George Santos, a former US Congressman and convicted fraudster, recently offered 55% off his Cameo video, probably because he’s facing a federal investigation for insider trading. Not insider trading on stocks. Insider trading on himself.
(A Cameo video is a personalised, paid video message from a celebrity, athlete, or influencer that you can buy for special occasions – eg, birthdays – by placing a prompt on the Cameo platform.)
Allegedly Santos posted on X the day before US President Donald Trump’s State of the Union address earlier this year: “I’m going to be there for the State of the Union in the gallery, guys.” Meanwhile, according to investigators, he had already placed bets on a platform called Kalshi that he would not attend. He then did not attend, posting “Watching Sotu on an airport TV was not part of the plan”. He used public prediction markets to bet he wouldn’t show up, and then he didn’t show up, and won his bet. Easy money. Except it’s completely unethical. This situation illustrates a big problem with these new prediction markets.
First off, what are prediction markets? Prediction markets let you bet on anything. Literally. You buy and sell contracts based on real-world outcomes. Who’s going to win the World Cup? Yeah, you can bet on that for sure. But what will the temperature be in Seattle tomorrow? You can bet on that too. It’s a bit nuts. If it can happen, you can bet on it. Which probably isn’t great for the social condition.
The two dominant platforms are Kalshi and Polymarket, which processed over US$41bil (RM170bil) in volume recently. So this is not a small niche, and it seems to be growing. Which will offer more problems for society.
But what’s astounding is the accuracy of these prediction markets. Their accuracy is measurably better than the punditry class that has been confidently wrong on television for decades. During the 2024 US presidential election, while polls showed a coin flip, prediction markets had Trump at 65% ahead well before election night.
An independent study of 2,500 markets found Kalshi correct 78% of the time, Polymarket 93%. Your fave political podcast host has no published accuracy rate and gets a book deal regardless.
The prediction markets are based on people looking at the questions and picking which outcome they think is more likely, and it turns out people en masse are pretty decent at predicting the future. The wisdom of the crowds is indeed wisdom.
Having established themselves as serious forecasting tools, the platforms then opened up the question of what counts as a forecastable event. The answer, apparently, is everything.
Recent markets have included whether specific words would appear in an earnings call by big data company Palantir, whether Elon Musk would win his court case against OpenAI, and whether aliens will be officially confirmed this year.
And then there is the Jesus market. Polymarket users have bet US$87mil (RM354mil) on whether Jesus Christ will return before 2027. The crowd gives it a 2% chance of happening.
Someone then created a secondary market betting on whether those odds would surpass 5%, at which point traders began pushing the original number up just to cash out on the secondary bet. People were gambling on the odds of gambling on the Second Coming.
There is something worth sitting with here. For most of human history, gambling was a vice, conducted in windowless buildings, done against parental warnings, ruining reputations. But prediction markets seem to be making it all very casual.
And when you can bet on anything, someone always knows something.
A US soldier was criminally charged for using classified information about the capture of Venezuelan President Nicolas Maduro to win US$400,000 (RM1.63mil). A Google engineer was charged for betting on search trends using confidential data. And, of course, there’s Santos allegedly betting on his own no-show.
It’s worth wondering how many people have had “inside information” and made money off prediction markets but never got caught. The chance for rigged tables is very real. But then, it’s probably not great to make everything into a gambling event.
I am not here to tell you this is bad. The accuracy numbers are real. Somehow gambling on everything brings out, at least for the time being, a clear-eyed view from the crowd about what might happen. And it’s been sort of accurate so far.
And while I think being able to bet on literally every event isn’t great, it’s interesting that prediction markets have become a decent place to check on certain things happening. Just to prepare yourself.
So I’ve bookmarked a few bets to go back and check on. Particularly, I’m checking if alien life will be disclosed anytime soon. And surprise, surprise the probability of this has been ticking up of late. Could it be nothing? Of course. But remember, in prediction markets, someone always knows something.
