What does the weather mean to you? As a PhD student in mathematics who focuses on climate-related risk, I think about the weather not just in terms of my outfit or commute. To me, the weather is a commodity.
Part of my research explores pricing models for weather derivatives, which are financial contracts that help firms mitigate risk and protect revenue from changes in the climate—acting almost like a hidden insurance policy for companies when the weather deviates from the “norm.”
Weather derivates are very popular in the energy sector. An energy company would use temperature-based derivatives to protect revenue during warm winters, when the average person uses less heat. Similarly, the company can also use weather derivatives to hedge its losses during cool summers.
We also see weather derivatives utilized in agriculture. Farmers can purchase temperature and rainfall derivatives to hedge against drought, excessive rainfall, or unseasonable temperatures—all of which could affect farmers' crop yield and revenue.
However, unlike actual insurance policies, weather derivatives require no proof of damage or loss; they are paid when specific weather metrics, like temperature and rainfall, are met. So, the energy company’s payout would come if the average temperature during the winter rose to a certain degree. The farmer would be paid if the total rainfall for a growing season fell below a certain amount.
By purchasing weather derivatives—having these “insurance policies” to hedge losses—the energy company and the farmer protect the consumer from price surges when adverse weather conditions affect their business. Meaning, energy prices stay stable during warm winters, and the price of almonds doesn’t skyrocket during a drought.
Weather derivatives are often traded in what is called an “over-the-counter” (OTC) market. In simple terms, this means the contracts are negotiated directly between buyers and sellers rather than traded on a centralized exchange. In these markets, price and specific weather metrics are negotiated directly between the two parties. This differs from more traditional exchange-based markets, such as the New York Stock Exchange (NYSE) where transactions are standardized, and intermediaries play a larger role.
The first weather derivative was traded in 1997 on the Chicago Mercantile Exchange (CME). Since then, weather derivatives have been a growing market, expedited by the increase of extreme weather-related events globally. With the market growing, the process of pricing weather derivatives has become more important.
However, as any meteorologist at a local news station knows, weather is hard to predict. So how can we find a fair price for weather derivatives? That is what I explored in my recent project with the iRisk Lab, the University of Illinois Urbana-Champaign Math Department’s experiential learning hub. My goal is to develop mathematical models that provide practical benchmarks for pricing and negotiating these contracts. In other words, I want to help create more reliable reference points for deciding what these contracts are worth. That matters because better pricing can support better risk management when weather affects costs and revenues, which, in turn, protects the price of consumer necessities—like heat and groceries—from price surges. I find these opportunities to use mathematical modeling to solve real-world problems very fulfilling.
In practice, pricing a weather derivative depends on the contract’s location, time period, and underlying weather index. For temperature-based contracts, historical weather patterns help estimate the likelihood of a payout, but the final price is also shaped by market conditions and how much risk buyers and sellers are willing to take on. Information on the daily weather that we actually had in the past is very niche and often proprietary and owned various government agencies, requiring expensive licensure to use, so sourcing historical weather data was one of the biggest hurdles I faced in my project.
I reached out to professors in the Gies College of Business, and although they couldn’t get me access to the data I needed, they did put me in contact with Anne Krema, director of commodity research and product development at CME Group, which is the operator of the world’s largest derivatives marketplaces such as the CME. I had the pleasure of interviewing Krema and our conversation gave me a deep understanding of CME Group’s trading practices and how weather derivatives are priced and traded in the real world.
Krema explained that CME Group primarily offers temperature-based weather derivatives tied to cities in the U.S., Europe, and Asia. Depending on the location, these contracts are priced using temperature indices. In U.S. cities, the main indices are Heating Degree Days (HDD) and Cooling Degree Days (CDD), which are commonly used as indicators of heating and cooling demand.
These indices are based on how far the mean daily temperature deviates from a reference point, usually 65°F. HDD measures colder-than-usual conditions by tracking how much the temperature falls below that benchmark, while CDD measures warmer-than-usual conditions by tracking how much it rises above it.
In European cities and Tokyo, CME Group also uses Cumulative Average Temperature (CAT), which sums daily average temperatures over a set period. Like HDD and CDD, CAT can help energy companies estimate demand, which can in turn affect the price of weather derivatives.
Krema offered many insights that will inform my iRisk Lab project going forward. A key takeaway from my discussion with her was that weather derivative prices are often shaped more by direct negotiation between buyers and sellers than by exchange trading alone. I saw this explained in the academic research I read for my iRisk Lab project. The research explained that over-the-counter markets tend to be less active and less standardized than many traditional financial markets, reliable pricing models become especially important. For that reason, strong pricing models can give firms a more informed basis for evaluating and negotiating these contracts.
Hearing this confirmed by Krema reinforced the need to develop models that can serve as rational benchmarks in negotiations while also deepening understanding of how climate-related financial instruments operate in practice.
Going forward, I plan to refine our models by making them closer to real-world conditions, while balancing realism with mathematical tractability. I am particularly interested in exploring how more advanced modeling techniques can capture the complexity of climate-related risks without sacrificing practical usability.
Ultimately, I hope this work contributes to more effective tools for managing climate risk in financial markets and beyond. In the long run, better tools for managing weather-related risk may help reduce some of the cost instability that extreme weather can create for businesses and consumers alike.
Want to read more?
Check out the transcript of Yasintorn's interview with Anne Krema.