Temperature, Price and Profit: Managing Weather Risk
last year's 21-percent warmer-than-normal weather in CNG's territories shaved off 33 cents a share, cutting almost $30 million out of the company's earnings. Having studied the industry's quarterly reports last year, Petrowski estimates some $500 million to $600 million in earnings was lost due to winter's warm weather and resulting low volumes. Since most regulated LDCs make money on volume alone, any increase or decrease in throughput flows straight to the bottom line, he notes.
"You can figure that if you have, say, 800 degree-days less in the Pittsburgh region, earnings could be affected, cumulatively, around $500 million, or about $700,000 per degree-day. Across the industry, that's as much as $700,000 to one million dollars per degree-day, on a cumulative basis," Petrowski says.
In addition, weather forecasts can affect how traders do business, says John Cochener, Gas Research Institute's principal analyst. Comparing temperature and prices in 1996, during which a harsh winter "blind-sided" gas markets, showed that, in the very short term, prices slightly lead changes in weather (see Chart 2).
Strengths and Weaknesses
Neither rain, nor snow, nor sleet, nor heat wave, nor ice storm, nor drought stops meteorologists from gathering objective historical and prediction data - forecast information imperative to weather derivative dealers.
A 1986 study conducted by the Gas Research Institute, entitled The Potential Value of Climate Forecasts to the Natural Gas Industry in the United States, reveals that five areas - gas storage, gas supply, service/sales to interruptible customers, maintenance and cash flow management - would most benefit from accurate forecasts. Quantitative estimates indicate that perfect climate forecasts could save companies as much as $8 million per year (imperfect forecasts could reduce operational costs around $3 million), thereby lowering gas prices to consumers.
Since understanding a weather forecast is critical for energy and weather derivative professionals, how can they differentiate between a low- or high-confidence forecast? According to Colin Marquis, a meteorologist for The Weather Channel, weather forecasting is a rapidly evolving science that should be approached with "cautious optimism."
Marquis identifies two factors as particularly important in forecasting: (1) volcanic activity and (2) conditions in the Eastern Pacific (like El Niño and La Niña). This coming winter, notes Marquis, is "very bullish," as La Niña presents a strong weather pattern.
The time of year also plays an important role. Winter forecasts tend to be the most accurate and most confident in long-range forecasts because winter weather patterns are better defined, depending more directly on peaks in big systems. By contrast, summer forecasts are more dependent on small-scale phenomenons and therefore are more apt to contain rapid changes. Spring and fall forecasts reflect transitional phenomena (manifesting lingering effects of winter) and are generally difficult to execute accurately, says Marquis.
As recently as June 1998, U.S. government forecasters confirmed that La Niña, a colder-than-normal South Pacific ocean temperature pattern that developed after El Niño, would likely influence weather conditions into 1999. Representatives from the National Oceanic and Atmospheric Administration had predicted that El Niño would bring warm and dry conditions to the Southwest during the summer, fall and