A trio of University of Miami professors are betting on a new way to predict where a hurricane will hit, an unorthodox approach they believe could help people living along the storm's path decide whether to evacuate.
The three have founded an electronic futures market that allows the public, students and trained forecasters to invest in shares representing selected coastline spots where they think the hurricane will strike. Those who forecast most accurately will get a payout.
The hope is that investors, because they have a financial stake, will draw an accurate consensus on the storm's path — much like bettors predicting which horse will win a race. Upsets happen, but betting favorites win most often.
But the project, expected to open next week, could put the professors in the path of a different storm: criticism from some top meteorologists, who feel forecasts should come from a source the public has come to trust, the National Hurricane Center.
"I don't view it as a game," said David Kelly, a Miami associate professor and economist. Along with David Letson, a fellow economist and associate professor, and David Nolan, assistant professor of meteorology, he developed MAHEM, Miami Hurricane Event Market, with help from trading specialists at the University of Iowa.
"The point is to use markets as a way of collecting and processing information," about where a dangerous storm will strike, Kelly said. "The National Hurricane Center bases their prediction on just three or four models." With the futures market, "information from many, many models is brought in to help figure out where the hurricane is going to land," he continued. "The hope is that this will be much more accurate."
But, Kelly said, "It may be the case that the National Hurricane Center will be more accurate and nothing's better. But there's only one way of finding out. Try it."
At the National Hurricane Center, Director Max Mayfield wasn't amused.
"If they think they're going to help us forecast hurricanes, I don't see how." And the idea that the center uses only three or four models "may be based on misinformation," Mayfield said. "We use at least a dozen sophisticated models" to forecast a storm's intensity, the radius of hurricane-force winds, and other data.
For years, the center's array of storm-related information has been "vitally important" to emergency response centers along the coastline, Mayfield said.
"You'd think if they're serious, they would have contacted the director of the National Hurricane Center, but nobody's called me," he said.
Mayfield had another observation, before turning his attention back to Hurricane Emily, which was barreling toward Texas. "One thing I want to be very clear about," he said. "Forecasters are not allowed to dabble in this sort of thing."
The idea behind MAHEM, said Letson, is to involve individuals with "a wider array of expertise," than the hurricane uses, such as reinsurers. "Obviously, the National Hurricane Center is the biggest, most trusted source," able to simulate the physics of hurricanes and deploy statistical models. "We're just asking people where they think it's going to go, and if the respondents happen to be architects of forecasting models, that'd be great."
Letson lives in Key Biscayne, which has had many uncomfortably close encounters with hurricanes, including four in 2004. While trying to decide whether to evacuate his family, including two small children, Amelia and Zachary, during one turbulent period, he happened to read a report on the accuracy of the University of Iowa's futures-based presidential election forecasting, and the hurricane market idea dawned.
"It's worth a try if we can learn something," he said. "The University of Miami and University of Iowa are not making money. I'm not making money. My colleagues are not making money" on the nonprofit venture. Traders are limited to a $500 spending cap per hurricane season, which runs from June through November.
The Commodity Futures Trading Commission, the federal agency that would oversee such markets, said it won't take action against the university's futures market projects as long as the organizers follow certain guidelines.
Using financial trades to encourage experts to sharpen their skills as they develop predictions isn't new. The Iowa Electronic Markets (IEM) of the University of Iowa has used the tactic to inspire health professionals to forecast where influenza might strike, said Forrest Nelson, an economist and one of IEM's founding directors. Unlike the hurricane market, which is open to anyone, the flu market was open only to health professionals, from doctors and nurses to epidemiologists.
Iowa is exploring a market based on President Bush's prospective Supreme Court candidate, Nelson said. "We're thinking of opening a market as soon as a nominee is named, it'll be a confirmation market. I don't think we've ever run a confirmation market — certainly not for the Supreme Court."
Nelson, who is involved in the MAHEM project, expects most hurricane traders to be meteorologists and hurricane experts. But other contributions can be valuable, he said. "Animals act strangely when an earthquake is approaching. Maybe someone who watches animals," has a valuable insight. TV forecasters, atmospheric scientists at universities, and longtime residents of hurricane districts might be good traders, too, he said.
While federal meteorologists may be prohibited from making a profit, Nelson added, they may be able to participate if their winnings go to charity. "This is not a market people are going to get rich on."
University of Louisville finance professor Russ Ray, a former economist with the State Department, says the hurricane project is the kind of "decision market" that has burgeoned in popularity recently because scientists, politicians and business people have found the markets strikingly accurate. Since its start in 1988, he said, Iowa's futures market has accurately predicted the popular vote outcome of every presidential election.
Futures markets have also predicted Academy Award winners and film box office receipts with uncanny accuracy. Companies like Hewlett Packard and Eli Lilly have based corporate decision-making exercises on the futures market concept, Ray said.
A predictive market like MAHEM reflects a consensus of expert and intuitive opinion.
"It's a flushing out tool that enables everyone to put their two cents' worth in."