Big Growth. 'Undemanding' Prices.

Morgan Stanley reiterated bullish calls on the hyperscalers—Alphabet, Microsoft, Amazon, and Meta—arguing they still offer strong forward earnings growth at "undemanding" valuations. The Roundhill Magnificent Seven ETF is up 27% since late March, while the group trades at a 30% forward P/E premium to the S&P 500, down from more than 70% in 2023. Analysts see AI-driven revenue growth and accelerating cloud demand outweighing near-term concerns about the sector's more than $700 billion in planned infrastructure spending this year.

Analysis

The market is re-rating hyperscalers from "AI capex stories" into quasi-quality growth compounders: the key second-order effect is that faster revenue reacceleration is compressing the valuation penalty normally assigned to rising capex. That matters because the buy-side has been positioned for margin erosion; if cloud growth and AI monetization stay ahead of spend, the multiple can expand even with negative free-cash-flow optics in the near term. In other words, the bulls do not need capex to slow — they only need payback visibility to improve.

The clearest winner is not just the four named platforms, but the broader ecosystem of data-center enablers: networking, electrical gear, power management, and liquid cooling should see the most persistent order flow as hyperscalers lock in multi-quarter build schedules. The loser set is more subtle: software names with weaker AI differentiation may face relative multiple compression as budgets consolidate around the few platforms that can monetize inference, distribution, and developer ecosystems. If enterprise AI adoption accelerates, the bargaining power shifts toward the cloud landlords, not the application layer.

The main risk is timeline mismatch. Over days to weeks, a pullback can still happen if investors focus on FCF deterioration or if rates back up enough to reprice duration risk; over months, the bigger threat is a capex reset cycle if utilization disappoints. But the article’s setup suggests the tape is still rewarding visible top-line acceleration, so any air pocket is more likely to be bought unless there is evidence that AI demand is saturating faster than spending.

The contrarian miss is that "undemanding" is relative, not absolute: these stocks can still underperform the market if earnings revisions plateau, even without a true fundamental break. The best tell will be whether revenue beats translate into upward estimate revisions faster than capex estimates rise. If that spread narrows, the value argument fails quickly.

Key Decisions for Investors

• Stay long GOOG/MSFT/AMZN/META on a 1-3 month horizon, but finance it by trimming weaker-duration software exposure; the setup still favors names with the highest AI monetization leverage and the cleanest earnings revision momentum.
• Initiate a relative-value long XLC vs short IGV trade for 4-8 weeks: hyperscaler-led communication services should outperform application software if enterprise AI spend continues to concentrate at the platform layer.
• Add a basket long in data-center suppliers/infrastructure beneficiaries over the next 1-2 quarters; the trade has better earnings visibility than chasing the hyperscalers after a 27% bounce and benefits from the same capex cycle.
• Consider a call spread on MSFT or GOOG rather than outright stock for the next 2-3 months: upside remains open if revenue reacceleration persists, but the spread caps downside if the market re-focuses on capex and FCF compression.
• Use any 5-8% pullback in the group as an entry window rather than fading strength immediately; the near-term tape is likely to punish underexposure more than overexposure until there is evidence AI demand is slowing.