- 10 -
Weather Derivatives
-Kaushank Khandwala
Weather Derivatives an Introduction
Weather has significant influence on many commercial businesses. Many businesses are impacted positively or unfavorably by weather. For this reason, new set of weather risk protection instruments called weather derivatives have emerged.
A derivative is a contract whose value depends on the value of underlying variable. The main types of derivatives are futures, forwards and options. Derivatives are primarily used as hedging instruments, though they also find place in speculation and arbitrage. Most common underlying instruments are stocks, bonds, or commodities. In case of weather, the underlying variable could be average Temperature, precipitation, snow fall or wind.
Weather derivatives are instruments used to reduce risk associated with unanticipated weather movements. For example: A shopping mall in a snowy area may use weather derivatives to stabilize its earnings
Weather derivatives v/s Weather Insurance
Weather derivatives cover low-risk, high-probability events. Weather insurance, on the other hand, typically covers high-risk, low-probability events, as defined in an insurance policy. For example, a company might use a weather derivative to hedge against a summer 3° more than the average (a low-risk, high-probability event). In this case, the company protects its revenues from adverse future weather. But the same company can purchase an insurance policy for protection against damages caused by a (typhoon)(high-risk, low-probability events)
Weather Derivatives History and Major Exchanges
Historically, the first weather derivative transaction was contracted in July 1996. Aquila Energy entered a commodity transaction with Consolidated Edison Co. This involved purchase of electric power in August from Aquila. Consolidated Edison was to be offered a rebate if August was cooler than predicted. These predictions were based on Cooling Degree Days measured at New York City's Central Park weather station.
Weather derivatives slowly began trading over-the-counter in 1997. As the market for these products grew, the Chicago Mercantile Exchange introduced the first exchange-traded weather futures contracts (and corresponding options), in 1999. The CME currently trades weather derivative contracts for 18 cities in the United States, nine in Europe, six in Canada and two in Japan. Most of these contracts track cooling degree days or heating degree days, but recent additions track frost days in the Netherlands and monthly/seasonal snowfall in Boston and New York.
In USA, weather derivatives are traded in the Chicago Mercantile Exchange (CME). In 2006, traded CME Weather derivatives had a notional value of $22 billion. In Europe, London International Financial Futures Exchange (LIFFE) is a major exchange for Weather Derivatives London. LIFFE traded Weather derivatives had an estimated notional value of $9billion.
In Japan, Weather Derivative products are not exchange -traded in Tokyo stock Exchange. Most of the contracts are Over-the-counter. The Japanese market for weather derivative products (OTC temperatures, snowfall) grew a third to about 80 billion yen ($680 million) in the year to April 2007. Japan's TIFFE is believed to start trading weather derivatives in early 2009.[1]
Modeling Weather Derivatives
Weather Derivatives are recent class of financial instruments. There are few widely accepted pricing and modeling methodologies for such instruments. Here we explore Black Scholes Model, Burns model and Monte Carlo Simulations for weather derivatives.
Black Scholes Model
Black Scholes Metron Model (BSM) is used to price derivative options.
Standard Black-Scholes Formula’s Inputs
• S = Underlying Asset Price
• X = Strike Price
• T = Time before Maturity
• σ = Volatility
• R = Risk Free Rate
BSM model has many assumptions which do not apply to weather variables and so cannot be used to price weather derivatives. BSM Model is probably inadequate because of following reasons:[2]
1. Weather does not “walk” quite like an asset price. In BSM, asset price can theoretically move from 0 to infinity. Weather variables such as temperature in general fall within a narrow bands because of mean –reversion tendency.
2. Weather is partially predictable in the short-run and partially random around the averages in the long run. So, it is not completely “random”.
3. Underlying variables (e.g. snow fall) are not tradable and market for weather derivatives are illiquid, so pricing cannot be free of economic risk aversion factors.
4. Black Scholes option payoff depends on the asset value at expiration. Weather derivatives would require average value of the variable and are more “Asian” in nature.
