Can AES Capitalize on Surging Data Center Energy Demand?

AES has 4.2 GW of data-center PPAs in operation and 8.2 GW signed, and in Feb 2026 signed long-term PPAs to supply power to Google’s Wilbarger County data center while developing, owning and operating co-located energy assets. Zacks consensus shows modest EPS estimate increases of +2.56% for 2026 and +1.98% for 2027; AES trades at a discounted forward P/E of 5.84x vs. the industry 16.11x and its shares are up 6.9% over six months (industry +9.2%), supporting a constructive but measured outlook.

Analysis

Data-center electrification is creating a bifurcation in the power value chain: owners of contracted, dispatchable capacity plus control systems will capture outsized economics versus pure merchant renewables. The structural premium comes from 24/7 load profiles that require either firming (gas, storage, hydrogen-ready assets) or guaranteed dispatch through behind-the-meter co‑located generation and energy-management services — this shifts margin pools from commodity energy sales to platform-like services (SaaS + O&M) with higher gross margins and stickier cash flows.

Capital markets will re-price firms differently depending on capital intensity and balance-sheet footprints: companies that can scale co-located assets and monetize ancillary services will justify higher multiples, while regulated utilities or asset-light developers may lag. Key operational risks that will compress value are interconnection and permitting delays, capital-cost inflation for firming capacity, and concentration risk if large tech off-takers renegotiate credit terms; these manifest over 6–36 months and can flip a growth narrative to an earnings shortfall.

Second-order winners include industrial HVAC/cooling suppliers, transmission and substation contractors, battery OEMs and software vendors that optimize hourly dispatch and demand response; conversely, pure wind/solar developers without firming optionality face margin pressure and basis risk. From a macro lens, accelerated data-center demand raises local capacity prices and congestion, creating arbitrage opportunities for merchant peakers and distributed battery fleets to capture high locational marginal values during peak AI training runs.

Key Decisions for Investors

• Long AES equity (12–24 months): overweight AES to capture re‑rating if the firm converts signed PPAs into operating cash flow and demonstrates margin expansion from energy-management services. Target +30% upside in 12–24 months; downside -35% if capex overruns, interconnection delays, or rate hikes impair free cash flow. Use 50% size initial position and add on pullbacks to key support levels.
• Pair trade — Long AES / Short XEL (9–18 months): express preference for asset owners that can attach firming assets and software margins (AES) vs regulated distribution/exposure (XEL) which will trade more like a utility. Expect pair to widen 10–20% if contract-to-asset conversion accelerates; risk is regulatory support for utility-scale offerings that narrows the spread.
• Long GEV (6–12 months) or buy-call spreads on GEV: play equipment backlog conversion as data-centers force add-on gas/firming builds. Reward is direct backlog-to-revenue recognition; tail risk is order cancellations if customers pivot to alternative firming tech.
• Buy AES long-dated call spread (18–24 months) instead of outright equity to cap downside: purchase near-the-money 18–24 month calls financed by selling higher strikes. This provides convex upside to a re-rating while limiting cash at risk to capex/ refinancing shocks.
• Hedge concentration: if owning NVDA exposure as AI proxy, protect 3–6 month downside with modest put protection (5–7% of position cost) — a meaningful tech drawdown would delay data-center builds and hurt near-term merchant power prices.