Commodities

Weather Derivative

Also known as: HDD/CDD swap, Temperature derivative, Weather hedge

A contract that settles on the temperature, not on any asset. Invented so an energy company could hedge a warm winter — the purest example of a derivative with no underlying you can own.

Asset class
Commodities (weather & energy)
Instrument type
Swap, option or future on an index
Traded
Mostly OTC; some CME-listed contracts
Typical users
Utilities, agriculture, retail, event organisers
1 · SnapshotThe one idea to remember
Key intuition: a weather derivative hedges volume risk, not price risk. The gas price hedge protects the margin per unit; the weather hedge protects the number of units.
2 · BeginnerWhat is it, really?

Most derivatives settle against a price. A weather derivative settles against a measurement — how cold a month was, how much rain fell, how much wind blew — at a named weather station.

The motivation is that many businesses are exposed to weather in a way no financial market covers. A gas utility sells less in a warm winter. An ice cream maker sells less in a cool summer. A ski resort has no season without snow. None of these losses are caused by a price move, so no price hedge fixes them.

The standard building block is degree days:

  • Heating degree days (HDD): for each day, how far the average temperature fell below a reference (18 °C / 65 °F). Cold days accumulate HDDs.
  • Cooling degree days (CDD): how far it rose above. Hot days accumulate CDDs.

A winter contract might pay €10,000 for every HDD below 1,200 across a season. If the winter is warm, HDDs are low, and the payment offsets the gas the utility did not sell.

3 · IntermediateHow it works in practice

The critical distinction: derivative, not insurance

FeatureWeather derivativeInsurance
TriggerAn index readingA proven loss
Proof requiredNone — the station reportsClaim, assessment, adjuster
Payout speedAutomatic at settlementWeeks to months
CoversOrdinary variationCatastrophic events
Basis riskYes — the index is not your lossMinimal by construction

Parametric settlement is the feature and the flaw. Payment arrives without argument, and it arrives whether or not you actually lost money — and, more painfully, sometimes fails to arrive when you did.

Basis risk, concretely

  • Geographic: your business is thirty kilometres from the reference station, and the weather differed.
  • Index: temperature was normal, but a single unseasonal week destroyed the selling season.
  • Business: revenue depends on weather plus competition, plus the economy — the correlation is real but never 1.
Worked example: a utility expects 1,300 HDDs and buys a swap paying €25,000 per HDD below 1,250. The winter delivers 1,150 HDDs — 100 below the strike — so the swap pays €2.5m against roughly €2.8m of lost gas margin. Most of the exposure is neutralised; the residual is basis risk, and it is a permanent feature.
4 · AdvancedPricing & valuation

Pricing without a tradeable underlying

Weather is not storable, not tradeable and not arbitrageable. No-arbitrage pricing has nothing to work with — you cannot replicate a temperature. Pricing is therefore actuarial:

$$ V_0 \;=\; e^{-rT}\Big( \mathbb{E}^{\mathbb{P}}[\text{payoff}] + \lambda\,\sigma_{\text{payoff}} \Big) $$
  • The expectation is taken under the real-world measure from historical station data — typically 20–40 years, detrended for urban heat-island effects and climate drift.
  • λ is an explicit risk-loading. There is no hedging argument to pin it down, so it is negotiated, and it is where the seller's margin lives.
  • More sophisticated desks model daily temperature directly — mean-reverting with seasonal mean and seasonal volatility — and simulate the index. Better tails, same fundamental dependence on the historical record.

The detrending problem is now the main problem

Pricing assumes the historical distribution describes the future. A warming trend violates that assumption in a directional way: HDD contracts systematically overpay sellers and CDD contracts systematically underpay them if the trend is ignored. Every desk detrends; nobody agrees on the right window or method, and the disagreement is a larger source of price dispersion than the volatility estimate.

Who is on the other side

  • Natural counterparties: a gas utility (hurt by warmth) and a power generator with heavy summer cooling load (hurt by cold) genuinely offset. These are the cleanest trades.
  • Reinsurers and specialist funds take the residual for the risk premium — weather is close to genuinely uncorrelated with financial markets, the same appeal that drives the cat bond market and one of the few real diversifiers in diversification.
  • Market size is modest by financial standards and concentrated in North American and European energy. The power and commodity swap markets dwarf it — but weather is the risk those markets cannot price.

The formulas above are standard textbook formulations, simplified for teaching. They explain the mechanism — they are not a valuation tool, and they will not reproduce a dealer’s price.

5 · Desk notesHow practitioners think about it
Practitioner note: the honest question before any weather hedge is how much of last year's revenue variance the index would actually have explained. Below roughly 60%, you are buying a lottery ticket correlated with your business rather than a hedge for it.