Yeah, lots of movement today…Wisconsin flipped…other D #s improved.
Basically, PM is now in line with the polls…
Polymarket:
| D | R | Δ |
|---|---|---|
| 251 | 287 | 56 |
PM’s overall prediction (right now) still stands 58-42 in favor of Trump…but yesterday is was 63-37…
Here’s the long version:
The Trump-Betting Whale Speaks Out: ‘I Have Absolutely No Political Agenda’
Self-described French trader says he bet more than $30 million on U.S. election after analyzing polls
By
Updated Nov. 1, 2024 2:51 pm ET
The shadow of former President Donald Trump is projected on a U.S. flag as he attends a rally in Green Bay, Wis.
The man who is betting more than $30 million on a Donald Trump victory wants you to know that he isn’t trying to manipulate the U.S. election.
“My intent is just making money,” the man, who called himself Théo, said during a Zoom call with a reporter from The Wall Street Journal earlier this week. He described himself as a Frenchman who had previously lived in the U.S. and worked as a trader for banks.
Théo’s huge wagers on Polymarket—a prediction market that isn’t open to Americans—drew broad attention last month after the Journal reported that four accounts on the platform had been systematically purchasing wagers on a Trump victory. The bets lifted Trump’s odds of beating Vice President Kamala Harris, as shown on Polymarket. Blockchain data showed that the accounts were all funded by the same crypto exchange, fueling debate about the motives of the “Trump whale” behind them.
Last week, New York-based Polymarket said it had contacted the Trump whale as part of an investigation into the wagers. The company described the person behind the bets as a French national with extensive trading experience and a financial-services background.
Here’s how election-betting markets work and what they can tell us about the outcome of the race. Photo Illustration: Jeremy Shuback
“Based on the investigation, we understand that this individual is taking a directional position based on personal views of the election,” Polymarket said in a statement.
The details in Polymarket’s statement lined up with how Théo described himself. He confirmed that he had spoken to a member of Polymarket’s legal and compliance team.
Election watchers are poring over any and all data in an effort to predict the outcome of one of the most contentious—and, seemingly, close—U.S. presidential elections on record. Prediction markets that allow people to place bets on a variety of potential events have emerged as one possible way to forecast election winners. Historical research suggests that, more often than not, the presidential candidate with the best odds in betting markets before Election Day goes on to take the White House.
Yet the emergence of the Trump whale exposes the limits of today’s prediction markets: Even though volumes have surged on Polymarket this year, it is still small enough for one wealthy, opinionated individual to push prices around with a multimillion-dollar bet.
Théo emailed the Journal after the publication of an Oct. 18 article about his wagers. To prove that he was behind the Polymarket wagers, the Journal asked him to place a bet on whether Taylor Swift would announce that she is pregnant in 2024—one of the many small, nonpolitical wagers available on the platform. Minutes later, Polymarket’s website showed that one of the four accounts, Theo4, had placed a small bet on Swift’s pregnancy.
During the Zoom call, Théo wore a gray Nike sweater and sported a short, neatly trimmed beard. Speaking English with a slight accent, he said he had made his bets after concluding that polls were underestimating Trump’s support. He denied speculation that his wagers were aimed at creating a sense of momentum for Trump.
Théo declined to give his real name, and the Journal wasn’t able to confirm all the elements of his story. Although Théo said he was funding the bets with his own money, it couldn’t be determined whether this is true. Nor could the Journal rule out links between Théo and any political organization or Trump allies.
“I have absolutely no political agenda,” Théo wrote in his initial email.
Théo said he didn’t want to share his name because his own friends and children don’t know the extent of his wealth, and he doesn’t want them to know about his Trump bet. He described himself as a veteran financial investor willing to risk tens of millions of dollars on high-conviction trades. But political betting was new for him, Théo said.
Théo said he took an interest in U.S. polling data earlier this year. He observed that many polls underestimated Trump’s support in 2016 and 2020, and concluded that if Trump outperformed again this year, he would beat Harris. Théo also cited the “shy Trump voter effect”—the idea that people were reluctant to tell pollsters that they supported Trump.
“I know a lot of Americans who would vote for Trump without telling you that,” Théo said. Asked about changes that pollsters had made in their methodologies in an attempt to fix the problems of 2016 and 2020, Théo was dismissive, saying he had “not seen anything substantial.”
Théo sent dozens of emails to the Journal reporter over a two-week period. In many of them, he criticized polls from mainstream-media outlets that he saw as skewed in favor of Harris. On the Zoom call, he alleged that Democrat-aligned media organizations were laying the groundwork for social unrest by stoking expectations of a close race, instead of the Trump blowout that he anticipates.
