Compendium

Weather and Energy Derivatives — HDD/CDD, Wind, Rainfall, Parametric, and Commodity Hedging

22 min read

Weather and Energy Derivatives — HDD/CDD, Wind, Rainfall, Parametric, and Commodity Hedging

Weather is the single largest exogenous driver of energy demand and renewable supply. A 1°F shift in average summer temperature across PJM raises electricity load by roughly 0.7-1.0%, or 1-1.5 GW; a colder-than-normal winter in the EU Continental synchronous area can drive €1-3B of incremental gas-procurement spending across the heating season; a low-wind year in Europe (the Dunkelflaute of December 2021 and the 2021 full-year UK wind deficit ~15% below normal) can convert an otherwise profitable wind operator into a loss-maker through PPA + corporate-procurement shortfalls. The financial product universe that has grown to hedge weather and energy risk spans temperature-indexed futures and options (CME HDD/CDD, multi-city), wind production-index swaps (Speedwell-developed, Nephila + Swiss Re underwritten), solar irradiance derivatives (early-stage, primarily UK/EU), parametric covers (Demex, Arbol, Skyline Partners), the deep commodity-derivatives stack (NYMEX/ICE/CME oil + gas + LNG + cracks + refined-product spreads), the ISO-traded power-futures complex (Nodal, ICE PJM/ERCOT/NYISO/CAISO + EEX/EPEX + ASX/AEMO), carbon derivatives (ICE EUA, UKA, RGGI, CCA), and the financial transmission rights (FTR/CRR/ARR) markets that hedge intra-ISO congestion. Trading is conducted by physical hedgers (utilities, retailers, gas shippers, agribusiness, ski resorts, sports leagues, amusement parks), proprietary financial participants (Trafigura, Vitol, Gunvor, Mercuria, Hartree, Citi, BNP, Macquarie, Shell Trading, BP IST, Equinor, ENGIE GEM, JP Morgan), and a niche of weather-specialist desks (Demex, Aclima, Arbol, Hartree weather, Vesttoo before its 2023-2024 collapse). This note covers the product taxonomy, pricing methodology, basis risk structure, and the standardized contract architecture that has emerged across these markets.

See also

1. Heating and Cooling Degree Days — the building block

The Heating Degree Day (HDD) and Cooling Degree Day (CDD) indexes are the workhorses of weather-derivative markets. The standard definitions:

Daily mean temperature   T_avg = (T_max + T_min) / 2
HDD_day = max(0, 65°F − T_avg)        [°F basis, US convention]
CDD_day = max(0, T_avg − 65°F)        [°F basis, US convention]
 
European convention uses 18°C and reports in °C·day units:
HDD_day = max(0, 18°C − T_avg)

The 65°F (18.3°C) threshold reflects the long-standing utility heuristic that buildings consume neither heating nor cooling energy near that ambient temperature. Both indexes are summed over a month or season to produce a tradeable index value. Typical monthly indexes for major US cities:

CityJan HDDJul CDDAnnual HDDAnnual CDD
Chicago O’Hare (KORD)1,150-1,250270-3606,200-6,800800-1,100
New York LaGuardia (KLGA)950-1,100280-3604,700-5,3001,000-1,300
Atlanta Hartsfield (KATL)500-650380-4502,700-3,2001,700-2,000
Dallas-Fort Worth (KDFW)420-550530-6202,200-2,7002,600-2,950
Houston Bush (KIAH)250-380540-6001,400-1,8002,900-3,200
Las Vegas McCarran (KLAS)600-720620-6802,200-2,5003,200-3,600
Sacramento Executive (KSAC)380-480410-4902,400-2,9001,400-1,800

The mathematical aggregation of HDD/CDD over a month or season creates the index value; CME contracts settle to NOAA NCEI station data with strict adjustment rules.

