When Peter Thiel’s fund, Thiel Macro LLC, sold its entire Nvidia position in Q3 2025 — over 537,000 shares representing nearly 40% of the fund’s portfolio — the headline wrote itself: “Billionaire exits AI’s crown jewel.”

But the more important story wasn’t the exit. It was the pivot.

Thiel didn’t move to cash. He moved to software. Alongside the Nvidia liquidation, Thiel Macro initiated new positions in Microsoft and Apple — trimming the fund’s total U.S. equity exposure by more than half while rotating toward what he and a growing number of institutional managers appear to believe is the next phase of the AI trade: application-layer monetization over infrastructure speculation.

Thiel wasn’t alone. SoftBank sold its entire Nvidia stake — worth approximately $5.8 billion — in the same quarter. Goldman Sachs data later showed that hedge funds collectively sold stocks at the fastest rate in 13 years, with Nvidia among the top names shed. Meanwhile, Goldman’s own research showed funds rotating toward names with more predictable earnings profiles and lower valuation risk.

At the center of where that money appears to be going: Microsoft (MSFT).

Market Temperature: The Shift from “Picks and Shovels” to “Who Uses the Shovels”

The first chapter of the AI trade was hardware. Build the infrastructure. Sell the chips. Nvidia was the obvious beneficiary, and it delivered — shares surged more than 1,180% over a three-year span through the end of 2025.

But markets don’t pay the same multiple forever. As Nvidia’s valuation stretched toward 50x earnings, the question institutional managers began asking was no longer “Will AI grow?” It became: “Who is actually converting AI spending into durable, recurring revenue — and at what price?”

That question points directly to the software layer. Specifically, to a company already generating over $300 billion in annual revenue, growing operating income at 21% year-over-year, and sitting at the intersection of every major enterprise AI workflow on the planet.

Thiel’s strategy, described by analysts as prioritizing AI users over AI builders, maps almost perfectly onto what the institutional data is now confirming. In Q4 2025, Microsoft ranked as the second most widely held AI stock by hedge fund conviction — held by hundreds of funds, with Ken Griffin’s Citadel boosting its Microsoft position by over 1,600% in a single quarter. Israel Englander’s Millennium Management added shares. Tiger Global’s Chase Coleman added shares. The institutional convergence is hard to ignore.

Company Introduction: What Microsoft Actually Is in 2026

Microsoft is not a software company that dabbles in AI. It is, increasingly, an AI company that still sells software.

The distinction matters. Microsoft’s AI integration runs across three massive business units: Azure (cloud infrastructure), Microsoft 365 (productivity software), and Dynamics (enterprise applications). Each of these products now has an AI layer — Copilot — embedded directly into the workflow of more than a billion users globally.

That distribution moat is what separates Microsoft from nearly every other AI play on the market. It doesn’t need to convince enterprises to adopt AI. Its customers are already inside its ecosystem — and AI features are being layered on top of subscription relationships that renew annually, generating predictable, high-margin recurring revenue.

As CEO Satya Nadella stated in the most recent earnings call: “We are only at the beginning phases of AI diffusion and already Microsoft has built an AI business that is larger than some of our biggest franchises.”

Data-Driven Deep Dive

Revenue: Consistent, Accelerating, and Large

• Q1 FY2026 (Sept 2025): $77.7 billion — up 18% year-over-year
• Q2 FY2026 (Dec 2025): $81.3 billion — up 17% year-over-year
• Trailing twelve months (as of Dec 2025): $305.5 billion — up nearly 17% year-over-year
• Full fiscal year 2025 (ended June 2025): $281.7 billion — up approximately 15% year-over-year

For a company generating over $300 billion in annual revenue, sustaining 17–18% top-line growth is a rare accomplishment. Most companies at that scale are fighting for 3–5% annually. The AI tailwind is doing real work here — and it’s compounding.

Profitability: Operating Leverage Is Real

• Q2 FY2026 operating income: $38.3 billion — up 21% year-over-year
• Q1 FY2026 operating income: $38.0 billion — up 24% year-over-year
• Net income (trailing twelve months, Dec 2025): $119.3 billion — up approximately 29% year-over-year
• Q2 FY2026 non-GAAP EPS: $4.14 — up 24% year-over-year

Operating income growing faster than revenue is the signal that matters. It means the business is generating more profit per dollar of revenue as it scales — classic operating leverage. That’s not easy to manufacture at $300+ billion in sales. It’s a structural advantage, not a one-quarter accounting quirk.

Azure: The Engine Inside the Engine

• Azure revenue growth (FY2025 full year): Up 34% to more than $75 billion
• Azure growth (Q1 FY2026): 40% year-over-year
• Azure has delivered 30%+ growth for 10 consecutive quarters
• Microsoft’s committed 2026 capex: approximately $94 billion — entirely directed at cloud and AI infrastructure expansion

Azure is not just growing — it’s accelerating. Ten straight quarters of 30%+ growth is an almost unheard-of streak for a cloud business at this scale. And with OpenAI committed to spending $250 billion of its computing budget through Azure, plus Anthropic’s models now available on Copilot Studio, the demand pipeline extends well beyond 2026.

Strategic Positioning: The OpenAI Bet

Microsoft holds approximately a 27% stake in OpenAI and privileged access to its intellectual property and models through 2032. That’s not just a financial investment — it’s a structural moat. Every major enterprise that wants to build on GPT-class models while staying inside a compliant, enterprise-grade cloud environment effectively routes through Azure.

