily work and uses fee-related prediction market contracts as a hedge.

“If Trump’s tariffs escalate against China or something, he can get out of the position and not be harmed,” Ridgely said.

He sees other evidence that gamblers specialize: Most aren’t good at everything. “You can see that a lot of traders who are very good at elections are not good at crypto. Or, if you are very good at crypto, you are not good at geopolitics,” he said.

Michael Feroli, chief U.S. economist at JPMorgan, has access to a wealth of knowledge from the bank’s political affairs team, country experts and equity researchers. But he still watches the markets to get a more accurate estimate.

Also read: “The first year is the one where you work and don’t get paid”, says giant founder

“When you talk to people in Washington, they say, ‘Well, I think they’re going to pass the budget.’ So what is the probability?” Feroli said. “It’s a different language. You often have to insist to get a quantitative answer.”

On the quantitative questions that are his focus, such as predicting changes in the consumer price index and gross domestic product, Feroli suspects something else is going on: betting markets are just following the experts. This could mean monitoring the Bloomberg consensus, reading research from major investment houses, or tracking the futures markets and investor expectations that groups like the Chicago Mercantile Exchange already aggregate.

Tara Sinclair, an economist at George Washington University who studies forecasts, agrees that’s likely. And therein lies a danger in prediction markets: if the crowd replaced professional experts, individual bettors would lose.

“They would be making their employees’ jobs more difficult, because now they have individual sources of information to consult,” Sinclair said. “If they replaced all of that, they would have nowhere else to turn.”

Most experts don’t worry about this because they do more than predict numbers. Each estimate comes with a detailed analysis of the factors behind the headline number, which is what investors and businesses need to decide how to spend money.

“Surprises happen, and people want to know, ‘What does this mean, what’s going to happen, what’s driving this?’” said Michael Pugliese, U.S. economist at Wells Fargo. “I think this is very important and detailed information that you want to have when making decisions as a trader in these markets.”

But prediction markets can become an input for complex forecasts, like those constructed by the Federal Reserve. Justin Wolfers, an economics professor at the University of Michigan who has studied and written about early versions of prediction markets, told Fed officials they should consider such markets. They have been hesitant, he said.

“There’s a deep problem, which is: If you did this, you would democratize decision-making,” Wolfers said. “Today, the senior economist has a lot of power. His vision prevails.”

Also read: We need more capitalists, not necessarily more capitalism

It may also be true that neither individual experts nor a collective of thousands are the best at predicting the future. Over the past decade, a group called Good Judgment developed a model for selecting people with a proven track record of getting predictions right. These “superforecasters” are applied to long-term issues for paying customers. They work collaboratively, but ultimately vote individually.

Warren Hatch, CEO of the organization, believes that prediction markets complement his group’s services because they focus on short-term issues and expand the use of probabilistic thinking.

Now he sees the emergence of another predictive force: artificial intelligence, which can synthesize large amounts of standardized information to generate reasonably good estimates. But AI may struggle with issues more related to humans and culture, and less about numbers and metrics.

“When data is sparse and the environment is in flux, machines are, by definition, retrospective,” Hatch said. “And that’s where I think the space for humans will remain.”

c.2026 The New York Times Company