5. Many Weather derivatives are capped in pay-off, unlike the standard Black-Scholes option.
Burn Analysis
Burn Analysis are simulations based on Historical Data. Using historical data, we can determine the expected payoff every year. The fair price of the option would be the average of the historical payoffs. Burn Analysis is widely used by insurance companies. Main steps involved are:
1. First, collect the historical data
2. Make adjustments to make the data comparable across different periods
3. Create an appropriate variable such as (Snow Depth/Fall, precipitation index)
4. For every year in the past, determine the option payoff.
5. Find the average of these payoffs.
6. Discount them to settlement date.
There are some problems in Burn Analysis:
1. Burn Analysis does not account for weather forecasting. Over a period of time, weather variables change. Even though, we assume mean-reversion behavior for weather variables, weather models /forecasts should be a part of analysis to obtain value.
2. Weather is local to a region. Gathering historical data can be cumbersome process and authentic weather need not be always available. Even when data is available, there are possible gaps and errors. The historical data must be adjusted before they are used for burn analysis.
3. It is quite tricky to consider the historical period. A city, town or village may undergo urbanization wherein due to heavy industrialization and construction, the weather grows warmer over time. These trends must be accounted for when pricing the option.
Monte Carlo Based Simulations
Monte Carlo is a simulated method of generating random numbers. Monte Carlo simulations provide a convenient way to contract and price derivatives.
This process can be used to statistically construct weather scenarios. For Monte Carlo simulations, it is important to choose the right model for the random walk of weather variable. In general, weather variables follow mean reversion. Models that only assume pure random walk behavior are not suitable.
Monte Carlo typically involves generating a large number of simulated scenarios to determine possible payoffs for the instrument. The fair price of the instrument is the average of all simulated payoffs, which is appropriately discounted to settlement date.
Drawbacks:
Monte Carlo process is very computationally intensive. Using stochastic or statistical weather models is complicated as there are many variables involved. Many trials would be required to arrive at a fair price.
Weather Derivative Structures:
Typical Weather Derivative Structures include Put/Call with Caps/Floors, Weather Swaps and Collars. The coming section would describe these with examples.
Call and Put Options with a Maximum Payoff (CAP)
Capped option is an option with a pre-established profit cap. A capped option is automatically exercised when the underlying security closes at or above (for a call) or at or below (for a put) the Option's cap price. Weather Derivatives typically have such structures wherein there is an upper limit on profit which can be obtained based on the movement of underlying variable.
Example The city council of a town has spent $3mn to remove 10 cm of snow. The city office estimates that additional inch of snow causes an increase of $250,000 of snow removal costs.
Solution: A Snowfall calls option which pays $250,000 per inch of snowfall above a strike of 10 cm to a maximum of 20 cm.
Call Option Features
Period = Nov-Mar
Strike = 10 cm
Limit = 20 cm
Tick= $250,000
Limit = $4,000,000
Price = $500,000
Weather Swaps
A swap instrument combines a call and put option with the same strike. Weather Derivative Swaps can provide revenue stability as depicted by the following example:
A leading Power company in Japan and Gas Company entered a contract for the period August-Sept 2001 for a temperature return swap. The underlying variable was temperature with base (standard temperature) at 26℃. Power Company purchased a PUT option with strike at 25.5℃ and the Gas company purchased Call option at 26.5℃.The profit /payoff diagrams are as shown below.
Time frame / August 1,2001 – September 30,2001Index / Average temperature (SYNOP Tokyo）
Standard Temp. / Approx. 26℃
Put strike / Standard Temp.-0.5℃
Call strike / Standard Temp. +0.5℃
Maximum payment / Approx.\700,000,000-
Actual temp. / 24.8℃
Actual payoff / TEPCO received approx.\320,000,000-
Contract between power company and Gas company for total return swap
Situation in Minami Uonuma City
Minamiuonuma city is situated in a valley in a mountainous region of Niigata Prefecture The city is bounded by Uonuma city and the Echigo-Sanzan mountains in the north, and Yuzawa, a popular ski resort town, in the south. The Uono river flows through most of the city. It is known as “Snow Country” (Yuki Guni) because of the heavy snowfall in winter.
Minami Uonuma City has 4-5 months of heavy snow fall. The region has many ski resorts. Snow significantly impacts the business and livelihoods of individuals and corporations in this region favorably or adversely. The region is famous for its rice production and also ski tourism.