This is the key insight that funds appear to be pricing in: Microsoft doesn’t need to win the AI chip wars. It needs to be the environment where AI gets deployed — and it already is.

The Valuation Case: A Rare Combination

Here is where the opportunity becomes genuinely interesting from a risk-reward standpoint.

After a meaningful pullback in early 2026, Microsoft’s forward price-to-earnings ratio dropped to approximately 30x — a level not seen in years for this business. For context:

• Nvidia’s forward P/E at peak concentration: approximately 50x
• Microsoft’s forward P/E after the 2026 pullback: approximately 30x
• 92% of Wall Street analysts rate Microsoft a buy or strong buy
• Median analyst price target implies approximately 55% upside from early 2026 lows

The business growing at 17–18% annually, with operating income compounding at 21–24%, is now available at a materially lower multiple than the AI infrastructure names it is effectively replacing in institutional portfolios. That asymmetry — strong fundamentals, compressed valuation — is precisely what large funds look for when rotating out of concentrated, high-multiple positions.

Three Scenario Ranges Worth Modeling

Scenario 1 – Azure Reaccelerates, Copilot Monetizes Broadly

• Revenue growth: 18%–22% annually over the next 2–3 years
• Operating margin direction: Expanding, as AI features carry higher attach rates
• Valuation: Multiple re-rates toward 35–40x as AI monetization becomes consensus
• Investor result: Earnings growth + multiple expansion = compounding outperformance

Scenario 2 – Steady Growth, Multiple Holds

• Revenue growth: 14%–17% annually
• Operating margin direction: Stable or slight expansion
• Valuation: Forward P/E stays in the 28–32x range
• Investor result: Stock performance roughly tracks earnings growth — a solid, lower-drama outcome

Scenario 3 – Capex Concerns Weigh on Sentiment

• Revenue growth: 10%–13% (Azure growth decelerates as competition intensifies)
• Operating margin direction: Compressed by $94 billion annual capex cycle
• Valuation: Multiple compresses toward 22–25x
• Investor result: Stock could be flat to modestly negative — but business remains highly profitable and strategically dominant

Notice the risk profile in Scenario 3 versus the equivalent bear case for a 50x multiple infrastructure name. At 30x forward earnings with $119 billion in trailing net income, Microsoft’s downside is structurally cushioned in a way that hyper-concentrated, high-multiple AI plays are not.

The Risks: What Could Break the Thesis

No editorial is complete without an honest accounting of what could go wrong. For Microsoft, the real risks are not existential — they’re margin and timing risks.

• Capex overhang: Microsoft is committing approximately $94 billion in capital expenditures for 2026. If AI adoption monetizes slower than expected, free cash flow could come under near-term pressure.
• Azure growth deceleration: The company noted one quarter of softer Azure growth due to internal infrastructure allocation rather than customer demand weakness. If that distinction blurs, sentiment could reset.
• OpenAI dependency risk: A 27% stake in a single AI partner introduces concentration risk. If OpenAI pursues its IPO and begins competing more directly with Azure offerings, the relationship dynamics could shift.
• Competition: Amazon Web Services and Google Cloud are not standing still. Azure’s 10-quarter streak of 30%+ growth will eventually face harder comparables and a more competitive pricing environment.
• Multiple expansion is not guaranteed: Even at 30x forward earnings, if the market re-rates the entire software sector lower, Microsoft is not immune to broader compression.

The Big Picture: Why This Rotation Makes Structural Sense

The AI infrastructure buildout — the data centers, the chips, the power — is a capital-intensive, cyclical business. Hyperscalers are collectively guiding toward hundreds of billions in data center spending through 2026 and beyond. That spending benefits Nvidia. But it also creates an enormous installed base of compute capacity that needs software to deploy, manage, and monetize.

Microsoft sits at that intersection. Every enterprise deploying AI workloads on Azure, every knowledge worker using Copilot inside Microsoft 365, every company running Dynamics with AI-embedded automation — all of that activity flows through Microsoft’s revenue line.

The AI arms race created the hardware winners of 2022–2025. The monetization phase — where AI spend converts into measurable enterprise ROI — may create the software winners of 2025–2028. That’s the macro thesis driving the rotation. And Microsoft, by scale, by distribution, and by early investment positioning, may be better placed than any other company to capture it.

Final Thought: What the Smart Money Signal Actually Means

Peter Thiel didn’t abandon AI. He rotated within it — from the infrastructure layer to the application layer. Ken Griffin didn’t get bearish on technology. He added to a position in a $3 trillion company growing at 17% annually at a 30x forward multiple. These are not fear-driven trades. They’re valuation-driven repositionings.

The question for individual investors isn’t “Should I copy what billionaires do?” It’s whether the underlying data — $305 billion in revenue, 10 consecutive quarters of 30%+ Azure growth, $119 billion in net income, $94 billion committed to infrastructure expansion — justifies putting Microsoft on a serious watch list as the AI trade’s next rising star.

Based on where institutional capital appears to be moving, the answer may already be forming.

This editorial is for informational purposes only and does not constitute investment advice. All financial data referenced reflects publicly reported figures. Past performance of any security does not guarantee future results. Investing in equities involves risk, including the possible loss of principal.

—

Rising Star Stocks Editorial Desk