Prediction markets giants Kalshi and Polymarket have recently shown a surge in bets that right-wing populist Abelardo de la Espriella will be the eventual winner of Colombia’s presidential election at the end of May.
The mayor is a far cry from previous polls in Colombia, which showed left-wing candidate Ivan Cepeda in second place, followed by center-right candidate Paloma Valencia.
But in the past week, bets on de la Espriella winning the election have increased compared to bets on his rivals.
On May 1, de la Espriella had a 28.8% chance of winning on the Kalshi betting market, while Cepeda had a 38% chance of winning. As of May 8, de la Espriella surpassed Cepeda by 1 percentage point, reaching 42%.
Polymarket is also showing a similar trend. De la Espriella’s odds have increased from 28% to 39% over the past seven days, but he still trails Cepeda, who sits at 41%.
As time went on, racing became more measured. According to Kalshi betting data, from September last year de la Espriella and Cepeda were fighting for first and second place until March, when de la Espriella exploded onto the scene after Valencia won a right-wing party primary in which he did not participate. She has since fallen in both prediction markets and opinion polls.
Colombian Director of Risk Analysis Sergio Guzmán is not confident about the predictive power of betting markets in Colombian elections.
He observed that before the Colombian primaries in March, Valencia had little interest in the betting markets. She won the primary with 3,236,286 votes, and her party received 16% of the vote in the general election. Meanwhile, de la Espriella refused to participate in the primaries, but his party national salvation movementThe vote rate was only 4%. Cepeda’s Party, historical conventionA primary election was held in October 2025 and he won with 1,186,095 votes.
“The best information we have suggests that Cepeda and Paloma have the most proven ability to get the vote,” Guzman said. Latin America Report. “To me, that means there’s something wrong with prediction markets.”
Poll weight index for news websites empty chair Combining polling data from five approved polling companies found that de la Espriella had a voter intention rating of 24%, while Cepeda had 38%.
That would be enough for the right-wing populist to secure a spot in the second round, but the same poll index showed de la Espriella would lose to Cepeda by three points in the second round.
Problems with the opinion poll prediction market
Prediction markets are not new, but the recent rise in popularity of sites like Kalshi and Polymarket presents challenges to democratic processes that require greater public understanding.
Unlike opinion polls, prediction markets measure investors’ intuition, not voting intent. This will allow users to trade cryptocurrency-based stocks in future events. A correct prediction pays $1 per share. Anything wrong loses value.
Prediction markets are therefore biased toward the predictions of those with more capital, rather than the average of users’ probabilistic beliefs, and are vulnerable to manipulation by those with the means to influence perceptions of the likelihood of outcomes.
There is some evidence that prediction markets have outperformed public opinion polls for predicting election margins, but both become significantly more accurate the closer they get to predicted events and should not be treated as reflecting the probability of an outcome.
Kalshi’s betting market for the Colombian presidential election is relatively shallow. At the time of this writing, only US$385,390 has been wagered, making the odds particularly volatile and potentially vulnerable to manipulation.
Coljuegos, the state-run agency responsible for regulating online gambling in Colombia, does not recognize Kalshi or Polymarket as approved operators.
Changes to Colombia’s Voting Process
The discrepancy between polling data and forecast market data may in part reflect the age of available polling data. In Colombia, the most recent approved survey was conducted no later than April 30. This means that all publicly available polls are now more than a week old.
In contrast, prediction markets like Kalshi and Polymarket are constantly updated and the odds change hourly as traders react to new information.
Most polling data is not instantaneous, but the increased delays are due to election polling rules introduced in Colombia last year. The rules effectively impose a gag order on opinion polls until the National Electoral Commission (CNE) reviews the methodology and respondent metadata.
As a result, the release of voter intention polls has been significantly slowed down, often by a week or more.
Spanish polling and consulting firm GAD3 has suspended its polling operations for Colombia’s presidential election, claiming the new requirements make reliable polling impractical.
“It is not feasible to conduct social research with minimal assurance of rigor and operability,” the company said in a statement.
As is the case globally, the prediction markets phenomenon entering Latin American politics will be an interesting trend to watch.
Guzmán suspects prediction markets are more reactive than forecasts, but understands that in the absence of regular poll updates, people need “somewhere they can go to soothe their anxiety.”
Featured image: Screenshot of Mayor Kalshi for Colombia’s upcoming elections, taken on May 9.
Image source: Kalsi
