David Bensoussan, known online as Scottlicious, has placed more than $1.5 million in wagers on midterm election outcomes, reflecting a growing trend of high-stakes betting in U.S. prediction markets. These online platforms, including Polymarket and Kalshi, allow users to trade contracts based on a wide variety of events, such as election results, effectively turning political contests into venues for large-scale financial speculation.

Bensoussan, 49, is a seasoned prediction market trader who has earned over $1.7 million in profits from his bets since 2024. His approach to forecasting election results combines polling data, economic indicators, and an intuitive sense of voter enthusiasm, developed through conversations and his background as a former political science Ph.D. student. Though his system is less reliant on extensive spreadsheets or databases, he consults a risk analyst to help temper emotional biases. While he has recorded significant wins, including profits from contests in France and South Korea, he acknowledges that losses are an inherent part of the activity. Over time, his net earnings exceed $2 million after accounting for both gains and losses.

This election cycle, Bensoussan has wagered on various key races, including $113,000 on Mary Peltola, the Democratic candidate in Alaska’s Senate race, and $114,000 on James Talarico, the Democrat contesting Texas’s Senate seat against incumbent Republican Ken Paxton. Texas’s race has attracted intense betting activity, with more than $39 million traded, as the state has not elected a Democratic senator in three decades. Polling shows a narrow lead for Talarico, while prediction markets currently assign him about a 65 percent chance of winning. These markets often respond quickly to news and social media trends—examples include a brief shift in favor of Paxton after a conspiracy theory about Talarico circulated, though the candidate denied the claim.

Prediction markets have gained prominence partly because they can adjust odds in near real-time, unlike traditional polling methods that can remain steady despite emerging developments. Researchers have noted that prediction markets accurately forecasted outcomes such as Donald Trump’s 2024 presidential win earlier than many polls. However, the markets are not infallible; for instance, bettors overestimated the chances of Flávio Bolsonaro winning Brazil’s recent presidential election — an outcome that did not materialize.

The rapid growth of election-related betting has raised regulatory and ethical concerns. Critics argue that these platforms may skirt state gambling laws or be vulnerable to manipulation and insider trading, issues the companies deny. Both Kalshi and Polymarket assert they operate as financial trading venues rather than gambling sites and monitor activity to prevent abuse. Anti-gambling organizations warn that these markets can encourage addictive behavior, given the high stakes and accessible nature of online wagering.

In the 2024 U.S. midterm cycle alone, betting volumes on prediction markets have surpassed $221 million, more than doubling the amount traded during the previous year’s presidential election. This surge has generated significant public attention, with some observers considering market-generated odds as a reliable barometer of electoral outcomes, sometimes even eclipsing traditional polling in media and political discourse.

Despite the risks, Bensoussan describes election betting as a compelling challenge that combines research, intuition, and market dynamics. He hedges his sizable wagers by focusing on outcomes where odds indicate at least a 75 percent probability, aiming to mitigate potential losses. The coming midterms will test whether his substantial bets pay off in expected profits or expose him to significant financial setbacks.