A world where every judgment has a price
Last week, prediction markets passed 47 million trades in a single week. That means 47 million judgments were backed by money and recorded in the market. The sheer amount of conviction tied to capital markets reaching this scale is an event in itself.
Prediction markets have already made more real than we imagined. During the 2024 US presidential election, Polymarket kept Trump's chances of winning between 65 and 67%. While major polls called it a dead heat, market prices were much closer to the outcome. Biden's withdrawal was even more dramatic. His chances of reelection, at 36% before the televised debate, plunged below 10% the next day. While the media cautiously suggested "accumulated fatigue" and "not feeling well," the market had effectively pronounced his reelection hopes dead. Just before Russia invaded Ukraine, the market held the odds of invasion between 60 and 70% while Western experts insisted it was "bluffing." When FTX collapsed, the probability of bankruptcy shot from 10% to 90% within hours of CZ's tweet. The news writes the autopsy of an event. Prediction markets read its heartbeat.
Why are markets faster and more accurate than experts? There is information in the world that no single institution can know on its own. A trader on the ground, an industry insider, someone who lives in the region. When each puts their piece of knowledge into the market in the language of 'price,' the sum comes closer to reality than any expert's analysis. This is the 'aggregation of dispersed knowledge' that the economist Hayek described. When money is at stake, people get serious. Someone might answer carelessly on a survey that pays 10,000 won, but when they put 10,000 won of their own money on a prediction, they draw on every bit of information they have. When that seriousness comes together, accuracy improves.
What trades in prediction markets today is only the beginning. Tesla's earnings next quarter, the odds of a particular startup closing its Series B, the probability that a new drug will pass Phase 3 trials. Intangible values like trust, reputation, and future potential, until now left to someone's intuition, all become tradable assets. When farmers in 18th-century Chicago began trading the future price of wheat they had not yet harvested, many people were puzzled. That became today's derivatives market, worth hundreds of trillions of dollars. Prediction markets are only now plowing their first wheat field.
This idea is not new. In a London coffeehouse in the 17th century, merchants began putting money on 'the probability that this ship will return safely.' That was the beginning of what is now Lloyd's of London and the global insurance industry. Traditional insurance has inherent limits. Insurers monopolize risk pools, underwriting takes months, the logic behind premiums is a black box, and coverage can be denied.
Combining prediction markets with DeFi turns this structure upside down. Liquidity providers deposit capital on "the probability of an extreme drought in South Asia within three months," and a smart contract automatically settles when a drought exceeding the threshold is recorded. No underwriting, no denials, immediate payment. Where traditional insurance "estimates the future using past data," prediction-market-based insurance has "market participants pricing risk in real time." Just as Lloyd's began in a coffeehouse, the next generation of insurance will begin in a liquidity pool.
Could traditional finance have built this structure? Smart contracts release funds as soon as conditions are met, eliminating counterparty risk. Decentralized oracles connect real-world outcomes to the blockchain through the consensus of many nodes, rather than a single institution. Liquidity pools with no nationality bring global capital into a single market without the constraints of borders and jurisdictions. This is not a philosophical preference but a structural necessity. Only permissionless infrastructure can open a market for questions like "the probability that North Korea will launch a missile next month."
Where is the line between gambling and prediction markets? Both put money on uncertainty. The difference is whether information becomes a public good in the process. In gambling, the casino designs the odds. In prediction markets, all participants create them together. Gambling tests God from the side of chance; prediction markets interview God from the side of ignorance. The two sit at the same table, but they look in different directions.
What happens when AI joins prediction markets? AI looks at a single market for 'next month's unemployment rate' and automatically generates 100 derivative markets. Markets form in real time for employment shifts by industry, regional effects, and the knock-on effects on consumption indicators. A world where, midway through a drug's Phase 3 clinical trial, prediction markets already signal its efficacy through price, and insurers begin coverage before FDA approval. AI creates questions humans have not imagined, markets put prices on those questions, and humans look at those prices to decide what to do.
When prediction markets grow large enough, the center of information production shifts. The market moves first, and journalists take on the role of explaining that movement. Experts drew their authority from "I know." Markets draw theirs from "we bet." In a world where trust is measured not by who said it but by how much they put on it, the existing power over information will inevitably be reshaped.
What limits, then, do prediction markets face today? The first is the liquidity paradox. A market needs enough participants to be accurate, but it needs to be accurate before enough participants will join. In practice, prediction markets on smaller issues have their prices swayed by just a handful of whales, and those prices are cited as though they were 'the conclusion of collective intelligence.' The second is the potential for manipulation. During the 2024 election, four traders used $30 million to artificially raise Trump's odds of winning, showing how easily prices can be distorted when a market is not large enough. The third is the oracle problem. Prediction markets settle based on real-world outcomes, but "who officially determines that outcome" is a more complicated question than it seems. When interpretations differ over questions like "Has war broken out?" or "Has the drought exceeded the threshold?", code cannot make the judgment. Even decentralized markets ultimately need 'a human who determines the truth' somewhere.
Imagine a world where every administrative decision in a city is tied to prediction market outcomes. Before city hall announces a new road, it first puts "the probability that this road will reduce traffic by 15% in ten years" on the market. If the market says 30%, the plan is revised. If it says 80%, construction begins. Take it a step further: the moment a market assigns low odds to "the probability that Korea's birth rate will rebound by 2035," that price changes companies' workforce plans, the government's immigration policy, and where property developers invest. Through prediction, the market pulls reality forward. A world where a startup founder puts a "revenue target within five years" on the market, and its price becomes the founder's credit rating. In that future, the most expensive asset is neither land nor stocks, but proven judgment.
The clash between the CFTC and prediction markets was a territorial dispute between the existing financial order and a new information infrastructure. In 2022, the CFTC fined Polymarket $1.4 million and shut down its US services entirely. Its reasoning was that "betting money on election results is derivatives trading and cannot operate without government approval." But Polymarket grew outside the US, drawing in global capital, and proved its worth in the 2024 presidential election. A court sided with rival platform Kalshi, ruling that the CFTC's ban on event contracts lacked sufficient legal grounds. Finally, in late 2025, the CFTC officially approved Polymarket's return to the US. The regulator itself reopened the door to the very market it had tried to block. It also meant there was now a legal basis for institutional investors and pension funds to enter.
The door is opening to a world where judgment becomes an asset. If politicians' promises are tested in real time, a drug's efficacy shows up in its price before approval, and disasters are reflected in markets before government announcements, the world will be a little less surprised. A little better prepared. Ten years from now, the institution that reflects the truth most honestly may be neither government nor the media, but prediction markets. And those markets will run not on someone's permission, but on code.
