Playing the weather game ; Firms cash in by staying one step ahead of Mother Nature
USA Today; Arlington, Va.; Dec 11, 2001; Del Jones;

Full Text:

Copyright USA Today Information Network Dec 11, 2001

Weather forecasting has improved enough -- and the stakes have shot up enough -- that companies no longer accept the weather as a pure gamble.

Energy giants such as Williams and Koch Industries are strategizing against Mother Nature, employing supercomputers and hiring meteorologists at more than twice what the National Weather Service pays in an aggressive attempt to profit from forecasting expertise.

Demantra, a consultant that has long tried to help companies such as Ford Motor, Sears and Wal-Mart predict swings in demand due to the economy and other factors, has teamed with weather and climate forecaster Planalytics to predict, for example, where spring weather will arrive a few weeks early so that home-improvement centers can stock lawn fertilizer.

Forty-four state departments of transportation pay Surface Systems (SSI) to tell them when highways should be salted for ice. SSI has 1,800 roadside monitors that measure pavement temperature, moisture and how much salt has been applied. Roads require just one-tenth the salt if it's applied just before the snow falls, not after. It's especially important to predict if the pavement won't freeze at all, which can save states $50,000 and more per winter storm, says George Reed, SSI general manager.

Until recently, the National Weather Service would have been more at home as part of the Agriculture Department. Today, it's fitting that it's part of Commerce. Amusement parks save on labor and food costs merely by knowing the odds tomorrow's forecasted rain will fall between 6 a.m. and 10 a.m., the hours when most customers decide to make a detour to the movies. The makers of allergy medicine fine- tune distribution and advertising campaigns by knowing the odds of when pollen will fly in Albuquerque or Albany.

If not for consumer backlash, vending machines would already be raising and lowering the price of a soft drink depending on the temperature.

More than $1 trillion of the U.S. economy -- from airline travel to orange orchards to skyscraper construction to sales of holiday sweaters and root beer floats -- is sensitive to temperature, precipitation, wind and humidity.

Industry still has no control over the weather. But dealing with it is being seen as less a game of chance and more like a game of chess that revolves around probabilities and tactics. That fits well into the business world, where tough decisions are rarely certainties but are driven by probabilities.

Coincidence, perhaps, but one expert in the new world of weather came from the world of advanced chess. "You must see several moves and several combinations ahead," says Anjelina Belakovskaia, a weather derivatives trader at Williams and three-time U.S. women's chess champion. "Both combine long-term strategy and short-term tactics. Everybody has almost the same information, and this is where skill comes into play."

Skill has its rewards:

* Bombardier, maker of Ski-Doo, boosted sales 38% by promising a $1,000 rebate on each snowmobile sold if snowfall in 44 cities was less than half that of the previous 3 years. Many people wouldn't spend $7,000 on a new snowmobile until it snowed enough to use it.

Trouble was, in most years it didn't snow enough until Dec. 25 had passed. Bombardier solved that by wagering against its customers and passed along the risk via a financial device known as a weather derivative (story, right).

* US Airways stands to lose $50 million if a major winter storm on the East Coast comes in worse than forecasted, strands jets and crews and snarls traffic as far away as California and beyond. Energy companies are giving chase, but airlines, not surprisingly, were the pioneer in hiring meteorologists.

* The Tennessee Valley Authority, which generates 4.8% of the USA's electricity, has a meteorologist who can save $100,000 and more on one extremely hot day with a forecast 1 degree better than that of the National Weather Service. Wrong forecasts cause TVA to fire up generation from hydro, nuclear, coal, wind, natural gas and oil plants that is not needed, or it is forced to buy expensive wholesale electricity at the last minute.

Before deregulation, energy prices that spiked on extremely hot or cold days were passed along to the utility customers, but now profits ride on controlling those costs. Energy is bought and sold in advance on the futures market, making the accuracy and timing of weather forecasting worth a fortune.

* Energy company Williams paid $10 million to the University of Oklahoma in August, the largest arrangement ever between a private company and a university meteorology program. Such partnerships between universities and the private sector have been common in engineering and biotech but are catching on in meteorology.

