Key takeaways:
-
AI-led profits helped push US equities to their best month since late 2020, brushing aside the rise in oil prices, further bolstered by consumer confidence, resilient US growth, and contained inflation.
-
The rally was not confined to tech: industrials, utilities, and healthcare outperformed, a sign that markets are beginning to price the second-order effects of the AI buildout: power demand, grid investment, and physical infrastructure.
-
The key risk is not adoption, but concentration. Returns will cluster by capital structure, geography, and physical capacity, warranting a selective approach as the cycle extends.
A year ago, markets rewarded the vision. Last week, they rewarded the receipts.
Bottom line up top
AI remains the engine driving US tech, but the market has moved on from applauding ambition. It wants results, and last week’s earnings largely delivered. Strong AI-led profits helped push US equities to their best month since late 2020, brushing aside the rise in oil prices, further bolstered by consumer confidence, resilient US growth, and contained inflation. The question now is whether upcoming data can keep pace – with US payrolls this week the next test as the war carries on.
Stronger growth, but a higher bar
Demand held up strongly, with cloud growth reaccelerating sharply. The fastest-growing cloud unit surged over 60% YoY, with enterprise backlog nearly doubling in a single quarter to close in on half a trillion dollars. One management team described demand for AI compute as “unprecedented.” No wonder when AI inference demand has surged tenfold in a year.
Across the sector, large cap tech delivered double-digit revenue growth, with early monetization of generative AI beginning to show through. Earnings held firm too, with operating leverage doing its part in core businesses. The Mag-7 is now on track for roughly 46% YoY Q1 earnings growth on ~25% revenue, well ahead of what analysts expected entering the season. And powering Q1 expectations to a whopping ~27% for the broader S&P 500.
But the story of the week was capital intensity and the tension it is creating. The five largest hyperscalers are on track to deploy close to $700B in AI capex in 2026 alone, spanning data centers, chips, networking – a figure that continues to climb each earnings cycle. The spending is fueling growth, but it is also putting free cash flow, depreciation timelines and returns on invested capital under increasing scrutiny.
Focus on financing the AI buildout
The funding mix is already shifting. Private markets deployed over $100B into data center deals and operators in 2025, double the prior year, as banks, credit markets, and infrastructure investors step in to finance what balance sheets can no longer absorb alone. Year-to-date private markets have provided additional financing for ~$17.5B of data centers. As the AI buildout scales, funding is likely to be balanced across asset classes: from syndicated credit, led by IG, to private markets for infrastructure and real estate, with banks acting as a bridge to permanent financing.
The divergence in market reaction to earnings was also telling. Where companies could draw a clear line between AI spend and incremental revenue, particularly through cloud and enterprise channels, stocks were rewarded. Where that line blurred – even alongside strong top-line growth – pressure followed as the bar for monetization has risen. Notably, the rally was not confined to tech: industrials, utilities, and healthcare outperformed, a sign that markets are beginning to price the second-order effects of the AI buildout: power demand, grid investment, and physical infrastructure.
From spending to getting paid
The investment case is shifting – from who is spending to who gets paid across the capital stack – as outlined in our latest ‘Investing Across the AI Supercycle’ report:
- Equities: Stay selective with focus on companies that can convert AI spend into durable margin expansion. Broaden beyond Mag7 into industrials, utilities, and healthcare where AI spillovers may be underpriced. Europe, Japan and EM offer selective opportunities.
- Fixed Income: Utilities entering a major grid-upgrade cycle offer regulated, inflation-linked cash flows. Power generation, transmission, grid upgrades, and data center infrastructure require long-term financing, creating a structural bid for high-quality duration assets.
- Private credit: Allows investors to tailor exposure to the AI supply chain, translating thematic conviction into contracted, cash flow-driven returns through underwriting discipline, covenant protections and asset backed structures.
- Real assets/infrastructure: Long-dated contractual revenues, high barriers to entry, stable cash flows. Offer opportunity to participate in long-term growth as AI capacity expands power generation, grids, storage, and cooling.
The key risk is not adoption, but concentration. Returns will cluster by capital structure, geography, and physical capacity, warranting a selective approach as the cycle extends.
Bottom line
The broader message is clear: AI is a powerful growth driver, but the patience of investors has limits. The narrative is shifting from the size of investment to the efficiency of it, and to the durability and visibility of what comes next. In this phase, it isn’t the biggest spenders that will win, but the ones that can show the returns – with receipts.
