Prediction market spikes distort LA mayoral odds, new data shows

Prediction markets for the Los Angeles mayoral race recorded over $47 million in trades, making it the most heavily wagered mayoral contest of 2026. Despite this high volume, the platforms failed to predict the winner, with Spencer Pratt finishing third in the June primary despite strong odds. The discrepancy between market projections and the final vote has intensified scrutiny over how these platforms function.

Spencer Pratt, a former reality television star, leveraged criticism of the city's response to the Palisades fire to build a campaign. On the eve of the primary, one market assigned him a 78% probability of advancing to the November runoff. He did not advance, and the error prompted the Los Angeles County Clerk's Office to address the confusion caused by public reliance on these platforms.

Dean Logan, the registrar-recorder for the office, stated that prediction markets can add to confusion when their projections are treated as more authoritative than actual vote counts. The failure to call the race correctly has fueled speculation that the markets may be manipulated to create an illusion of support for specific candidates.

Exclusive reporting by CNBC, supported by data from the Anti-Corruption Data Collective, confirms that such manipulation is possible. The nonprofit found that 38 of 167 congressional markets worth more than $100,000 experienced price spikes of at least 30 cents. In the Pratt race, a trader spent $1.25 million on "Yes" contracts on May 20, causing the price to jump from 29 cents to 45 cents before reverting.

Kalshi, one of the largest platforms, reported that the price self-corrected within nine seconds. However, ACDC analysis showed that the price remained three cents higher than the original level for approximately two hours. This sustained shift could influence public perception, as a Brennan Center for Justice study found that voters were 10 to 20 percentage points more likely to believe a candidate would win if prediction markets favored them, regardless of polling data.

The ease of moving prices is highest in markets with low trading volume. ACDC noted that 88% of 18,000 active congressional markets could see their prices shift by 10 cents with a single wager under $1,000. In one instance, $8,000 in trades drove the odds for Illinois Senate candidate Juliana Stratton from 47 cents to 99 cents briefly. Stratton won the primary two weeks later, and her office pointed to a poll released the same day as the spike.

Robert DeNault, head of enforcement at Kalshi, said rational traders typically correct such anomalies quickly. The company added transparency features to its midterm election markets this month, including labels for low-activity markets and live trade feeds. Polymarket also launched a surveillance system to detect manipulation attempts, though both platforms are governed by the Commodity Futures Trading Commission, which sets rules to prevent insider trading and market manipulation.

Despite regulatory oversight, the identity of traders remains largely opaque to the public. Polymarket partnered with Chainalysis in April to track suspicious activity on its international platform. The two major platforms, Kalshi and Polymarket U.S., collect customer information including legal names and Social Security numbers, but outside observers lack direct insight into who places each bet.