2. CME HDD/CDD futures and options — the canonical market

The Chicago Mercantile Exchange launched HDD futures in September 1999 as a hedging tool for utility procurement and gas-shipper portfolio management; CDD futures followed in 2000. Over time the market grew to cover 12+ US cities, 11 European cities (London, Paris, Amsterdam, Berlin, Madrid, Rome, Essen, Stockholm, Oslo, Helsinki, Prague), 6 Japanese cities (Tokyo, Osaka, Hiroshima, Nagoya, Sapporo, Fukuoka — discontinued briefly in 2022 then relaunched 2023 as energy crisis renewed interest), and seasonal strip products (full-winter Nov-Mar, full-summer Apr-Sep).

CME contract mechanics:

  • Notional value: $20 per index point (US contracts); €20 (European); ¥2,500 (Japanese)
  • Tick size: 1 index point
  • Settlement: cash, against the monthly aggregated HDD/CDD value reported by Speedwell Weather Derivatives Ltd. on behalf of CME (Speedwell is the appointed index calculation agent)
  • Trading hours: nearly 24-hour electronic on CME Globex
  • Months listed: typically 7 cooling-season months (April-October) for CDD and 7 heating-season months (October-April) for HDD per city, on a rolling basis
  • Options: European-style cash-settled on the monthly index; key strikes around the seasonal mean

Trading volumes peaked in the mid-2000s at roughly $3-5B of notional outstanding before the 2008 financial crisis crippled the OTC weather market; since 2018 volumes have recovered as renewable-portfolio hedging, ESG-driven climate-risk disclosure (TCFD, SEC Climate Rule pending), and parametric-insurance demand has expanded. By 2024 weather-derivative notional outstanding is estimated at $15-20B globally (WRMA + Hartree weather desk + Aclima estimates), with the bulk OTC bilateral and CME-cleared contracts representing the price-discovery anchor.

Industry users of HDD/CDD derivatives

  • Electric utilities (PG&E, ConEd, Duke, Dominion, Southern Co, NRG, Vistra, Exelon, TVA, ENGIE, EDF, RWE, Vattenfall) — hedge weather-driven load risk in regulated and competitive markets.
  • Natural gas LDCs (National Grid US, ConEd Steam + Gas, Atmos Energy, Southwest Gas, Centrica, E.ON) — hedge winter throughput risk against heating-driven gas demand.
  • Gas shippers and traders — hedge pipeline capacity utilization vs heating demand.
  • Retail electric providers (Texas REPs especially: Reliant, TXU, Constellation, Direct Energy, Octopus, Rhythm) — hedge customer load against TOU exposure.
  • Agribusiness — Cargill, ADM, Bunge (corn + soy yield correlated with growing-degree days)
  • Ski resorts — Vail Resorts, Alterra, Boyne Resorts use snowfall + temperature derivatives
  • Theme parks and outdoor venues — Disney, Universal, Six Flags, Cedar Fair hedge attendance vs weather
  • Sports leagues + outdoor event organizers — MLB, NFL outdoor venues, Wimbledon (London rain), French Open (Paris)
  • Construction firms — hedge cold-weather concrete-pour delays

3. Pricing weather derivatives — burn analysis vs index modeling

Pricing a weather derivative requires forecasting the distribution of the index value at settlement. Two predominant approaches:

Burn analysis (“historical simulation”)

Compute the index value that would have settled in each of the last 20-30 years from station-quality temperature records. The empirical distribution of those values, possibly weighted toward recent years, becomes the pricing distribution. Burn analysis is appealing for its simplicity, robustness against model misspecification, and stakeholder-defensibility. Critiques: it cannot extrapolate beyond the observed range; it under-represents climate-change-trended distributional shifts; sample sizes (~30 years) yield wide confidence intervals on tail percentiles.

Standard burn analysis adjustments:

  • Detrend the historical series to remove climate trend (typically subtract a linear-regression trend on average temperature)
  • Reweight recent observations more heavily (exponential decay with half-life 10-15 years)
  • Composite with longer reanalysis data (ERA5, NCEP/NCAR reanalysis) for outliers

Index modeling (parametric / time-series)

Fit a stochastic model to daily mean temperature, then Monte-Carlo simulate forward to compute the index distribution. The canonical reference is Davis (2001) — “Pricing Weather Derivatives by Marginal Value” — which derives a no-arbitrage option-pricing approach where the underlying is a non-traded asset, and shows that the marginal utility approach (pricing in terms of the buyer’s risk-adjusted utility) is more appropriate than risk-neutral pricing for non-tradeable underlyings.

