Many years ago I came into the habit of offering (and taking) bets on absolutely everything. Conversationally, it could be anything from factual and statistical statements to principles or author names, often when I thought I knew something better than my conversation partner.
The point was not the money. It was what happened after losing a few bets: I learned that I often overestimated the things I so confidently knew. Sometimes, convictions held too strongly will cost me.
That is a useful corrective in a time when everyone seems to know everything, or will blindly defer to what a handy LLM produces. From fact-checking to fake news, mythbusting to AI-generated slop, it is getting progressively harder to know what we know. Everyone has an opinion; fewer people have an incentive to find out whether they are right.
Prediction markets have put this ethos and emphasis back on the menu. They’re “civilization’s way of knowing what it knows,” said Eliezer Yudkowsky hyperbolically at a who’s who gathering of the prediction market world in Berkeley, California, in June. That’s a pretty tidy, idealistic summary of largely unregulated sports gambling.
Participating in prediction markets forces people to put a price on their beliefs. And sometimes that is enormously valuable.
Words are Cheap
Economics at its very essential, even metaphysical, core is about human actions — entrepreneurially making choices for an uncertain future.
We act even though we do not know the future. Entrepreneurs raise and deploy capital today in the hope of satisfying customers tomorrow. Consumers make purchases based on perceived future circumstances. Every commercial transaction is, in some sense, a wager on an uncertain future.
Doing things, then, particularly as it pertains to commercial transactions, is the subject matter of the discipline. At high enough a level, it’s what financial markets do, too; investors buy securities in hopes of future cash flow. When they guess correctly and don’t overpay for the assets, they profit. When they’re wrong, the (short) seller benefits. Dynamically, asset prices are our collective, capital-backed, best guesses about the future state of the world.
Prediction markets likewise leverage this money-tested, skin-in-the-game, best-guess model for the future.
Prediction market contracts are zero-sum: one trader’s gain is another trader’s loss, with some slippage for trading fees and the platform itself. So why should society care? Why should we regard a market where two people bet on whether the Fed will cut rates, whether a hurricane will hit Florida, or whether Trump will say “tariff” in a speech as anything other than gambling? How could zero-sum contracts possibly enrich society?
Because relevant, correct information is valuable. And as it is created by these priced transactions, it spills over to the rest of us, free of charge.
Knowledge Creation Is Valuable
The social value of a marketplace isn’t limited to the assets changing hands and the profits of (a few) participants. Prices transmit information, coordinate expectations, tell others whether to deploy more capital or less on a certain idea, whether to expand production or try something else. Prices tell people something about what other market participants know, believe, fear, or expect.
One trading party wins, another loses, but we understand that the system as a whole — asset prices, relocation of capital, a deep and liquid market for issuing securities — is positive sum for society. The spillover knowledge generated by a prediction market contract or an oil price option exploding in price has positive meaning for market actors everywhere. Information is valuable; knowing what it means, probabilistically speaking, for various states of the uncertain future is precisely what financial markets are here to do.
Friedrich Hayek’s famous example in “The Use of Knowledge in Society,” revolved around the market for tin; economics textbooks usually use orange crops and Florida. Prices for goods convey what’s going on. Asset prices do, too.
An oil producer gains valuable information by seeing futures six months out move in price. As a consumer, seeing those moves helps me estimate costs for a tank of gas or airfares, and to judge whether or not to take a certain trip. Traditional financial markets cover some of that, no problem. But the variety of valuable information in a future state isn’t limited by what a regulated exchange or traditional market maker is willing to support. Prediction markets work on the same basic principle, with a much wider definition of what’s potentially tradable.
Contracts traded on prediction markets, however quirky and frivolous, have that same element to them. But just because Kalshi and Polymarket, the two most well-known platforms, are in this sense mostly regulatory arbitrage for sports gambling, it doesn’t mean nothing important is happening underneath all this apparently frivolous speculation.
“Insider” Information and Externalities
Another line of objections to prediction markets is that they make insider trading easier. Prediction markets are riddled with insider trading, say the casual and morally indignant critic. They’re narrowly correct about that claim.
Last month a teleprompter operator working for Trump reportedly made almost $100,000 on Kalshi via mention markets. Through his work, he knew in advance what words would be in a speech, and bet accordingly.
While this feels wrong on some fundamental level — insider enrichment by trading on specific knowledge — it also illustrates that the existing information is valuable. As one virtue of prediction markets is the spillover information it conveys to everybody else, having accurate inputs reflected in public asset prices earlier should be a good thing.
The Trump teleprompter story is just one story among many indicating that yes, prediction markets are rife with insiders profiting off confidential information.
We can think of it as growing pains for a future industry, or as a whistleblower bounty. One purpose of prediction markets, the more anonymous and unregulated they are, is to financially incentivize insiders to disclose important information to the world.
And their existence has positive externalities for everybody else.
The point is that nobody is obliged to bet; traders taking the other side of that bet (at least the more sophisticated ones) are well aware of the possibilities for insider trading and asymmetric information. Also, the information still leaks to the rest of the nontrading public: Because somebody else is willing to put their funds at risk, I can gain some information by seeing what the smart money thinks.
Rewarding the occasional insider is a tolerable price for society to gain access to, and spread more widely, information that benefits us all.
If the price of contracts about whether a leader will be toppled or Israel will launch missiles against Iran suddenly shoots up, lots of people directly or indirectly affected by such events receive vital information that something is coming. Even in the pretty horrid and extreme examples of forest fires, military invasions, bombings, and assassination markets, information is valuable. Large, sudden moves in the odds of these events could, at best, give advance warning and time to prevent disasters, and at least, provide a financial trail for law enforcement to follow later.
Being Confidently Wrong Should Cost More
In an AI age of infinite generation but limited attention, the economic value of information also explains why I don’t find the proliferation of increasingly bizarre prediction markets quite as troubling as their critics do.
Take Trump’s social media posts, access to which is now sold opportunistically to some Wall Street firms in advance. There is obvious value in knowing what the president has said before everyone else does. News organizations pay for speed (Reuters, Bloomberg); traders pay for speed (high-frequency trading). As outlined popularly in Michael Lewis’ book Flash Boys, the race to place computers closer to a financial exchange’s main servers, or indeed running fiber-optic cables between New York and Chicago for trading information microseconds faster, seems largely wasteful to society and closely approximates economic rent to insiders.
If it’s critical enough information to have market value, the earlier some semblance of it can exist in the public domain — clearly, milliseconds count — the better off is humanity.
If you think Truth Social selling early access to posts reeks of pay-for-play and insider enrichment, you’ve already conceded the point that valuable information is otherwise kept away from people and the market, mistakenly leading people to transact on insufficient information. A widespread financial market including contracts on all manner of things (prediction markets) decentralizes and erodes the scarcity value of such advance information and distributes it — meritocratically, we might say — to everyone with skill, devotion, and patience to find it (and profit off it).
Plenty of ideologically motivated commentators in the noisy landscape that is media appearances and opinion pages — plus economists, journalists, think tanks, political strategists, and certainly central bankers — never pay a price for being wrong. They miss currency inflation, recessions, elections, wars, AI timelines, housing crashes, or the value of bitcoin, yet return next week with another similarly confident forecast. Reality rarely seems to discipline them toward humility. Prediction markets can help here, too.
When someone predicts that housing markets will crash this winter, or proclaims there’ll be a government shutdown by October 1, we shouldn’t be afraid to ask: “What are you willing to bet?”
