Finance that harnesses the weather: uncertain temperatures, secure returns

Financial operators who trade natural gas futures are exposed to weather risk as much as farmers are, but with one fundamental difference: the farmer is subject to the weather, while the financial operator can choose whether to expose themselves to the uncertainty of forecasts. And when they do, according to the results of a recent study, they earn a risk premium that delivers higher returns than the S&P 500 index.

As Stefano Caselli, Manou Monteux, Maria Cristina Arcuri and Gino Gandolfi observe, forecasts of extreme cold generate a risk premium in natural gas futures that can be exploited systematically. In other words, investors willing to base their decisions on uncertain weather forecasts can, in the run, earn extra returns.

The strategy advanced by the authors produces returns higher than the S&P 500 both in absolute terms and on a risk-adjusted basis. This means that, for each unit of risk investors bear, the weather-forecast-based strategy generates more returns than a traditional equity investment.

Gas prices and futures prices

The link between weather and energy has long been well known. Natural gas is strongly influenced by temperature: when it’s cold outside, demand for heating obviously increases; when it’s hot, there is more demand for cooling. For this reason, economic literature has long studied how weather affects gas prices, especially in terms of volatility. What is less clear, however, is if and how the weather influences the returns on gas futures. In particular, most research does not clearly distinguish between observed weather (what has already happened) and forecast weather (what might happen in the coming days or weeks).

The study fits precisely into this gap and revolves around three questions:

• Do markets react to observed temperatures or to forecasts?
• Are forecasts of extreme events and normal weather treated differently?
• Is the risk linked to forecast uncertainty compensated by the market?

Differences in value

The researchers constructed a massive dataset combining 30 years of data (1990–2019) on daily prices of US natural gas futures (Henry Hub), temperatures observed hourly, and weather forecasts at 1-day, 1-week, and 2-week horizons. In total, the analysis is based on more than 7,400 days of observations, a solid foundation for drawing robust conclusions. The main methodology is the event study, used to understand how markets react to clearly identified events. In this case, the events are forecasts of temperatures much colder than normal, defined as those falling within the most extreme 10% relative to seasonal averages.

The strategy under investigation consists of buying the nearest-maturity natural gas future (NG1) and selling the slightly longer-dated one (NG2) when forecasts for the following two weeks indicate extreme cold. The rationale behind the strategy is that for the most part, cold weather drives up the value of gas that is needed immediately, not gas that will be needed later. So by buying the first and selling the second, this difference can be exploited. In other words, if the forecast cold materializes, investors gain by selling NG1 and buying back NG2. But forecasts can be wrong, which means operators take on a risk, which the market prices and rewards.

In essence, the bet is not on the price of gas in general, but on the difference in value between gas needed immediately and gas needed later, a difference that tends to widen when cold weather is imminent. Working on this difference exposes the operator only to weather risk. If, instead, they simply bought NG1, they would also be exposed to other risks that affect gas prices (for example, geopolitical ones). Since these risks affect both NG1 and NG2, buying the first and selling the second offsets them.

A better risk–return trade-off

Weather forecasts can generate excess returns. In particular, forecasts of extreme cold one or two weeks out trigger strong movements in the front part of the futures curve, creating a risk premium that can be exploited. Over the long term, this strategy not only produces high returns, but does so with a better risk–return trade-off than the S&P 500. The strategy does not earn more simply because it takes more risk, but instead because it uses risk more efficiently. As economists would say, it has a higher risk-adjusted return. The Sharpe ratio, which measures how much reward is obtained for each unit of risk accepted, is three times higher for this strategy than that same ratio for the S&P 500.

The paper discusses the implications of these results in the context of climate change. The expected proliferation of extreme weather events makes the weather risk premium potentially more relevant over time, with effects both on costs for final consumers of natural gas and on risk management strategies in the energy sector. In light of these trends, the study contributes to a better understanding of the interaction between climate, information, and financial markets, opening new avenues of research on risk premiums linked to exogenous variables and their forecast uncertainty.

Manou Monteux, Maria Cristina Arcuri, Gino Gandolfi, Stefano Caselli, “Can extreme weather forecasts lead to a risk premium? Evidence of a non-linear response in U.S. natural gas futures.” The North American Journal of Economics and Finance, Volume 80, 2025, 102494. DOI: https://doi.org/10.1016/j.najef.2025.102494.