In 2024, an online market priced Donald Trump's chance of winning Pennsylvania at 65% — three weeks before any major US pollster moved the state from "toss-up" to "lean Republican."
That market was not running polls. It was running prices. And if you've never heard of a prediction market, this is the explanation you actually need.
The world's strangest stock exchange
A prediction market is exactly what it sounds like — a place where people buy and sell contracts whose payoff depends on what happens in the real world.
"Will the S&P 500 close above 6,000 on December 31?" That becomes a contract. You buy it at 38 cents. If the S&P closes above 6,000, the contract pays $1. If not, it pays zero.
The price — 38 cents — is the market's real-time probability. The crowd is saying there's roughly a 38% chance.
Every trade is a vote with money attached
Someone bought because they think the answer is worth more than 38 cents. Someone sold because they think it's worth less. Both sides are forced to be honest because their wallets are on the table.
When thousands of those small disagreements happen, the price becomes something powerful: an instant, public estimate of what the future actually looks like.
It's the same logic as the stock market. The only difference is what's being priced — not earnings, but events.
Why prices beat polls
Polls ask people what they think. Prediction markets ask people what they think enough to act on. The difference is enormous.
In 1988, the University of Iowa launched the first real-money prediction market for US elections — the Iowa Electronic Markets. Across five presidential cycles, IEM forecasts were closer to the actual outcome than national polls 74% of the time, with a mean error of 2.65 percentage points compared to 4.49 for polls.
The further out from election day, the bigger the gap. Markets stayed calm while polls bounced with every news cycle.
The 2024 election proved it again
By October 2024, traditional pollsters had Harris and Trump tied within the margin of error in seven swing states. Polymarket — the largest crypto-native prediction market — had Trump at 60%+ to win the electoral college for most of October.
The polls were not wrong on every state. But the market was right on the overall outcome, and right early enough that you could read the result before election night.
Kalshi, the regulated US prediction market, told the same story. So did Manifold. So did the offshore betting exchanges. Every prediction market that allowed enough liquidity converged on the same answer — and that answer was the one that turned out to be right.
Three things prediction markets are not
First, they are not casinos. A roulette wheel has no information in it — the next spin is random by design. A market on the next jobs report collects real beliefs from people who study the economy. The price is signal, not noise.
Second, they are not opinion polls dressed up. Polls measure how people feel today. Markets measure what people are willing to lose money over. Those are different things.
Third, they are not always right. They are right more often than the alternatives — and when they're wrong, the price moves fast enough that you can see exactly when the consensus broke.
Where this is going
Polymarket cleared more than $9 billion in trading volume in 2024. Kalshi is now SEC-regulated and listed on Robinhood. Major news outlets — Bloomberg, the Financial Times, the New York Times — quote prediction market prices as routinely as they quote analyst forecasts.
Within five years, prediction market prices will likely be embedded in every major news article about an uncertain future event. Inflation. Elections. Court rulings. Sports. The format works.
The question is no longer whether prediction markets matter. It is whether you understand how to read them.
How Juno fits in
Juno is building a prediction market designed for the next billion users — not the trading desk in Manhattan, but the curious investor anywhere with a phone. Easy onboarding. Clear questions. Real money. Real prices.
Whether you want to back your view on the next Fed cut, the next Apple earnings beat, or who wins the 2028 election, Juno is where the crowd will tell you what it actually believes.
The 38-cent question
That 65% Trump price three weeks before the polls caught up was not luck. It was thousands of strangers agreeing through their wallets that the consensus was already wrong.
The next time a confident expert tells you what will happen, ask a better question: what would the people who have to be right say?