Oil Hit $110. AI Bought the Power Grid.

GOOD MORNING, Monday showed both sides of the AI boom at once. Brent traded above $110 and the 10-year Treasury yield touched 4.63%, its highest since February 2025, as stalled Iran talks revived fears that an energy shock could keep inflation and borrowing costs high. The Nasdaq fell about 0.5%. Yet capital kept pouring into the infrastructure underneath AI: NextEra agreed to buy Dominion Energy for $66.8 billion, largely to gain exposure to the power demand around Virginia’s data-center corridor, while Google and Blackstone launched a new AI cloud venture. Higher rates are squeezing AI valuations. AI itself is still reshaping the physical economy.

MARKETS | TLDR

Tech lagged as yields climbed: The Dow gained roughly 0.3%, the S&P 500 finished little changed and the Nasdaq fell about 0.5% as investors took profits in technology stocks.
The 10-year Treasury yield hit 4.63%: The benchmark yield reached its highest level since February 2025 as high oil prices reinforced concerns that inflation and borrowing costs could stay elevated.
Brent traded above $110: The Strait of Hormuz remained largely closed and global inventories were being depleted, keeping the energy shock at the center of the rates outlook.

POWER

AI Just Put a $66.8 Billion Price Tag on Electricity

NextEra Energy agreed to acquire Dominion Energy in a $66.8 billion all-stock transaction, creating one of the world’s largest utility companies. The strategic prize is Dominion’s position in PJM and Northern Virginia, home to the world’s densest concentration of data centers.

The deal matters because AI infrastructure is becoming a power-market story as much as a semiconductor story. Dominion has nearly 51 gigawatts of contracted data-center capacity, with customers including Google, Amazon, Microsoft and Meta. The combined company would be better positioned to finance and build the generation, transmission and grid infrastructure required to serve a projected wave of new data-center electricity demand.

That makes power one of the clearest second-order beneficiaries of AI capex. Nvidia sells the accelerator, but the server still needs a data center, cooling system and reliable electricity for years. The risk is financing: the same AI boom increasing power demand is arriving while long-term Treasury yields are near multi-year highs, raising the cost of building the infrastructure needed to satisfy it.

RATES

Oil Is Raising the Discount Rate on the AI Rally

Brent crude moved back above $111 Monday as efforts to end the Iran conflict appeared to stall and the Strait of Hormuz remained largely closed. Reuters noted that only a fraction of normal shipping was getting through and cited estimates that as much as 1 billion barrels of crude supply could be lost by the end of May if disruption persisted.

The bond market is where that energy shock becomes an equity problem. The U.S. 10-year Treasury yield touched 4.631%, its highest since February 2025, while the 30-year reached 5.159%. Higher energy prices raise inflation expectations; higher inflation keeps monetary policy restrictive; and higher long-term yields reduce the present value of future earnings.

That transmission matters most for the market’s expensive growth leaders. AI earnings remain strong, but Reuters noted that just 20 stocks accounted for almost all of the recent upside surprise in S&P 500 earnings, while estimates excluding AI and energy were flat for 2027. With leadership this concentrated, Nvidia’s results this week have to validate both the earnings story and the valuation built on top of it.

HEADLINES

Google and Blackstone launched an AI cloud venture: Blackstone will invest an initial $5 billion to bring 500 megawatts of data-center capacity online in 2027 using Google TPUs, another sign that private capital is moving deeper into the AI infrastructure stack.
The Google-Blackstone venture could reach $25 billion: Including leverage, the project could become a major new source of compute-as-a-service capacity and another competitor for independent AI clouds.
Storage stocks bucked the broader tech weakness: SanDisk jumped 11.9%, Seagate gained 6.3% and Western Digital rose 5.9%, showing that AI-linked hardware demand was still producing pockets of strength.
China’s April data showed signs of oil-driven demand pressure: Retail sales rose only 0.2% versus 2.0% expected, while industrial output also disappointed, offering an early glimpse of how sustained energy costs can damage growth.
Nvidia entered earnings week with expectations already elevated: Wall Street expected revenue around $78.5 billion, roughly 80% higher year over year, leaving the AI bellwether with an unusually high bar to clear.

UPCOMING

Home Depot earnings — May 19: The retailer will provide a read on housing-related demand and whether high borrowing costs are feeding into big-ticket consumer spending.
Nvidia earnings — May 20: Nvidia is the week’s central AI catalyst, with investors focused on data-center growth, margins, hyperscaler spending and whether guidance can justify the sector’s valuation.
FOMC minutes — May 20: The April meeting record will show how policymakers were thinking about inflation before the latest jump in oil and long-term yields.
Walmart earnings — May 21: Walmart will offer a broad read on household demand and the extent to which higher fuel and goods prices are changing consumer behavior.

DEEP INSIGTHS

Bonds Get a Taste of Oil’s Demand Destruction

Read this for the full macro transmission chain. Hormuz disruption is no longer just an oil-market problem: inventories are being depleted, yields are rising and high energy costs are beginning to show up in weaker activity. The key question is when supply destruction becomes demand destruction — and whether the Fed can respond while inflation remains elevated.

NextEra to Buy Dominion for $66.8 Billion as AI Power Demand Booms

This is the clearest example of AI expanding beyond semiconductors into regulated infrastructure. Dominion’s exposure to Virginia’s data-center corridor makes electricity capacity itself strategically valuable. The AI capex cycle is increasingly becoming a grid, generation and financing cycle — with utilities now competing for the same growth narrative that first re-rated chipmakers.