Standard temperature dynamics models:

  • Ornstein-Uhlenbeck mean-reverting Brownian motion — the simplest specification:

    dT_t = κ(θ(t) − T_t) dt + σ(t) dW_t

    where κ is mean-reversion speed, θ(t) is the seasonal mean function (sinusoidal annual cycle plus climate trend), σ(t) is seasonally-varying volatility (higher in winter, lower in summer for mid-latitudes). Calibration via maximum likelihood on detrended residuals.

  • ARMA-GARCH on daily anomalies — Alaton, Djehiche, Stillberger (2002) is the canonical academic reference for this approach; fit ARMA(p,q) to deseasonalized daily mean temperature, fit GARCH(1,1) to residuals for time-varying volatility. Yields a richer distribution than simple OU but adds complexity.

  • CARMA (Continuous-time ARMA) — Benth-Saltyte Benth approach (2007, 2013) extends to continuous-time analogs and handles option pricing analytically.

  • Stochastic volatility / regime-switching extensions — capture Polar Vortex regime + ENSO regime states.

Burn vs index — practical synthesis

Most desks use both approaches in tandem: burn analysis for the central estimate and historical-percentile range; index modeling for option-strike-specific pricing (where you need a smooth distribution function), tail-risk assessment, and scenario stress-testing under specified climate-shift assumptions. The implied volatility surface for HDD/CDD options is sparse but observable for the most-liquid cities (Chicago, New York, Atlanta), informing model calibration.

4. Basis risk — the structural challenge

A weather derivative pays based on a station-measured index (typically a single NOAA NCEI station, often the major airport — KORD for Chicago, KJFK or KLGA for New York, KATL for Atlanta, etc.). The hedge user’s actual exposure is over a much wider geographic area. Basis risk is the residual between station-index payoff and user-exposure economics:

  • Spatial basis — distance between the station and the user’s load center; a generic Chicago city HDD hedge may not perfectly track a Chicago suburban utility’s load
  • Sensor basis — even at a single airport, microclimate (asphalt heat-island, instrument shelter age) can introduce a small bias
  • Index-definition basis — the standard 65°F threshold may not exactly match the user’s load-temperature elasticity break-point (commercial cooling kicks in around 60°F for some HVAC stocks; residential heating in cold-climates kicks in below 55°F for well-insulated homes)
  • Population-weighted vs station basis — utility load is correlated with population-weighted temperature across service territory; standardized contracts use a single station, leaving residual

Basis risk has driven the growth of custom OTC products — bespoke indexes built from multiple stations with weights matching the user’s load distribution, or from proprietary mesh datasets (PRISM, NOAA Climate Prediction Center analyses, ERA5 reanalysis interpolation). These custom products trade off liquidity for hedge effectiveness.

5. Wind production derivatives

Wind output is the integral over time of 0.5 · ρ · A · v³ · Cp, with very strong cubic dependence on wind speed v. A 10% miss on annual average wind speed translates to ~30% miss on output. This non-linearity makes wind-speed indexes inadequate for hedging; instead, wind-production indexes based on either turbine-specific power-curves or measured park outputs have emerged.

The Speedwell Wind Index family

Speedwell Weather (UK, founded 1999 by Stephen Doherty) pioneered the standardized wind-production index. Key products:

  • Speedwell Wind Production Index — site-specific, calculated from MERRA-2 reanalysis wind data put through a standardized power curve, with bias correction against actual on-site measurements where available.
  • Speedwell Daily Production Index (DPI) — daily granularity for shorter-tenor structures.
  • Speedwell Site Variability Index — a measure of expected production variance for capacity-pricing.

