Heating degree-days (HDDs) and cooling degree-days (CDDs) are the standard used in the weather derivatives industry for measuring the change in an average day’s temperature from a usual base of 65°F (18° Celsius). In the US, temperature data is supplied by the US National Weather Service. HDDs are calculated by taking the average of the high and low daily temperatures, subtracted from 65°F, and CDDs are calculated by subtracting 65°F from the average.

The same weather derivatives that help energy marketers could also benefit consumers. While marketers are set back by warm winters or cool summers that mean low prices, consumers suffer under the opposite conditions: extreme temperatures and high prices. But most consumers are simply too small to offset their risk by hedging in the weather or price markets. Simplifying the process of weather and price hedging for the consumer is a business opportunity.

A 1996 survey conducted by Pittsburgh, Pennsylvania based WeatherWise, which provides services to reduce weather-related energy risks, found that more than 60% of consumers would be interested in predictable energy bills. More than three-quarters of those who expressed such an interest said that predictable bills would increase their loyalty to a supplier. The WeatherProof bill, an energy bill developed by WeatherWise, is the first product to take advantage of the need for a consumer to hedge its weather risk, and at this point the company do not see any active competition. It guarantees a total cost for the consumer’s energy supply, regardless of the weather conditions. Unlike budget bills – which average the consumer’s projected annual energy usage and use this to offer bills based on fixed monthly payments – there are no under- or over-payments, to be paid or refunded. Budget bills are already used by about 20% of US consumers. WeatherProof bills also enable suppliers to offer a simplified and predictable bill, thus differentiating the supplier from its competitors in the eyes of the consumer.

The bill addresses unit price fluctuations and consumption, the two major weather risks in residential and commercial energy bills. Unit price fluctuations are eliminated by using a fixed unit rate for energy. Fixed unit rates are available from most major energy suppliers, but are typically only available to large consumers.

WeatherWise then analyses a consumer’s energy use in relation to weather conditions and develops a WeatherProof bill for them. This is the total amount the customer will pay for energy for the year. Typically, WeatherWise uses a year of data and needs at least five actual readings including two in summer and two in winter. With these factors, the bill is calculated, and gives the consumer a fixed total price for energy, typically for a year. Over this period, WeatherWise takes on the risk of increased consumption caused by weather-related factors.

For example, the company might calculate a consumer’s gas heating WeatherProof bill at $1,000 per year. The consumer’s energy supplier collects the $1,000. If the consumer uses more energy – because of a cold winter – WeatherWise pays the energy supplier for the additional energy used. If the weather is warm and the consumer uses less energy, the energy supplier pays WeatherWise the difference. The energy supplier remains in the same financial position as if the bill were not used.

WeatherWise then buys financial instruments – typically swaps – on the weather derivatives market to attempt to reduce its exposure to weather changes. The company takes an opposite position to the major US energy suppliers – such as Enron or Koch – making them ideal counterparties. For example, in winter WeatherWise has a commitment to pay in colder weather and a commitment to receive value in milder weather.

In developing the bill, the company is subjected to a number of difficulties:

Most utility data must be corrected before the company inputs it into its regression models. Inaccuracies arise from imprecise readings or estimates and short periods to accommodate rate changes, among other things. If uncorrected, these factors can cause modelling errors of up to 20%. The company’s technology reviews the data, corrects and adjusts readings based on proprietary analyses and yields a corrected data package for regression modelling. The regression models are then used to calculate the bill.

There is also a degree of model risk as models that work as a class, but do not perform with high precision on individual accounts, can cause “selection bias”. This must be kept to a minimum. To use a hypothetical – and extreme – example, let us presume that the average utility bill, based on typical weather conditions, totals $1,000. To take into account consumers’ different usage patterns, the modelling produces half the WeatherProof bill offers at $600 and half at $1,400. Although the models work well on the average, the consumers that are offered payments of $600 will accept them and those that receive the $1,400 offer will buy their energy elsewhere. This yields a net loss of $400 per customer per average year on the individuals that accept the offers.

In addition to the weather sensitivity of each individual account, the aggregate weather sensitivity must be estimated to calculate the hedge that the company needs to undertake. As consumer consumption is non-linear, the hedge must also be non-linear. But the weather derivatives market only offers linear weather contracts. This means that a complex series of small monthly linear hedges needs to be opened. For example, aggregate consumer heating energy consumption increases exponentially as HDD’s increase. A linear hedge at low energy/HDD levels would leave out the exponential increase at higher HDDs. Therefore, two or more hedge contracts must be combined in order to achieve an accurate hedge – one at the lower rate of use and one or more at the higher rate, or rates.

The hedge must address the differing weights of degree-days in different months, and the concentration of degree days within each month, and eliminate basis risk. Basis risk arises if weather risk in one location is hedged using a temperature contract in a different location, for example in a nearby city. This can present complex mathematical problems, even if the two locations show a high correlation in terms of temperature. Inappropriate hedging using a nearby city can leave most of the risk in place, but translate the risk from temperature phenomena to random phenomena.

What is termed “moral hazard” needs to be taken into account when hedging consumers’ energy use. Moral hazard is an indication of consumers’ behaviour. It includes shifts in energy consumption as a result of changes in lifestyle, appliances or the household. And, in the case of a fixed-price product, a moral hazard can include the possibility that consumers will use more of the product, because they consider that additional use is free. Addressing moral hazard requires a detailed understanding of modelling, engineering, and human behaviour.

The WeatherProof bill first came onto the market in 1997, and by the end of this year, the company expects that 150,000 consumers will be using it, mostly in the US. Users include individual households and small commercial customers. The WeatherProof bill also offers energy marketers a differentiating energy-related product to simplify and enhance marketing and stabilise their earnings. Innovative energy marketers, including regulated utilities, are now offering the WeatherProof bill. These include Shell Energy Services, based in Atlanta, a subsidiary of the oil major Royal Dutch/Shell Group, and US utility Kansas Gas