Practical problems sink climate-prediction markets

David Jeffery at Oikos has been compiling research and writing on the usefulness of prediction markets for weather. These are commonplace in politics now. The typical system sees willing participants “buy contracts” representing victory for a particular candidate or referendum side, at some price less than $1; if the candidate wins, the buyer gets $1, and if not, the buyer gets nothing. Many of these markets do an excellent job predicting winners, even down to the seat-counts in complex parliamentary elections.

There’s much less central control than in a typical betting arrangement, which sees house oddsmakers setting payoffs and whatnot. Here, the going prices are set simply according to what existing contractholders are willing to sell for. It’s their money, so they have a powerful interest in getting it right.

Confused? Look at an example.

The University of Iowa runs several such markets, including two on who’ll be the major parties’ nominees for president in 2008. (There’s one on Canadian elections at the University of British Columbia, too.) As I write, the participants collectively figure Hillary Clinton is the favourite for the Democratic nomination, with a $1 contract going for 44.1 cents, and Barack Obama in second at 30.4 cents.

If you want, you can imagine these as percentage chances, as collectively assessed by the people who decided to play. They have real money on the table so they have a powerful interest in dispassionate examination of the facts, trends, history, and all other evidence they can get their mitts on. The odds of unexpected events (Bill Clinton has another “bimbo eruption,” Obama is found to have written a viciously racist paper in university) are “priced in” to the numbers as they stand — though, of course, if one of those things actually did happen, the numbers would change dramatically.

Applied to the weather, the hope is that the wisdom of crowds could help us anticipate global temperature changes and other climate and weather events. Jeffery’s found a hurricane market, for instance, where participants try to predict where a given storm will make landfall. The idea, where climate change is concerned, is to come up with a collective best guess, incorporating all the science and all the common sense of hundreds or thousands of market participants to make long-term predictions that would be better than any single computer model or even set of them.

Mind you, plenty of financial giants already do make such predictions, and bet money on the outcomes, even if there isn’t a market you and I can speculate on from our desktops. Insurance companies, in particular, need to predict the climate and the weather to know how to price all sorts of property-insurance policies. Solar-energy companies bet on sun; wind-power companies bet on wind. They don’t necessarily trade securities on an exchange, but bet they do.

Unfortunately, exchange-traded weather securities got a bad name thanks to the company that pioneered the industry: Enron. The idea of trading weather-based securities comes in for quite a bit of scorn in the excellent documentary on Enron’s collapse, The Smartest Guys in the Room, seeming to represent as it did the very depths of the company’s slot-machine mentality toward customers’ money, but it still made quite a bit of sense.

If you’re in the solar-power business, you might hedge against the possibility of a cloudy summer — spend a little money now on a contract that pays off big if there are X days of cloud between June and August. If it’s sunny, you’re out the price of the contract, but you’ve made more in solar-power profits so you’re OK. If it’s cloudy and you’ve no electricity to sell, you probably still lose money, but your cloudy-weather contract pays out so it’s not so bad. Other commodity companies — coffee, oil, you name it — do this all the time by buying and selling inputs and outputs at inflated or reduced prices well in advance, just to have the advantage of certainty, so why shouldn’t energy companies?

The fact that trading weather-based securities was an idea with merit regardless of all the nasty things Enron did is indicated by the fact that nearly all the 50 or so people who worked on Enron’s “weather desk” were snapped up in the months after the company’s collapse, at least according to this 2002 story from Environmental Finance:

But six months on, the majority of Enron’s 50-plus weather team – by far the largest in the market – has found new homes. Rather than depressing activity in the market, brokers say, the dispersal of Enron’s weather team has significantly bolstered existing weather desks, as well as promising to bring new participants into the market.

Perhaps the most significant development is the hire of five Enron employees – including Mark Tawney, the former head of weather at the firm – by Swiss Re. This promises a dramatic increase in activity in the market by the reinsurance giant, an enthusiastic early weather derivatives vendor, but one which scaled back its activities dramatically at the end of 2000. It largely pulled out of the traded weather market, concentrating instead on the structuring of larger, one-off deals for clients.

But a rethink at the firm coincided with the availability of staff following Enron’s collapse.“ Swiss Re has been in the weather business for three and a half years and was in the process of identifying steps that would ensure it would become a more effective player in the market when the team came to our attention,” says Phil Lotz, co-chief operating officer of Swiss Re Financial Products (SRFP).

In the U.S., the Chicago Mercantile Exchange has a weather-derivatives market, but it’s all short-term — season-to-season at most, not 50 years down the line.

For the rest of us, as Jeffrey writes, it sure would be useful to have a trackable market in climate futures, so we could see what the big players are predicting.

But we’re talking about contracts that come to fruition in 20 years, or 50, or 100, not at a political convention next summer. Given the thin trading even in short-term weather futures, it seems unlikely we’d see big market plays in long-term climate predictions by giant companies such as Swiss Re. Even more unlikely is the prospect of a “toy” market like the University of Iowa’s, since agreeing to operate one is a decades-long commitment for the benefit of, at best, a couple of hundred interested amateurs. Once the money’s in, the market operator has to see it through to the end.

I think practical problems are going to sink this idea.

Advertisements

One response to “Practical problems sink climate-prediction markets

  1. Rather than on the weather itself you can now bet on the policies that might influence it; i.e. international climate change negotiations…more details here.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s