Microsoft Is a Mess. Is the "Magnificent Seven" Stock a Buy in May or Better Avoided?

Microsoft delivered strong fiscal Q3 results, with revenue up 18% year over year, operating income up 20%, and AI revenue reaching a $37 billion annual run rate, but the stock remains down 15.7% year to date. The company is guiding to more than $40 billion of capex in fiscal Q4 and about $190 billion for calendar 2026, pressuring free cash flow, while commercial bookings fell 46% on lower OpenAI commitments. The revised OpenAI arrangement adds flexibility and Copilot adoption is accelerating, but investor concern over heavy AI spending and OpenAI dependence is weighing on sentiment.

Analysis

MSFT’s issue is not demand destruction; it’s capital intensity re-rating. The market is moving from valuing a software compounder on incremental gross-margin expansion to valuing a hyperscaler on returns on incremental capital, and that transition typically compresses multiples before the operating leverage shows up. In that regime, even strong revenue growth can underperform if the next 12 months of spend is perceived as front-loaded and uncertain.

The OpenAI reset matters less as a revenue cliff than as a competitive optionality event. Losing exclusivity to the best-known model partner reduces MSFT’s ability to force demand through a single ecosystem, while simultaneously creating a chance for Microsoft to monetize Copilot across a broader model stack if it executes well. That is a subtle but important shift: the bull case is now platform distribution and enterprise workflow capture, not model access scarcity.

Second-order beneficiaries are the vendors selling the picks-and-shovels around GPU deployment, networking, power, and data-center buildout, because the spending race is pushing urgency into the supply chain regardless of who “wins” the model layer. The quieter risk is that custom silicon leaders with lower unit economics eventually widen the cost gap, leaving MSFT with heavy depreciation before enough workload migration occurs. That could keep free cash flow growth lagging earnings for several quarters, which is exactly the setup that tends to cap multiple expansion.

Contrarian take: the selloff may be overdone if investors are treating the capex spike as purely discretionary. If AI demand is still supply-constrained, then near-term spend is more like inventory build than waste, and the first sign of a better return cycle will be accelerating paid-seat conversion in Copilot rather than headline Azure growth. The inflection point to watch is the next 1-2 quarters: if bookings stabilize while Copilot monetization keeps compounding, the market can re-underwrite MSFT as a durable AI distribution layer instead of a capex story.

Key Decisions for Investors

• Stay tactically underweight MSFT for the next 1-2 quarters; the risk/reward still favors waiting for capex normalization or evidence that Copilot monetization is outpacing depreciation drag.
• Pair long AMZN / short MSFT over 3-6 months: AWS has a cleaner narrative on external AI monetization and custom-silicon economics, while MSFT faces a tougher proof burden on return on invested capital.
• Own GOOGL as the preferred large-cap AI platform exposure for 6-12 months; its model stack, chips, and distribution are converging into a more self-reinforcing cost advantage than MSFT’s partner-dependent setup.
• For traders willing to express a mean-reversion view, buy MSFT 12-month call spreads on any post-earnings pullback of 5-8%: the upside is a multiple reset if Copilot seats and commercial bookings stabilize, while premium is defined.
• Watch networking and power-infrastructure beneficiaries for a lagged long: MSFT’s spending cadence should continue to support names levered to data-center bottlenecks over the next 2-4 quarters.