Underwriters of wind production swaps (typically a producer + a swap counterparty exchange a fixed production amount for the floating Speedwell-index amount, settled monthly or quarterly):

  • Nephila Capital (Bermuda-based ILS specialist; now Markel-owned) — one of the largest catastrophe + weather underwriters
  • Swiss Re Corporate Solutions — Zurich; large weather + parametric book
  • Munich Re — major reinsurer with weather book
  • Hartree Partners — proprietary weather + commodity desk
  • Statkraft, Vattenfall, Enel Green Power, Iberdrola internal weather risk-management groups

Wind production swaps are commonly structured as Wind Index Swaps with:

  • Notional: typically MWh-equivalent (e.g., 50,000 MWh annual at a multiplier of €40/MWh = €2M notional)
  • Strike: the long-term-mean P50 production
  • Cap and floor: to manage extreme-tail tail premiums
  • Tenor: 1-10 years (longer tenors increasingly common with renewable PPA dis-application)

Solar irradiance derivatives

Solar derivatives are commercially less developed than wind, partly because solar variance is lower at annual horizons (cloud variability averages out somewhat) and partly because solar PPAs themselves have provided implicit hedges. Emerging products:

  • Speedwell Solar Index — irradiance + bias-correction index for site-specific PV production
  • CelsiusPro Solar Variability Cover — Swiss parametric provider
  • UK + DE pilot solar production swaps under early underwriting by Swiss Re and Nephila

6. Rainfall and ENSO derivatives

The CME launched Rainfall futures for select US cities in 2011 and an ENSO Index futures product in 2013 covering the multivariate El Niño Southern Oscillation index (computed by NOAA from Pacific sea-surface-temperature anomalies). The products are thinly traded but provide reference pricing for OTC structures. Industry users:

  • Hydroelectric operators (PG&E, BC Hydro, Northwest IPPs, Brazilian Eletrobras + Itaipu + IPP fleet) — hedge against drought-driven hydro shortfalls
  • Agribusiness — corn/soy/wheat yield correlations to growing-season precipitation
  • Insurance reinsurance market — parametric ENSO covers as proxy for hurricane and drought risk

7. Parametric weather covers — the insurance overlap

A parametric weather cover is structured legally as insurance but financially as a weather derivative — payout triggered by index value reaching a threshold, with no claims-adjustment. The advantage over indemnity insurance: instant settlement, no loss-causation dispute, no moral hazard around proof-of-loss. The disadvantage: basis risk between index and actual loss.

Major parametric weather underwriters:

  • Demex (US, founded 2020 by Bill Clark, ex-Swiss Re) — climate-volatility parametric covers for renewable developers, utility procurement, agribusiness
  • Arbol (US, founded 2018) — blockchain-anchored parametric weather covers; agricultural focus
  • Skyline Partners (UK) — parametric specialist for SME and corporate
  • Descartes Underwriting (France) — parametric covers for natural-catastrophe and weather
  • Floodbase (formerly Cloud to Street) — parametric flood covers
  • Mitiga / Climate-i / Sola Climate — emerging climate-parametric platforms

The 2023 collapse of Vesttoo (Israeli InsurTech that had structured ~$3.5B of parametric and reinsurance products allegedly backed by fraudulent letters of credit) tarnished the parametric reinsurance market briefly but did not derail the broader product trend.

8. Commodity derivatives — the deep stack

The energy-commodity derivatives market is dominated by NYMEX (CME), ICE, EEX, and ASX, with bilateral OTC volumes substantially exceeding cleared. Major products:

Natural gas

  • NYMEX Henry Hub natural gas futures (NG) — the canonical North American benchmark; $/MMBtu, monthly contracts out to 12+ years, options on futures. Daily volumes ~$10-20B notional in 2024. Henry Hub is a physical hub in Erath, Louisiana, with 13 interconnecting interstate pipelines.
  • ICE TTF (Title Transfer Facility) — the European gas benchmark, traded as monthly cal-month, quarter, and seasonal contracts in €/MWh. TTF replaced NBP (UK National Balancing Point) as the dominant EU benchmark progressively 2010-2020; the 2022 gas crisis cemented TTF dominance with TTF prices exceeding €300/MWh in August 2022.
  • ICE NBP UK natural gas — UK benchmark, in p/therm; smaller volumes than TTF but key for UK hedging.
  • ICE JKM (Japan-Korea Marker) LNG — the canonical Asia LNG benchmark, in $/MMBtu; published daily by S&P Platts. JKM futures launched ICE 2014, now the most-traded LNG benchmark globally.
  • CME NGSL (Henry Hub spreads) and CME Brent NG spread products — basis trades.
  • NYMEX HH options — extensive American-style options market with deep liquidity 12+ months out.

Crude oil and refined products

  • NYMEX WTI (West Texas Intermediate) crude futures (CL) — the canonical US crude benchmark, $/bbl, Cushing OK physical delivery, monthly out to 10+ years. WTI dropped to negative $37/bbl on 20 April 2020 (the front-month May contract) as COVID demand collapse + Cushing storage overflow forced sellers to pay for offtake.
  • ICE Brent (BFOE: Brent-Forties-Oseberg-Ekofisk-Troll) — the global crude benchmark, North Sea physical with paper trading. Spread to WTI averaged $3-5/bbl 2015-2023.
  • NYMEX Heating Oil / ULSD (HO) — distillate futures
  • NYMEX Gasoline RBOB (RB) — gasoline futures
  • NYMEX Crack spreads — 3:2:1 crack spread (3 bbl crude = 2 bbl gasoline + 1 bbl distillate); 5:3:2 variant. Major refiner hedging.
  • Argus Far East Index + Argus Sour Crude Index (ASCI) — published indexes for non-WTI/Brent benchmarks; some cleared products at CME and ICE.
  • Jet fuel — primarily OTC, with Gulf Coast jet kero, Singapore jet, North Sea jet differentials.
  • Gasoil (ICE Gasoil) — European distillate benchmark.
  • Marine fuels (VLSFO, MGO, HSFO) — post-IMO 2020 sulfur cap, new product complex.

Coal and emissions-coal

  • ICE Newcastle Coal (Newcastle thermal coal, Australian export) — Pacific benchmark, $/tonne
  • ICE API2 (ARA — Amsterdam-Rotterdam-Antwerp coal) — European import benchmark
  • CME Powder River Basin coal — US Western coal
  • All thermal coal markets have declined sharply in 2020s as coal-generation retires.

9. Power futures — Nodal Exchange, ICE, CME, EEX

US ISO-traded power futures clear through:

  • Nodal Exchange — the dominant US power futures venue, owned by EEX (acquired 2017). Settles in cash against ISO-published nodal LMP indexes. Major products:
    • PJM Western Hub Real-Time Peak / Off-Peak
    • PJM AEP-Dayton Hub
    • PJM DOM Zone
    • ERCOT North / Houston / South / West Hub Real-Time and Day-Ahead
    • ERCOT 7x24 + 5x16 + 2x16H + 5x8 strips
    • NYISO Zone A / Zone G / Zone J Real-Time and Day-Ahead
    • ISO-NE Mass Hub Real-Time and Day-Ahead
    • CAISO SP15 / NP15 / ZP26 Real-Time and Day-Ahead
    • MISO Indiana Hub / Michigan Hub
  • ICE — competitor venue, similar product suite plus PJM Western Hub, NEPOOL Mass Hub, EPEX SPOT German Phelix futures.
  • CME — PJM West DA Off-Peak, ERCOT North, CAISO SP15 cleared products.
  • EEX (European Energy Exchange) — German + French + Italian + Spanish + UK power futures (Phelix-DE, Phelix-FR, IPEX, PUN, OMIP, N2EX). EEX Power Futures Cal-Y (calendar year), Q (quarter), M (month) products are the deepest in the EU.
  • EPEX SPOT — day-ahead and intraday physical clearing across the Continental synchronous area.
  • Nord Pool — Nordic + Baltic + UK day-ahead clearing (EUPHEMIA algorithm via Nord Pool); DS Futures for futures hedging.
  • ASX Energy — AEMO NEM power futures (NSW, VIC, QLD, SA peaks and baseloads).