Getting it right more often

Weather forecasting has steadily improved. National Weather Service director Jack Kelly says today's 3-day forecast is as accurate as the 24-hour forecast of 20 years ago because of computer models that can map out where high pressure and storm systems will be weeks in advance.

By law, National Weather Service forecasters can't invest in energy futures because of their ability to swing prices by millions, even billions, of dollars. But private companies like Williams and Duke Energy can buy or lease computers that are 30 times faster than those used by the weather service in 1995 for one-fortieth the cost.

Williams is configuring a new supercomputer capable of 300 billion computations per second that will provide real-time forecasts. Its partnership with the University of Oklahoma gives it a window to the most powerful supercomputer among the Big 12 universities, a computer dedicated to forecasting that can perform 500 billion computations per second.

Each computer is as powerful as the network of computers clustered at the National Center for Atmospheric Research in Colorado, says David Montroy, head meteorologist at Williams.

It's unlikely computers will ever be able to tell you in May to carry an umbrella on the Fourth of July. "If you had a computer big enough, it would generate so much heat, it would change the weather," says Bob Brammer, chief technology officer of the Northrop Grumman engineering arm TASC.

But private forecasters aren't trying to replicate the National Weather Service. They use the government data collected at 8,000 sites, then employ their sophisticated computers and software to focus on a single weather challenge, such as ice on a stretch of highway or a heat wave in a major city where the accuracy stakes are highest. They can focus on extremes such as the 100-degree day, rather than the 75-degree day. They customize, while the government must forecast nationwide.

Increasingly, private forecasting is going long range to let companies manage inventories. Long-range forecasting was little better than a coin toss a few years ago, but a century of good data and an understanding of global warming and events like El Nino have improved those odds beyond guesswork.

This is a tough winter to predict, but the Southwest United States has a 43% chance of enjoying temperatures above normal and a 33% chance of being close to normal. This, of course, comes with no guarantee, but meteorologists widely agree that there is less than a 1-in-4 chance of the winter being colder than normal in the Southwest.

The limits of forecasting

Of course, meteorologists also predicted 3 inches of snow over North Dakota in October, then watched it become a foot-deep blizzard.

Aquila Energy runs a contest that pays $50,000 twice a year to the meteorologist who makes the best temperature forecast for 13 cities. While the winner is within 1 degree on average, the participants who do the worst miss the mark by an average 8 to 9 degrees.

The weather derivative market is so new that company officers assigned to managing risk don't really know how to use it. "Risk managers will be the first to tell you they are totally at sea," Brammer says.

Many companies remain skeptical. Burlington Coat Factory CEO Monroe Milstein says Enron tried to sell him a weather derivative to hedge against a warm winter a couple of years ago. There is no denying that warm winters are bad for coat sales, he says, but Enron wanted to charge Burlington a $2 million premium in exchange for $12 million that Enron would give Burlington if the winter came in warmer than it's been in 150 years, Milstein says.

That's a 6-to-1 bet, Milstein said. Enron eventually offered Burlington 8-to-1, or $16 million for a $2 million premium, but Milstein turned that down, too. "The warmest winter in 150 years certainly sounds like at least 10-to-1," he says.

(Enron spokesman Eric Thode says recent Enron financial problems that have put the company into bankruptcy protection had nothing to do with its relatively small weather derivatives business.)

Another hitch is that private weather companies market themselves as being far more accurate than the government without independent verification. Kelly says it would not surprise him if private companies are beating his forecasters in specific areas of focus, but there is no outside audit.

None of the drawbacks has put a damper on the race for an edge. The weather service tries to keep the playing field level by releasing forecasts and data to all interested parties at once. "The question is: How to get the data into your system and make use of it," says Lee Branscome of Climatological Consulting.

Doppler radar has saved untold lives by making tornado warnings 14 to 38 minutes in advance vs. 6 minutes 15 years ago.

In the world of energy trading, crunching the forecast numbers minutes or seconds faster than the competition can save untold fortunes.

