People can now bet on wildfires on prediction markets. What could go wrong?

Prediction markets have entered the natural disaster space.

You can bet on the probability of tornadoes, earthquakes, hurricanes, and new reporting from the Colorado-based website High Country News shows a prediction market platform focused just on wildfires. “Wyldfyre’s” tagline is “You can’t predict fire, but you can trade on it.” Some people see this as ethically murky. Others say natural disaster betting could be useful.

Kaitlyn Trudeau, an applied climate scientist at Climate Central, read that people are betting money on how fast wildfires might spread and what cities they’ll reach.

“Honestly, my first reaction was, that’s gross,” she said.

On Tuesday, Trudeau’s friend in El Dorado County, California was almost evacuated after a fire broke out. That’s the human cost, she said.

“On the other side, people (are) just like callously, kind of coldly betting on their phone about, you know, whether someone's house is going to burn down,” she said.

Trudeau worried that could lead to people starting fires to make more money. But one could flip that argument, said Koleman Strumpf, an economics professor at Wake Forest University.

“If you could incentivize enough people to clear brush and take preventative actions, you could use these markets and bet on that as well,” he said.

That’s just another way markets can potentially change behavior. And with natural disasters, prediction betting could act as an informal type of insurance, said Robin Hanson, an economist with George Mason University.

“If I had a beachfront property or something where there was a flood risk, then I might consider that because it actually tends to be pretty hard to get flood insurance,” he said.

Another possible benefit: prediction betting could help forecast where a wildfire might spread.

“But we already have that, we already know,” said Craig Clements, who leads a team that does that kind of forecasting at the Wildfire Interdisciplinary Research Center. “It gets very complicated very quickly when we're trying to model fires in mountainous terrain and complex fuels with changing weather patterns.”

And Clements is not betting that people using Kalshi or Polymarket are doing all that work.