For our Project, we present a detailed analysis of snow fall, potential impact on snow removal companies and Minami Uonuma City office and potential weather derivative products.
We interviewed and collected data from Minami Uonuma City office and leading snow clearing companies for the last 15 years.
Our Methodology:
Main Steps in our project were:
1. Identification of potential businesses directly affected by snow fall
2. Interviewing of a few of these businesses and collecting data
3. Analysis of snow depth or snow fall data in surrounding data
4. Choosing the correct underlying variable for our weather derivative product
5. Analysis of the financial impact of variable/extreme snow fall for businesses
6. Exploration of appropriate Weather Derivative products based on Snow fall
The following table depicts the businesses impacted favorably or unfavorably by snow fall.
Businesses impacted Positively by extreme Snow fall / Businesses impacted negatively by extreme Snow fall1. Tourism, Ski industries
• Hotels
• Resort operators
• Ski equipment sellers
2. Retailers
Winter goods sellers
3. Agriculture (need to secure water supply)
4. Snow Removal Business (construction companies) / 1.Retailers,Supermarkets
2.Construction Companies
3.Community-based shops
4. Road Transportation
5.Amusement Industries :Pachinko shops, arcades
6. City Governments
Ski resorts are primarily dependent on tourists from Kanto region and abroad for their revenues. Heavy Snowfall is favorable for many tourism dependent businesses. On the other hand, construction businesses, retailers and road transportation prefer low snowfall. Also, agriculture is significantly impacted by the duration and snowfall depth. Heavy Snowfall would mean a delay in the harvesting season since snow would require more time to melt and clear off.
For our project we interviewed Uonuma City government, Retailer (AEON) and a leading snow removal company. We also collected data from them about the snow fall and the cost budget for the last 15 years.
Findings from Interview with Minami Uonuma City office, and Snow removal company:
From our survey, we could find that the snow removal activities are carried out by construction companies and businesses in Minami Uonuma region. In fact, the president of a leading construction company quoted ““In winter, we rely on snow removal business”.
Construction companies have the following main advantages to engage in snow removal business:
- In winter the constructions business is minimal. Engagement in snow removal business provides a good opportunity in winter months.
- Employees at construction companies are familiar with operating machines for snow removal.
- Employees at construction companies have drivers license of large-size car. So the employees can be directly deployed in snow removal activities.
Types of Snow Removal businesses:
Type / Client1 / Road Maintenance / Government (s)
2 / Avalanche Patrol
(Elimination of Avalanche Risk) / Government (s)
3 / Maintenance of Snow-Covered Roofs / Individuals
4 / Motor Park Maintenance / Private Companies
5 / Piste Maintenance / Ski Resorts
There are 5 main types of snow removal businesses that these companies engage in. Details are shown in the table above. Road maintenance is the biggest business. City government and municipalities are the clients. Avalanche patrol is also done by construction companies to eliminate or mitigate avalanche risk. The Snow removal companies do Motor Park maintenance and maintenance of snow-covered roofs for individuals and private companies.
Piste refers to a designated path down a mountain for snow sports. Minami Uonuma Area has many ski resorts. Maintenance of the Piste is done by Construction companies.
Issues with snow removal business:
Forecasting of snow fall:
We interviewed the President of a leading Snow removal company (Anonymity requested).
As per our interview, we found that it is almost impossible to forecast how snowfall they would before winter season. However, the amount of snow fall tends to fall within the certain range. For example, if November and January had more snowfall than average, February and March tend to have less snowfall.
Regarding short-term (daily, weekly) forecast, the company uses 1) isopiestic line, 2) aerotonometer and 3) meteorological chart at Noto area to alert and station work force for snow removal activity.
Wage costs:
For a snow removal company, fuel costs and labour costs are the major cost drivers.
Sales= (wage rate contracted with city office) x (working hours). Working hours could be affected by snow fall, whereas wage rates have been pushed down by city office recently. Thus, the president suggested that it will be meaningless if we simply compare sales with snowfall. We should consider the wage rate paid by city office.
Construction companies usually sign a contract with city office one year before snow season. Fluctuations in fuel price can affect profitability during this period.
We try to investigate suitable hedging techniques using weather derivatives for the road construction business. In the next section we will describe analysis of snow data followed by investigation of suitable derivative structures.