Down 35% on Artificial Intelligence (AI) Panic: Why Smart Investors Are Loading Up on Microsoft Stock

Microsoft is trading about 35% below its all-time high and is described as the cheapest it has been on an earnings basis in a decade, with diluted EPS up 23% in the latest quarter. The article highlights strong AI momentum, including Copilot at a $37 billion annual run rate growing 123% and Azure growth of 40% in fiscal Q3 2026. It also notes Bill Ackman’s reported $2 billion Microsoft position, framing the stock as a potential long-term buy despite broader market skepticism.

Analysis

The setup is less about Microsoft being “cheap” in isolation and more about the market questioning the durability of AI monetization. When a mega-cap with multiple recurring-revenue engines rerates down while earnings still compound, it usually signals either multiple compression from higher-duration discount rates or skepticism that AI attach rates will stay as high once customers optimize spend. The second-order risk is that Microsoft’s own AI success becomes its own headwind: as Copilot adoption broadens, customers will pressure bundle economics, and Azure AI growth can decelerate faster than headline cloud growth once workloads move from experimentation to cost control.

The more interesting read-through is to the AI ecosystem. If Azure remains the preferred inference/training substrate for frontier models, MSFT’s weakness can create a buying opportunity in adjacent infrastructure providers that benefit from sustained capex regardless of Microsoft’s multiple. But if the market is repricing MSFT on worries that AI ROI is lagging hype, that is a warning signal for NVIDIA and high-multiple software names that rely on a similar “AI monetization later” narrative. In that case, the next leg is usually not a collapse in fundamentals but a slower-than-expected conversion of AI usage into operating leverage.

The contrarian view is that the selloff may be overdone because this is exactly the type of name institutions rotate into when they want AI exposure with balance-sheet quality and earnings visibility. The risk window matters: over the next 1-3 months, sentiment can remain weak if hyperscaler capex scrutiny rises; over 6-12 months, sustained 20%+ EPS growth with reacceleration in Azure should force a multiple repair. The key catalyst is not another product announcement but evidence that AI is lifting per-seat monetization and cloud consumption without triggering margin dilution.

Key Decisions for Investors

• Go long MSFT on further weakness, staged over 2-4 weeks; use the name as a high-quality AI compounder with downside buffered by recurring cash flows. Target a 12-month multiple reversion scenario; risk is a prolonged de-rating if Azure AI growth slows into the low-30s.
• Pair long MSFT / short a basket of higher-duration AI software names with weak profitability and no clear monetization path. The thesis is that capital rotates toward proven cash generators if AI spend becomes more discriminating.
• Buy NVDA only on confirmation that Azure and other hyperscalers are sustaining capex, not just launching pilots; otherwise the trade is vulnerable to a sentiment air pocket over the next quarter.
• Consider MSFT call spreads 6-9 months out to express a re-rating view with defined risk; structure for a rebound in the stock to prior valuation bands while limiting carry cost if the market stays skeptical.
• Trim exposure if Copilot or Azure AI growth decelerates materially for one quarter; that would suggest the market is discounting a real monetization issue rather than a temporary valuation reset.