Opinion: Microsoft Probably Should Have Rallied After Those Solid Earnings

Microsoft reported Q3 EPS of $4.27 versus $4.07 expected on revenue of $82.886 billion, up 18% year over year, while operating cash flow rose 26% to $46.679 billion and Azure growth accelerated to 40%. The key concern was capital spending of $30.876 billion, up 84% year over year, which has kept the stock down nearly 6% after the earnings call despite a fourth straight beat. Commercial remaining performance obligations reached $627 billion and AI annual run-rate revenue surpassed $37 billion, underscoring strong long-term demand even as investors focus on capex intensity.

Analysis

The market is not discounting MSFT’s demand problem; it is discounting duration risk. When capex steps up this aggressively, the question shifts from “is AI monetizable?” to “how many quarters of margin drag can the stock tolerate before operating leverage reappears?” That makes the stock vulnerable to a valuation reset even while fundamentals remain excellent, because the next two prints are likely to be judged on free cash flow conversion rather than revenue growth.

Second-order, this spending wave is a competitive moat for Azure and a tax on smaller cloud/AI vendors. The beneficiaries are upstream infrastructure suppliers and the ecosystem that sells picks-and-shovels into hyperscaler buildouts, while mid-tier cloud players face a tougher funding environment as customers benchmark against Microsoft’s scale and pricing power. If capacity is the binding constraint, the next leg higher in MSFT should come only when utilization catches up; until then, the capex line will keep capping multiple expansion.

The contrarian takeaway is that the post-earnings drawdown may already be doing some of the work the bears want. A high-quality compounder de-rating on capex fear can create a better entry than a straight chase after a beat, especially if the market is implicitly asking for proof over the next 60-90 days rather than next year. The risk is not that AI demand disappears; it is that investors decide the payback period is longer than they’re willing to underwrite, which can keep the stock range-bound even if Azure remains strong.

Catalyst-wise, the setup improves if management can show sequential moderation in capex growth while preserving Azure acceleration, because that would flip the narrative from ‘spend now, worry later’ to ‘spend is already converting.’ Conversely, if capex stays elevated into the next quarter without a clear operating margin payoff, the stock likely underperforms despite continued fundamental beats. The window for this debate is weeks, not years: the next earnings call will probably determine whether the recent selloff is a buying opportunity or the start of a longer multiple compression phase.

Key Decisions for Investors

• Sell near-dated MSFT upside against stock exposure: buy 1-2 month put spreads or collar existing longs to monetize the post-earnings volatility reset; best risk/reward if shares fail to reclaim the $400 area within 2-4 weeks.
• Go long MSFT on weakness only via staged entries, not all at once: accumulate in tranches over the next 30-60 days if the stock remains de-rated, targeting a 12-month re-rating once capex growth decelerates and FCF conversion normalizes.
• Pair trade: long MSFT / short a basket of lower-quality AI infrastructure beneficiaries that trade on narrative rather than earnings visibility; the idea is that hyperscaler spend ultimately consolidates share toward the strongest platform, but the basket will be more vulnerable if AI funding tightens.
• Favor upstream capex beneficiaries for 1-2 quarters over software pure plays: look to semicap and datacenter infrastructure names as the nearer-term winners from Microsoft’s buildout, with the trade thesis fading once the market starts demanding ROI proof instead of just deployment.
• If you need equity exposure to AI with less balance-sheet sensitivity, rotate part of MSFT risk into the strongest cash-generative megacap peers on any relative weakness; MSFT is now the cleanest ‘show me’ name in the group, so relative performance may lag until the market sees margin inflection.