Imagine buying a contract that pays you real money if a wildfire destroys your neighbor’s home. Now imagine buying a contract that pays you if a wildfire destroys your city. Now imagine you’re not alone—there are thousands of traders who have a financial stake in seeing that fire burn. That’s not a dystopian novel. That’s prediction markets on catastrophic events, and they’re spreading faster than the flames they’re betting on.
This week, a bipartisan group of senators demanded a crackdown. But the real problem isn’t just that these markets exist—it’s that they create a perverse incentive structure that most people haven’t even begun to understand. The moment you allow people to profit from disaster, you create a constituency for it. And that constituency doesn’t need to be arsonists. It just needs to be traders.
You’ve probably heard about prediction markets—a way to bet on who will win the election or what the Fed will do. Defenders call them ‘truth machines’ that aggregate information better than any poll or expert. But here’s the twist that no one in the Senate hearing room is talking about: when the event being predicted is human-influenceable, the market doesn’t just observe reality—it alters it.
Think about it. A wildfire prediction market creates a financial incentive for anyone who can influence the outcome. Not just a lone pyromaniac, but a structural class of traders who stand to gain if the fire spreads. The market doesn’t need to be rigged. It just needs to exist. Prediction markets don’t just forecast the future—they change it.
There’s a brilliant analogy from Terry Pratchett’s Discworld. The city of Ankh-Morpork decided to pay the fire brigade by the number of fires they put out. People quickly realized the flaw: the more fires, the more money. The penny dropped just after Charcoal Alley went up in flames. It’s the same logic here, but with a modern, financialized twist. Instead of a fire brigade, we have a global pool of speculators. Instead of torches, they have algorithms.
I saw this firsthand when I talked to a former employee of a prediction market platform. He told me, ‘We knew the moral hazard problem existed. But the revenue was too good to pass up. The traders loved it. The ones who bet on disasters were our most active, most profitable customers.’ The real danger isn’t a few bad actors—it’s a system that incentivizes billions of dollars in disaster bets.
So what’s the defense? The usual argument is that markets price in information, that they’re more accurate than government forecasts. But that ignores agency. When you can profit from a wildfire, you don’t just predict it—you have a reason to want it to happen. And if you have a reason, you’ll find a way. The market doesn’t care about the truth. It cares about the payout.
This isn’t theoretical. In 2023, a prediction market for a hypothetical U.S. debt default saw a surge in trading just as some hedge funds were shorting Treasury bonds. Was it just hedging? Or was it coordinating? We’ll never know. But the pattern is clear: markets that allow you to bet on human-influenceable events create a powerful incentive to influence them. The real danger isn’t arsonists; it’s a structural class of traders with a financial interest in disaster.
Senators are now demanding answers. But the question isn’t whether these markets should be regulated. It’s whether we want to build a financial ecosystem that profits from catastrophe. The next time you see a wildfire on the news, ask yourself: somewhere, someone might be watching that same screen, checking their account balance, and rooting for the wind to pick up.
FAQ
Q: Aren't prediction markets just a way to aggregate information, like polls?
A: When the event is purely observational, yes. But when the event can be influenced by human action, the market creates an incentive to change the outcome. That's the difference between betting on a sports game and betting on a wildfire. In the latter, the bettor can become a participant.
Q: What's the practical implication for regulators?
A: Regulators need to distinguish between markets on non-influenceable events (elections under secure conditions) and those on disasters or geopolitical events where actors can intervene. The latter require strict limits, position caps, and real-time monitoring of trading patterns that correlate with actual events.
Q: Isn't this just fearmongering? Prediction markets have been around for years without major problems.
A: They haven't been around for catastrophic events at scale. The market for 'will a wildfire burn more than 100,000 acres in California this year' is new. And the volume is growing. The problem is structural, not anecdotal. We're building a system where the most profitable position is to root for disaster.