Centrica H1 2026 slides: transformation costs weigh on earnings

Centrica reported H1 2026 adjusted EBITDA of £737m, down 18% from £900m, and adjusted EPS of 6.8p vs 7.0p prior-year—both missing analyst expectations. Revenue fell to ~£10.5bn vs a £12.4bn forecast, while transformation spending drove free cash flow to a £570m outflow (vs a £200m inflow) and net cash down to £709m from £1.5bn. Despite a 9% dividend increase to 2.0p and progress on nuclear/contracted assets (e.g., Sizewell B CFD at £70.50/MWh), the shares dropped 10.18% premarket toward the 52-week low as 2026 capex rose to ~£1.1bn from ~£700m-plus and near-term segment outlooks were cautious.

Analysis

This is a classic “good strategy, bad quarter” setup: the market is punishing the equity for front-loaded capex and negative free cash flow before it is willing to pay for the cleaner earnings mix. The real mechanism is multiple risk — if management keeps converting merchant exposure into contracted/infrastructure earnings, the stock can re-rate toward a utility/infra multiple, but only once investors see cash conversion stabilize. Near term, the selloff likely reflects the loss of buyback capacity and the prospect that the balance sheet will be asked to fund growth longer than expected.

The second-order winners are not obvious on the surface. Nuclear life extensions and incremental baseload spending are a structural tailwind for the uranium/nuclear supply chain, so UUUU is a cleaner thematic beneficiary than trying to trade the utility itself. By contrast, the biggest hidden risk is consumer credit deterioration: elevated bad debt is a margin issue today and a demand issue tomorrow if affordability stays stretched, which would pressure retail economics even if wholesale conditions improve.

Time horizon matters. Over the next 1-3 months, the stock is vulnerable to any miss on cash flow, cost savings, or project execution, and the market is unlikely to pay for 2030 targets until it sees H2 normalization. Over 6-18 months, the contrarian view is that the market may be underpricing the shift toward contracted earnings and the option value of scarce power assets, but that only works if management avoids construction slippage and political intervention on storage/social tariff issues. GOOGL/TSLA look like noise here; this is more a power-infrastructure and capital-allocation story than an AI/EV equity trade.

Key Decisions for Investors

• Short/underweight CPYYY on any 1-2 week bounce; the setup favors further de-rating until free cash flow turns positive and management proves that capex is temporary, not structural. Falsify if H2 cash conversion improves materially or buybacks are explicitly reintroduced.
• Buy UUUU as a 3-6 month thematic long on nuclear durability and fuel-cycle sentiment; use it as a higher-beta proxy for the growing baseload/extension narrative. Stop if uranium spot weakens sharply or if UK nuclear policy loses momentum.
• Pair trade: long UUUU / short CPYYY for a relative-value expression of the same power-demand theme, with the long leg tied to nuclear supply-chain optionality and the short leg tied to capital intensity and retail credit risk. Keep it small; this is sentiment-sensitive, not a high-conviction fundamental spread.
• Do not chase GOOGL or TSLA off this print; the article does not create a direct earnings revision for either. Revisit only if power-cost headlines begin to hit data-center economics or EV affordability in a measurable way.