A Polymarket trader known as 0xA2 topped the platform's daily leaderboard with $917,000 in profits after Democrats took both the House and the Senate in the U.S. midterm elections.

Most of those gains are unrealized, sitting in an active market wagering that Republicans will not win any swing states while Alaska waits on a final result. Polymarket gives an Alaska upset a 14% chance. Two weeks ago, its traders gave Lula a 20% chance in Brazil, and he won.

A "blue wave" swept the U.S. midterms as the Democrats took control of both the House of Representatives and the Senate, in a crushing blow to President Donald Trump.

X post

Per the site's daily leaderboard, 0xA2 is the platform's most profitable trader over the past 24 hours, currently sitting on $917,000 in profit—although most of that remains in an active position that carries a real, albeit small, chance of being upset.

The user wagered $609,644 on Republicans not winning any Senate swing states—a market covering the seven tossup Senate states of Alaska, Iowa, Kansas, Maine, Michigan, Ohio, and Texas—when odds sat at just 36%, meaning traders then priced a 64% chance that Republicans would win at least one of those seats. The market has since surged to 87%. The bet turned the stake into $1.48 million, putting 0xA2 $860,000 in the green. It is important to remember: these are unrealized gains and could drop at any moment as 0xA2 waits for Alaska's final verdict. Polymarket traders put a 14% chance on an upset—and while that may seem small, the site has seen similarly sized upsets recently, as in Brazil's presidential election two weeks ago.

0xA2 also profited $106,115 by predicting that the Democrats would win the Texas Senate—profits that have been realized. Oddly, the user's second-largest active position is $1.42 million on the U.S. not confirming that aliens exist before 2027.

"Election night showed how quickly round-the-clock markets move when traditional exchanges are closed. As polls began closing at 6pm ET, Kalshi's odds of a Democratic sweep went from about 74% to 99%," Kelly Ye, head of research at digital asset management firm Talos, said in a note shared with Decrypt. "Prediction-market trading peaked between 8pm and 10pm ET, led by a small number of large trades."

It was a big day for Polymarket users, with the top 10 on the platform profiting more than $170,000, per the leaderboard. To put that into perspective, on a fairly random day in March this year, the top daily winner on Polymarket profited $154,000, according to a Wayback Machine snapshot, and the 10th-biggest winner that day gained $29,700.

Prediction markets and elections

The midterms attracted major activity on the leading prediction markets Polymarket and Kalshi. According to Precition.com data, midterm markets drew more than $546 million in notional volume over the past two weeks.

"The 2024 election was the breakthrough moment for prediction markets. It was their first major real-world test and the event that really drove volume and open interest," Dash, the pseudonymous data lead at Prediction.com, told Decrypt. "Today, sports dominate trading volume, but politics is still the key driving factor for open interest."

During the 2024 presidential race, Polymarket became what Bernstein analysts called the world's largest election prediction market, with its flagship market topping $3.6 billion in volume. Perhaps more importantly, predictors favored Trump for most of the campaign, while traditional polls kept him neck and neck with Kamala Harris. Prediction market users took this as a sign that prediction markets are more accurate than traditional political polling.

"Increasingly, people are looking for accurate forecasts on election results," Dash said, "and prediction markets are one place they can be found."

Those forecasts aren't always accurate, though. Just two weeks ago, Polymarket predictors whiffed on Brazil's presidential election, with users giving Luiz Inácio Lula da Silva about a 20% chance of winning ahead of election day before he ultimately beat Flávio Bolsonaro to take a fourth term. Experts told Decrypt at the time that Brazil's prediction market ban had left the odds in the hands of traders with little local knowledge to correct them. Experts also warned that a 20% chance of something happening is still a very real chance.

"Ultimately, something that has an 80% chance of happening does not happen 20% of the time," Luke Nolan, senior researcher at CoinShares, said.

So the 14% chance that 0xA2 loses his $860,000 in profit should be taken seriously.