10. Financial Transmission Rights — congestion hedging

When transmission constraints bind, LMPs differ between source and sink. Financial Transmission Rights (FTRs) — and the analogous Congestion Revenue Rights (CRRs) at CAISO and ERCOT — let market participants hedge congestion. Mechanics:

  • An FTR is a financial instrument with a defined source node and sink node and a defined MW quantity.
  • It pays the holder (LMP_sink − LMP_source) × MW × hours for the contract period.
  • If LMP_sink > LMP_source, the FTR pays positive cash to the holder; if reversed, the holder pays.

Two FTR variants:

  • Obligation FTRs — pay/receive the full congestion differential, regardless of sign.
  • Option FTRs — pay only when the differential is positive (holder has the option to walk away from negative side); cost more in the auction.

Each ISO/RTO runs:

  • Annual FTR auction (typically 3-year horizon in PJM, 1-2 year in others)
  • Monthly FTR auction — additional procurement
  • Long-term FTR auctions — multi-year strips (PJM, NYISO offer up to 3 years; CAISO up to 10 years)
  • Auction Revenue Rights (ARRs) — load-serving entities receive ARRs that pay the auction revenue when FTRs clear; ARRs can be converted to FTRs in self-scheduled or annual auctions.

FTR market structure: deeply liquid in PJM ($2-4B annual auction revenues), MISO ($1B), NYISO + ISO-NE ($300-500M each), CAISO ($200-400M), ERCOT (~$300-500M). Primary participants include:

  • Load-Serving Entities (utilities and REPs) — hedge their physical load LMP exposure
  • Generators with bilateral PPAs into foreign congestion zones — hedge counter-flow
  • Financial participants — JP Morgan, DC Energy, Vitol, Castleton Commodities, Twin Eagle, Boston Pacific, IEEE, and specialist hedge funds (e.g., Modo Energy, Hartree, Greenwood Energy, EDF Trading, others)

The 2007 DC Energy and Allegheny Energy FTR investigation exposed the magnitude of financial participation; FERC and ISO market-monitoring units have refined surveillance over time. The PJM uplift-cost-allocation disputes have repeatedly highlighted that FTR market participants externalize uplift costs onto load — a chronic complaint.

11. Carbon derivatives

Cap-and-trade carbon markets have spawned futures and options:

  • EU ETS Allowances (EUAs) — ICE EUA futures are the canonical carbon derivative. Phase 4 (2021-2030) prices ran €60-95/tonne CO₂ in 2023-2024. Options on EUAs cleared at ICE.
  • UK ETS Allowances (UKAs) — post-Brexit UK ETS launched 2021. ICE Endex platform. Prices generally tracked EUA with modest discount.
  • RGGI Allowances — Regional Greenhouse Gas Initiative quarterly auctions; secondary trading on ICE. 2024 prices $15-25/short-ton CO₂.
  • CCA (California Carbon Allowances) — under California’s Cap-and-Trade Program (linked with Québec since 2014). ICE CCA futures + options. 2024 prices $30-45/metric-ton.
  • WCI (Western Climate Initiative) — the joint CA + Québec system; Washington’s Climate Commitment Act linkage initiated 2023.
  • CFE (CFE2 nature-based offset futures) — ICE-cleared voluntary carbon-offset benchmark futures launched 2022.
  • CBL Group voluntary-market spot + futures — voluntary carbon-credit standardized contracts on the Xpansiv platform.

EUA-EUR-Cal-25 (calendar 2025 ICE EUA future) had open interest of ~500K lots (each lot = 1,000 tonnes) in mid-2024 — ~500 Mt CO₂ notional, on a market that issues ~1,500 Mt/year EU + UK allowances.

