BlackRock, EQT Pay $33.4 Billion for AES in Bet on AI's Power Grid Problem

BlackRock and EQT's $33.4 billion acquisition of AES signals a structural shift: AI data center demand is straining the power grid, and private capital is moving to own the infrastructure.

BlackRock, EQT Pay $33.4 Billion for AES in Bet on AI's Power Grid Problem

AES Corp's $10.7 billion take-private by BlackRock's Global Infrastructure Partners and EQT is not a routine utility deal. It is institutional capital making a multi-decade bet on the most under-discussed constraint in the AI build-out: the power grid cannot keep up, and private markets are moving to own the infrastructure that public markets have been ignoring.

The numbers are substantial. The enterprise value reaches $33.4 billion when including approximately $22.7 billion of existing debt. The per-share cash price of $15.00 represents a 40.3 percent premium to pre-rumor trading levels from July 2025. The consortium committed 100 percent equity—no debt bridge—insulating the deal from credit market volatility. Yet AES shares dropped 17 percent on announcement, a 13 percent discount to the February 28 close. Both the premium and the drop are true, depending entirely on when investors entered.

The Power Grid Is AI's Bottleneck

Every major technology company is racing to build AI data centers. Every one of those data centers needs power. A lot of power. The US electrical grid—built for a pre-cloud, pre-AI demand profile—does not have enough of it.

AES Corp owns AES Ohio and AES Indiana, two state-regulated electric utilities serving the Midwest, plus a significant clean energy generation portfolio. AI data centers are being built at unprecedented scale, primarily in the US. Each hyperscaler—Microsoft, Google, Amazon, Meta—is committing tens of billions in capital expenditure annually. Every dollar of that spending requires a kilowatt-hour of electricity.

US power demand had been flat for 20 years. AI broke that trend. The Department of Energy projects 15 to 20 percent growth by 2030. Utilities that own regulated generation and transmission assets in high-demand corridors are suddenly among the economy's most strategically valuable infrastructure assets.

AES beat Q4 2025 earnings per share estimates by 80 percent—$2.09 actual versus $1.16 expected. That is evidence of structural demand acceleration the public market had not fully priced.

The Deal Timeline

Acquisition rumors first surfaced in July 2025, with AES shares trading at a premium even then. By the time AES reported Q4 earnings in early 2026, the fundamental confirmation of the AI power thesis was unmistakable. Advanced talks emerged in February 2026, with BlackRock's Global Infrastructure Partners leading, joined by EQT, CalPERS, and the Qatar Investment Authority. Sovereign wealth and pension participation signaled long-duration conviction.

A definitive agreement was signed March 1–2, 2026. The consortium agreed to a $321 million break fee, confirming deal conviction. Regulatory review began March 13, with FERC and DOJ review expected to take 9 to 18 months. The expected close is late 2026 or early 2027, following a shareholder vote. AES Ohio and Indiana will then become private regulated utilities.

Why the Consortium Paid a 40 Percent Premium

Public utilities face structural tension: dividends, quarterly guidance, and analyst coverage. AES Board Chair Jay Morse noted the deal provides "enhanced financial flexibility" without "dividend cuts or heavy equity raises." Private ownership eliminates quarterly pressure entirely.

The demand driver is AI data center construction. Each 100 megawatts of data center capacity requires enough power for approximately 80,000 homes. EV adoption will add 10 to 20 percent to grid load between 2026 and 2030. Industrial reshoring is creating manufacturing power demand on a 3- to 7-year build timeline. And AI inference loads are baseload, not peak demand, meaning they run continuously.

The Midwest, where AES Ohio and Indiana operate, is an active data center corridor. It offers available land, lower real estate costs, and existing power infrastructure. That combination is increasingly rare.

Why AES Fell 17 Percent on Announcement

The 17 percent drop on announcement is not a rejection of the deal. It is arithmetic. A 40 percent premium to pre-rumor levels and a 17 percent drop can both be true, depending on entry price.

Pre-rumor holders who bought near $10.70 received $15.00, a 40 percent gain. Post-rumor buyers who entered between $14 and $16 broke even or took small losses. Late speculators who bought above $17 lost 12 to 13 percent. Deal arbitrage funds that bought near $14.60 captured a 2.7 percent spread over 9 to 12 months, a modest return for regulatory risk. Many of those funds exited, driving the drop.

The Consortium

BlackRock's Global Infrastructure Partners leads as the infrastructure specialist with a 10- to 15-year horizon. EQT AB joins as a European private equity and infrastructure co-lead, attracted by the stability of US regulated utility frameworks. CalPERS represents US pension capital with 30-year liabilities that match utility asset duration. The Qatar Investment Authority brings sovereign wealth seeking exposure to digital infrastructure globally.

The deal uses 100 percent committed equity. There is no debt bridge. Deal economics are not sensitive to credit markets.

Post-Close Operations and Regulatory Path

PUCO in Ohio and IURC in Indiana must approve the ownership change. Customers will see no operational difference. Local branding and rate-setting continue. The private structure enables reliable capital expenditure for grid upgrades without quarterly earnings pressure.

FERC review typically takes 4 to 8 months, with approval odds above 90 percent for infrastructure deals meeting standard conditions. DOJ antitrust review runs 3 to 6 months and is considered low risk. Precedent utility buyouts have been routinely cleared.

Why This Utility Drew a 40 Percent Premium

AES traded at 12.5 times EV/EBITDA before the deal, a discount to the XLU utility ETF at 14.2 times, NextEra at 18.1 times, and Southern Company at 13.8 times. Only 65 percent of AES revenue came from regulated utilities, compared to 85 percent or more for peers. That mix gave the consortium room to value the regulated assets at a premium while the market focused on the unregulated pieces.

The data center corridor exposure was direct. The Midwest is active. The valuation discount relative to peers was visible. And the 80 percent EPS beat in Q4 confirmed the thesis.

Which regulated utility in a data center corridor trades at a similar discount today? That question may determine where institutional capital moves next.

What This Deal Signals

This transaction is not an isolated event. It is a template. Institutional capital is identifying structural constraints in the AI build-out—constraints public markets are not yet fully pricing—and moving to own the solutions.

The power grid is the most capital-intensive, least-flexible layer of the AI stack. It cannot be software-updated or cloud-scaled. It must be built, upgraded, and maintained over decades. That timeline matches infrastructure capital's natural holding period.

For investors, the question is not whether more such deals will follow. The question is which utilities in data center corridors are trading at similar discounts to their strategic value. The AES transaction provides a blueprint for identifying them.

DISCLAIMER:

This market narrative is published for educational and informational purposes only. It does not constitute financial advice. All references to market conditions reflect conditions as of the timestamps noted. Trading involves substantial risk of loss. Always conduct independent research.