Think you know who’s going to win the World Cup? So does Goldman Sachs
By Auzinea Bacon, CNN
(CNN) — Think you know who’s going to win the World Cup? Think you know better than Goldman Sachs?
Yes, that Goldman Sachs.
Goldman Sachs, possibly the most buttoned-up of all the big investment banks, released a report on Friday led by chief economist and head of Global Investment Research, Jan Hatzius, listing the odds of potential World Cup winners.
The bank says Spain has a 26% chance of winning, followed by France (19%) and Argentina (14%). Goldman’s prediction model considers a team’s historical performance, scoring talent, team momentum, geography and other factors. The team also considered “the winner’s slump,” warning that Argentina may underperform after winning in 2022.
Goldman listed the probabilities of advancement for all 48 nations, including a spot in the Round of 16 for host countries Canada (50%), Mexico (68%) and the United States (39%).
How reliable are Goldman’s predictions?
Goldman Sachs has previously tried to predict the World Cup’s winner, including the 2014 and 2018 tournaments.
But the firm’s prediction is just that — an estimated guess. The model’s power is “limited,” and soccer is full of “inherent unpredictability,” the economists warned.
Goldman missed the mark in 2018. After one million simulations, it projected Brazil had a roughly 18% chance to take home the trophy in a title match against Germany. But Brazil was knocked out in the quarterfinals and the final saw France defeat Croatia. Goldman Sachs had given France the second-best chance of winning.
Goldman called it an “exciting and unpredictable” World Cup compared to previous tournaments, noting that the firm’s prediction model was sensitive to new information as the game progressed — including Russia’s unexpectedly positive performance.
The Goldman team’s forecast includes an analysis of nearly 20,000 matches since 1978, which can estimate how many goals a team will score in certain match-ups.
While the economists say they have accounted for more factors in this year’s prediction model, the team said that their projections fall short on some ends, such as considering players’ health, managerial experience and more.
Since the research model acts on limited information, it’s more of a “fun exercise” that is further from reality than prediction markets, said Jacek Dmochowski, an engineering professor at The City College of New York.
“The information that is going into (Goldman’s) model is a tiny sliver of all the information that’s in the possession of the millions of people that have bet into (online) prediction markets,” he said.
Prediction and betting markets
In its latest World Cup report, Goldman compared its model to online prediction markets — sites that allow users to bet on almost anything from politics, sports, finance, weather and, more recently, the Iran war. That includes platforms like Polymarket and Kalshi, where users typically trade by taking a position on yes-or-no questions.
“These markets tend to be quite efficient,” said Victor Matheson, an economics professor at the College of the Holy Cross. The markets absorb “all of the available information about a sports contest into the price you’re paying for the bet,” he said.
Both Polymarket and Bet365 predict Spain has less than a 20% chance of winning, and place France’s odds similarly.
But prediction markets are not perfect, according to Dmochowski. There may be overreactions to player injuries and bias towards unlikely outcomes.
Ultimately, it’s “impossible to know who was right,” Dmochowski said.
The-CNN-Wire
™ & © 2026 Cable News Network, Inc., a Warner Bros. Discovery Company. All rights reserved.
