S06.12 · Energy, Power & Climate
Marketing, trading and price-risk management for power, gas and environmental commodities, where algorithmic trading is already standard practice.
Energy trading, retail supply and risk covers the marketing, trading and price-risk management of power, gas and environmental commodities. No single figure spans trading and retail together: global electricity retail sales revenue alone is estimated at over $2.5T in 2025 (IEA-derived), while wholesale power, gas and environmental-commodity trading notional volumes are far larger but not comparably reported. Retail-supply growth tracks electrification and load growth at mid-single digits; trading volumes grow faster wherever renewables and storage additions force more active balancing. Unusually for this sector, AI here has moved furthest inside an existing function rather than reshaping demand for it — algorithmic trading is already standard practice at major energy trading desks.
Deregulated retail markets — Texas, the UK, parts of the EU, Japan and Australia — support competitive retailers and independent trading desks; regulated markets confine supply to incumbent utilities. Concentration differs by activity. Trading concentrates among major energy companies and specialist traders such as Vitol, Trafigura, Shell Energy and utility trading desks, while retail supply is fragmented across hundreds of licensed retailers active in liberalized markets.
The segment sits between generation and production on one side and end-use utilities and consumers on the other, with environmental-commodity trading linking directly to broader carbon markets. Trading margins are volatile and depend on volume and volatility; retail-supply margins are thin and constrained by regulated price caps in many markets. Capital intensity is low — the business runs on working capital and collateral rather than fixed assets — but regulatory licensing is a hard gate on entry. The competitive basis follows from that balance-sheet shape: desks compete on balance-sheet strength and data access rather than on physical infrastructure, which sets this segment apart from the rest of the sector.
The expansionary adjacency is environmental-commodity trading — carbon, renewable energy certificates — which monetizes the energy transition as a tradeable commodity flow, alongside demand-side management and distributed-energy aggregation.
Agentic AI automates price forecasting, portfolio optimization and algorithmic trade execution, collapsing the trader and analyst headcount line, and the moat shifts from human trading intuition toward proprietary forecasting models and data access. Algorithmic trading, as noted, is already standard practice at major energy trading desks. The margin effect is asymmetric rather than rising-tide: desks with proprietary data feeds and models capture a larger share of a more volatile, renewables-heavy market, while desks reliant on generic forecasting tools are more likely to watch their share compress than to find the market easier for everyone.