Abstract:
Many companies remain skeptical. Burlington Coat Factory CEO Monroe Milstein says Enron tried to sell him a weather derivative to hedge against a warm winter a couple of years ago. There is no denying that warm winters are bad for coat sales, he says, but Enron wanted to charge Burlington a $2 million premium in exchange for $12 million that Enron would give Burlington if the winter came in warmer than it's been in 150 years, Milstein says.

A: A ski area could pay a $250,000 premium to collect, say, $100,000 for every inch of snow this winter under the "strike" amount of 100 inches. This is like a "put" option. The ski area is out the premium whether or not snowfall is inadequate. Or, it could enter into a "swap" with another party, paying no premium and getting $100,000 for every inch under 100 and paying $100,000 for every inch over. Increased ticket sales in good winters would cover the cost.

A: Small fortunes can ride on weather instruments that are fallible. Gauges have gone unfixed for years. The National Weather Service sometimes moves instruments, and we've all seen rainstorms that hit on one side of the street but not the other. Unlike insurance, which spreads premiums and risks over time, derivatives are one-shot, risky deals that create big winners and losers in a hurry. It looks ripe for lawsuits, but traders say that hasn't been the case. Also, financial markets need simplicity to trade in large blocks, but weather derivatives usually need to be customized to individual needs.
 

TEXT OF INFO BOX BEGINS HERE

How weather derivatives work

Q: What is a weather derivative?

A: It's a financial instrument that seems like an insurance policy but is more like an option. Most weather derivatives are based on how much the temperature goes above or below 65 degrees. But weather derivatives can be based on anything measurable.

Q: Can you give an example?

A: A ski area could pay a $250,000 premium to collect, say, $100,000 for every inch of snow this winter under the "strike" amount of 100 inches. This is like a "put" option. The ski area is out the premium whether or not snowfall is inadequate. Or, it could enter into a "swap" with another party, paying no premium and getting $100,000 for every inch under 100 and paying $100,000 for every inch over. Increased ticket sales in good winters would cover the cost.

There is also a "call" option where the ski area receives a premium of $250,000 and pays $100,000 for every inch over the strike 100 inches, again assuming higher revenue with heavy snowfall.

Q: I understand why companies would hedge against the weather. Why would anyone assume financial risk by taking the other side?

A: Some are speculators. They believe they understand the probabilities of weather and are willing to wager just as a football fan might bet against a football team because he believes the quarterback is injured. But most parties take the other side of a weather derivative because they also are hedging. Cities, for example, might want to hedge against heavy snowfall because of the cost to clear the streets.

Q: What makes a weather derivative different than other derivatives?

A: All derivatives are used to hedge against bad news. Airlines use derivatives to protect against soaring jet fuel prices. Derivatives are commonly used to avoid fluctuations in interest rates or foreign currencies. What makes weather derivatives unique: They are not derived from anything with an underlying value.

Q: How big is the weather derivatives market?

A: Not very. The first wasn't sold until 1997. It has grown to $12 billion and there are signs that some small companies are interested. Last June, the Rock Garden in London became the first restaurant to hedge against the cool weather that keeps customers from populating its outdoor tables. Aquila predicts $50 billion in weather derivative contracts by 2005, while the value of all derivatives traded worldwide is $100 trillion.

Q: Are there problems with weather derivatives?

A: Small fortunes can ride on weather instruments that are fallible. Gauges have gone unfixed for years. The National Weather Service sometimes moves instruments, and we've all seen rainstorms that hit on one side of the street but not the other. Unlike insurance, which spreads premiums and risks over time, derivatives are one-shot, risky deals that create big winners and losers in a hurry. It looks ripe for lawsuits, but traders say that hasn't been the case. Also, financial markets need simplicity to trade in large blocks, but weather derivatives usually need to be customized to individual needs.

[Illustration]
PHOTO, Color, Dave Crenshaw for USA TODAY; Caption: Weather expert: Anjelina Belakovskaia, a weather derivatives trader and chess champion, likens her job to a chess game: You must see several moves and several combinations ahead. Both combine long- term strategy and short-term tactics."