12. Option strategies — the trader’s toolkit

Energy and weather option strategies are largely identical to financial options:

  • Long call / put — directional volatility play.
  • Straddle / strangle — long volatility, betting on large moves either direction.
  • Calendar spread — buying near-month, selling deferred-month (or vice versa); plays time-decay + roll-yield.
  • Crack spread option — option on the refiner spread between crude and refined products.
  • Heat-rate-linked option — option on the spark spread, useful for gas-fired generator hedging.
  • Collar — long underlying, long put (downside floor), short call (premium-financing); used by utility procurement for capped + floored gas/power prices.
  • Risk-reversal — long call + short put (or reverse); volatility-neutral directional bet, common for sophisticated procurement hedging.
  • Ratio spreads — typically 1-by-2 ratio (e.g., long 1 call ATM, short 2 calls OTM); high-volatility view at low net premium.

For weather options, the analogous structures apply but the underlying is non-tradeable, requiring care in delta-hedging (delta computed against the implied temperature index forward, not a directly tradeable instrument).

13. ISDA, master agreements, and clearing

The legal architecture for energy + weather derivatives standardized progressively from the 1999 ISDA 1992 Master Agreement, the 2002 ISDA Master Agreement, and the ISDA Equity Definitions + ISDA Commodity Definitions through to the ISDA US Power Annex (2007 revised 2017) + ISDA Weather Derivatives Definitions (1999 with multiple supplements). Standard contract terms include:

  • Bilateral OTC trade execution — confirmation via long-form ISDA confirmation
  • Collateralization via Credit Support Annex (CSA) — daily mark-to-market, posting of variation margin
  • Close-out netting under ISDA — bankruptcy-proof bilateral netting
  • Clearing for cleared products — CME ClearPort, ICE Clear, LCH SwapClear (for some weather/commodity trades), EuroNext.liffe (formerly ICE).

Dodd-Frank Title VII (2010) and EMIR (EU 2012, refit 2019) mandate clearing of standardized swaps, with several major energy derivatives within scope. Anti-manipulation rules (FERC EPAct §1283, CFTC EPAct §1283 parallel, MAR EU, MAD UK) cover trader conduct.

14. Major dealer / market-maker landscape (2026 snapshot)

The energy + weather derivatives market is dominated by a handful of major dealers, trading houses, and prop desks:

Trading houses (physical + financial):

  • Trafigura — Geneva-based; major physical metals + oil + gas + LNG trader; large carbon + power book
  • Vitol — Geneva + Singapore; largest independent oil trader by physical volume
  • Gunvor — physical oil + gas, expanding power and renewables trading
  • Mercuria — Geneva; oil + gas + carbon + power
  • Hartree Partners — NYC + London; specialist energy + weather + carbon prop desk; major weather-derivative book
  • Glencore — major physical commodities, trading book
  • Cargill — agricultural + power + carbon trading (post-2022 wind-down of metals + energy desks but still active)
  • BB Energy — Geneva-Beirut, growing power
  • Vesttoo (collapsed 2023-2024) — historical participant

Oil major trading arms:

  • Shell Trading and Supply — London + Houston + Singapore; largest oil major prop desk
  • BP IST (Integrated Supply & Trading) — major NYMEX participant
  • Equinor Trading and Marketing — Stamford + Stavanger; major Atlantic gas + LNG
  • TotalEnergies Trading — Geneva
  • ENI Trading — Rome + London
  • Chevron Marketing and Supply
  • ExxonMobil Trading

Banks (commodities desks remaining post-Volcker Rule + post-Dodd-Frank exit waves):

  • JP Morgan Commodities — kept most commodities desk; major power + FTR + carbon
  • Citi Commodities — significant power + gas
  • BNP Paribas — Paris; major EU power + gas + carbon
  • Macquarie — Sydney + London + NY; major US power + gas + carbon
  • Morgan Stanley — commodities (downsized 2014-2018; some return 2022+)
  • Goldman Sachs — commodities (significantly downsized 2017-2021; partial return)

Utility / generator trading desks:

  • ENGIE Energy Management (GEM) — Paris; major EU power + gas
  • EDF Trading — major French utility prop arm
  • NextEra Energy Resources marketing — Juno Beach FL; major US renewables + power
  • Constellation (Exelon) — competitive power book
  • Vistra Energy — Texas + national; growing power book
  • Talen Energy commercial desk — PJM heavy
  • RWE Supply & Trading — Essen; major EU power
  • Statkraft Energy — Oslo; Nordic power + green-trading
  • Iberdrola Trading — Madrid + London
  • Centrica Trading — UK gas + power

Specialist weather + parametric desks:

  • Nephila Capital (Markel) — Bermuda + Hamilton; largest weather + cat ILS
  • Swiss Re Corporate Solutions — Zurich; weather + parametric
  • Munich Re Markets — Munich; weather + parametric
  • AXA Climate (formerly XL Catlin Climate) — Paris; weather + parametric
  • Aon Reinsurance Solutions — broker + risk-finance for weather
  • Marsh McLennan Climate & Resilience — weather + parametric advisory
  • Demex — US climate-volatility parametrics
  • Arbol — blockchain parametrics
  • Speedwell Weather — UK index calculation agent + custom indexes

15. The frontier — climate-trended pricing, AI forecasts, and integrated risk

The classical burn-analysis assumption that historical weather is the best proxy for the future is breaking down under accelerating climate change. Modern weather-derivative pricing increasingly incorporates:

  • Climate-conditioned burn analysis — historical observations re-weighted by similarity to current and forecast climate state (ENSO phase, NAO/AO, decadal modes)
  • Reanalysis-based pricing — ERA5 (ECMWF, 1940-present, 0.25° grid) and MERRA-2 (NASA, 1980-present) provide consistent multi-decade time-series at gridded resolution, replacing station-only history
  • Seasonal forecast integration — ECMWF SEAS5, NCEP/NMME, UK Met Office GloSea for sub-seasonal-to-seasonal (S2S) forecasts; some desks have systematized incorporation of S2S signals into option-pricing models
  • Climate-scenario stress testing — pricing tail percentiles under SSP1-2.6, SSP2-4.5, SSP5-8.5 IPCC scenarios; particularly important for multi-year structured products
  • AI forecasting models — Pangu-Weather (Huawei, 2023), GraphCast (DeepMind, 2023), FourCastNet (NVIDIA, 2022) outperform classical NWP at 1-10 day horizons; integration into trading desks is in early stages but accelerating

The integration of weather risk with the broader energy + commodity stack is accelerating: utilities increasingly model joint distributions of HDD/CDD, gas prices, power prices, wind production, and load to optimize hedging across the full risk surface rather than product-by-product.

Further reading

  • Davis, M. (2001), “Pricing Weather Derivatives by Marginal Value,” Quantitative Finance 1(3), 305-308 — the foundational utility-based pricing paper
  • Alaton, P., Djehiche, B., Stillberger, D. (2002), “On Modeling and Pricing Weather Derivatives,” Applied Mathematical Finance 9(1), 1-20 — canonical ARMA-GARCH temperature model
  • Benth, F., Saltyte Benth, J. (2007, 2013), Modeling and Pricing in Financial Markets for Weather Derivatives (World Scientific)
  • Jewson, S., Brix, A. (2005), Weather Derivative Valuation: The Meteorological, Statistical, Financial and Mathematical Foundations (Cambridge)
  • CME Group, Weather Futures and Options Product Specifications (current edition)
  • ICE, Energy Futures + Options Contract Specifications
  • Hersbach, H., et al. (2020), “The ERA5 Global Reanalysis,” Quarterly Journal of the Royal Meteorological Society 146(730), 1999-2049
  • ISDA, 2002 Master Agreement and Commodity Derivatives Definitions (current supplements)
  • WRMA (Weather Risk Management Association) annual surveys
  • FERC EPAct 2005 §1283 (16 USC §824v) and 18 CFR §1c.2 — Energy Market Manipulation Rule

Adjacent

Graph View