Théo said he was caught by surprise when his trades drew public attention. He started quietly in August by betting several million dollars on Trump, using an account with the username Fredi9999. At the time, Trump and Harris had roughly even chances on Polymarket.
Théo spread out his wagers over multiple days and weeks to avoid causing a price spike. Still, as his bets grew, Théo noticed other traders were backing away from quoting prices when Fredi9999 was buying. That made it harder for Théo to get attractive prices. He created the other three accounts in September and October to obscure his purchasing, Théo said.
Now, Théo could get a payday of more than $80 million-—more than double his investment—if all of his expectations of a sweeping Trump victory come true. Besides his main wager on Trump winning the Electoral College, Théo has bet millions more on Trump winning the popular vote—a scenario that many political observers consider unlikely—and on Trump winning individual swing states such as Pennsylvania, Michigan and Wisconsin.
If Harris wins, Théo could lose most or all of his $30 million, which he described as the majority of his available liquid assets.
He is such a big trader on Polymarket that he is effectively stuck, unable to exit his wagers without crashing the market. The four “Trump whale” accounts collectively hold about 25% of the contracts on Trump winning the Electoral College and over 40% of the contracts on Trump winning the popular vote, according to data provider Polymarket Analytics.
Théo admitted feeling nervous. He voiced confidence that Trump would win—assessing his odds of winning the election at 80% to 90%—but fretted that his bets could be thrown off by an unexpected last-minute news development.
“A surprise can always occur,” Théo said.
Write to Alexander Osipovich at alexo@wsj.com
Theo la baleine
I expect them to somewhat cleave to statistical models. ie. whatever Nate Silver says. That’s not the same as “the polls” since a 1% lean in the polls is (vaguely) like a 30% lead in odds.
I don’t think the models are perfectly rational. Sometimes because model makers are not objective, and see things how they want. Sometimes because they are a little too cautious in keeping only the cleanest assumptions.
And I don’t think the gamblers are rational at all. They are just gamblers. They have human biases like “I like Trump”. I expect any predictive market would have a partisan lean, in whatever direction gamblers tend to lean.
1 Like
Biden was 60/40 just before the election. Clinton was similar in 2016. The real question is still if there is overcorrection in the polls for the Trump effect seen the last two cycles.
Trump is spending a lot of time in North Carolina. That should suggest something…
1 Like
Yeah, fwiw, in Clinton v Trump , intrade(?) gave Clinton a nearly 100% chance of winning. That market also suffered From being very thinly traded.
Nate Silver’s latest newsletter points out that there are too many polls showing a dead heat in the key states – if the polls were the results of fair, unbiased sampling, the nature of random errors would drive more of the polls to have greater separation.
He’s hypothesizing that, after having been burned a couple of times with misses, polling firms are making decisions that force their results to reflect the a-priori expectation of a coin-flip outcome.
2 Likes
That’s a much more articulate way of making the point i was suggesting back around post #50 when knoath posted the swing states all basically 50/50. That can’t possibly be right.
2 Likes
This is a concerning trend, as perhaps we have few true polling results. I think it is wise for pollsters to try to tease out any biases present in their polls observed from past misses, and I think the good polling orgs do try to do this. However, the herding thing is poor work and a CYA approach.
1 Like
Predictit is basically 50-50
Custom prediction map btw.
Trump wins PA and WI.
Harris wins GA and NC.
273-265.
The other thing about prediction markets is that they shouldn’t be able to deviate far from each other. Otherwise there’s an arbitrage opportunity.
If there’s not an arbitrage opportunity, then there must be a lot of fees or legal barriers. Which also would imply it’s not a place for rational investors.
2 Likes
With too much uncertainty, all betting markets will go to 50/50. And polling firms, scared of looking foolish, will go to 50/50 as well. Essentially nobody knows wtf will happen next week.
By the way, if you’re concerned with cheating, or any kind of state level gaming, the correct way to approach it is actual result vs expected result.
So not “Does Harris win NC and MI?” but how much better does she do in those two states than in similar states.
/actuary
My forecast (lol): Trump wins NV,NC,GA,AZ, Harris wins MI, WI, PA
And Harris wins 270 vs 268. It’s possible that the comedian at MSG just saved the world from Trump.
This I find very interesting. It’s actually possible that there is a difference in betting markets if you apply a betting limit on one market and not on the other. The whales on polymarket can’t hedge on predictit.
I’m expecting demographic shifts to swing some states out of order from 2020. NC is the tipping point.
In theory, its people like us that should max bet Trump on predictit and then match that with Harris on Poly. But I know that there are some fees in both markets.
I’ve been thinking that for weeks now.