Bill Ackman Explained Why He Just Bought Microsoft. But There Are Even More Reasons to Buy This "Magnificent Seven" Stalwart

Bill Ackman disclosed that Pershing Square USA built a core position in Microsoft in Q1, selling most of its Alphabet stake to fund the purchase. The article argues Microsoft’s AI fears are overblown, citing its 21.8x forward P/E, OpenAI stake, Azure diversification, and new custom AI models/chips as reasons the stock could re-rate higher. The piece is broadly supportive of Microsoft, but it is opinion-driven rather than a new fundamental catalyst.

Analysis

The market is treating Microsoft like a monolithic SaaS risk, but the more important dynamic is that it is becoming a capital allocator across the AI stack: frontier model exposure, distribution, and custom inference hardware. That structure is qualitatively different from software peers whose margins are more vulnerable to model commoditization, because Microsoft can reprice the stack internally and shift economics from third-party model rent to owned infrastructure economics. The biggest second-order winner is likely Microsoft’s own balance sheet, not just its P&L: every point of AI traffic it can route onto proprietary silicon or in-house models is a direct offset to external vendor dependence and a long-duration margin defense.

The overlooked issue is competitive entropy at Alphabet and, to a lesser extent, any enterprise software company without a hardware or hyperscale substrate. If Microsoft can successfully bundle AI into the existing trust layer of M365 and security/compliance workflows, the disruption threat becomes less about seat loss and more about slower pricing expansion — a much lower-severity outcome than the market is discounting. Conversely, if Copilot adoption remains weak, the downside is not just product disappointment; it risks a slower Azure monetization path because the market has been implicitly underwriting AI-led cloud share gains that may now need a longer time horizon.

The contrarian read is that the selloff likely overreacted to a temporary narrative unwind. Microsoft does not need to win the model race outright to justify re-rating; it only needs to avoid being structurally disintermediated while monetizing AI through a hybrid pricing model and internalizing more compute economics over 12-24 months. The real catalyst path is execution evidence: better Copilot usage, tighter Azure AI attach, and proof that custom silicon meaningfully lowers cost per token, which would directly compress the bear case on gross margin erosion.

Near term, the stock can remain range-bound if investors keep rotating into names with cleaner AI beta, but over a 6-18 month horizon the risk/reward improves if AI demand broadens from training hype to enterprise inference. That favors Microsoft because inference economics reward scale, distribution, and low-cost compute more than pure model quality. The main tail risk is that the market reassigns Azure growth to lower-quality, lower-margin AI demand, but even that is manageable unless capital intensity spikes faster than operating leverage.

Key Decisions for Investors

• Add MSFT on weakness over the next 2-6 weeks; target a 12-month upside re-rating if the market accepts that AI is a margin-defense story, not a moat-collapse story. Risk/reward favors entry below ~22x forward earnings because downside is buffered by software recurring revenue and balance-sheet optionality.
• Pair long MSFT / short GOOGL for a 3-6 month relative-value trade. The thesis is that Microsoft’s enterprise distribution and vertical integration should translate AI into monetization faster than Google’s more search-adjacent AI defense; stop if Google demonstrates clearer enterprise attach or Microsoft Copilot metrics disappoint.
• Own NVDA only tactically, not as a core relative winner against MSFT. If Microsoft’s custom silicon and in-house models gain traction, it pressures the premium on third-party GPU demand at the margin; prefer buying NVDA on drawdowns when hyperscaler capex data is accelerating, not on narrative alone.
• Consider MSFT Jan-2027 call spreads to express a 12-24 month re-rating tied to Azure AI monetization and custom silicon savings. This offers defined downside versus outright stock in case near-term AI spend shifts to lower-margin inference before operating leverage shows up.
• Fade sustained GOOGL weakness only after evidence of AI product re-acceleration; until then, the market may continue to reward Microsoft for being the cleaner 'enterprise AI control plane' while penalizing model-dependent peers.