By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
The panel is largely bearish on Microsoft's AI-driven growth prospects, citing slow Copilot adoption, high capex, and potential regulatory risks. They agree that the Gates foundation exit is not a bearish signal, but the market's skepticism towards Microsoft's AI-driven margin expansion is a significant concern.
Risk: The single biggest risk flagged is the potential for both hyperscalers to overspend on inference capacity while enterprise customers delay adoption, leading to a sector-wide margin trap (Claude).
Opportunity: The single biggest opportunity flagged is the potential for Copilot and OpenAI-based offerings to translate into revenue per seat and overall ARR growth, which could lead to multiple re-rating (ChatGPT).
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Moneywise and Yahoo Finance LLC may earn commission or revenue through links in the content below.
The Gates Foundation Trust, founded by Bill Gates and Melinda French Gates, recently disclosed that it sold its final 7.7 million shares of Microsoft during the first quarter of 2026 — a roughly $3.2 billion exit that ends a decades-long position in the company Gates cofounded (1).
A very different story had already broken on the morning of the disclosure. Hours before the Gates filing hit the SEC on May 15, Bill Ackman, founder and CEO of Pershing Square Capital Management, used a lengthy X post to announce a brand-new Microsoft position (2). Pershing’s 13F filing, filed later that evening, showed roughly 5.65 million shares, worth around $2.09 billion, on the books at quarter-end (3).
The following morning, Ackman explained he used Pershing’s holdings of Alphabet, the parent company of Google, to pay for it. “To be clear, our sale of $GOOG was not a bet against the company,” he wrote on X (4). “We are very bullish long term on Alphabet. But at current valuations and in light of our finite capital base, we used $GOOG as a source of funds for $MSFT.”
The 7.7 million-share sale is the final tranche of a multi-year sale by the Gates Foundation Trust. The trust held roughly 28.5 million Microsoft shares at the end of the first quarter of 2025, trimmed to 7.7 million by year-end and zeroed out this quarter.
This comes after Gates announced in May 2025 that the foundation will sunset operations in 2045 and spend roughly $200 billion on charitable work over the next 20 years (5).
“There are too many urgent problems to solve for me to hold onto resources that could be used to help people,” he wrote. “That is why I have decided to give my money back to society much faster than I had originally planned.”
A foundation winding down its endowment doesn’t have a choice but to sell — it has to fund the giveaway.
Read More: Here’s the average income of Americans by age in 2026. Are you falling behind?
Pershing started stockpiling Microsoft in February, right after the company's fiscal Q2 2026 earnings sent the stock tumbling (2). It kept buying through a stretch where Microsoft was down sharply on the year and well off its July 2025 record high.
Two worries had spooked the market.
First, Copilot adoption. Microsoft has converted only about 15 million of its 450 million paid Microsoft 365 commercial seats into paying Copilot users. Independent research showed Copilot’s market share fell from 18.8% in July 2025 to 11.5% by January 2026 (6). That decline prompted CEO Satya Nadella to reorganize the AI division in March and sideline the AI executive he paid $650 million to recruit (7).
Second, the AI capex bill. Microsoft is spending $190 billion on capital expenditure in 2026. Some investors think the math doesn’t work.
Ackman thinks they’re wrong on the capex part. Azure revenue grew 39% in constant currency last quarter. Microsoft’s AI business hit a $37 billion annualized run rate, up 123% year-over-year (8). Ackman called the $190 billion “growth capex that should drive future revenue generation” rather than a margin threat (2).
He also thinks the market is undervaluing the core franchise. Microsoft 365 and Azure together throw off roughly 70% of Microsoft’s overall profits. M365 monthly ARPU sits around $20 — less than half what customers would pay for the underlying applications individually. Investors, he wrote, “underestimate the resilience of the M365 franchise given its deeply embedded role across enterprises and highly attractive price-value proposition (2).”
Then there’s the OpenAI stake. Microsoft owns roughly 27% of OpenAI economically. At the most recent funding round, that’s about $200 billion — or 7% of Microsoft’s market cap. Ackman says the share price doesn’t reflect any of that.
Pershing got in at roughly 21 times forward earnings. Ackman called that “broadly in line with the market multiple and well below Microsoft’s trading average over the last few years.”
The Google side of the rotation is where the conviction shows.
Pershing held more than 6.1 million Alphabet Inc. Class C shares at the end of 2025. By the close of the first quarter of 2026, that was down to about 312,000, or a 95% cut, worth roughly $1.64 billion at quarter-end pricing. Class A holdings fell from around 678,000 shares to 32,000 over the same stretch. The remaining position was fully liquidated in the second quarter, according to a person familiar with the portfolio, as cited by Reuters (9).
Pershing had owned Google for three years at an average cost of about $94 a share. Class C was near $392 the Friday the 13F dropped. That’s roughly four times Ackman’s cost basis.
He rolled the proceeds straight into Microsoft, then went on X the following morning to make it clear that the sale wasn’t a call on Google’s prospects.
To summarize, Ackman zeroed out a three-year, multibillion-dollar Google position, picked up Microsoft at 21x forward earnings and bet that the market has mispriced the enterprise franchise against its AI uncertainty. The Gates sale, while being reported on social media like a bear signal, is its own thing — a foundation funding $200 billion in giveaways. Selling Microsoft because of that would be misguided.
Ackman’s trade is the one with a thesis behind it: Copilot’s paid base is running at about 3% of its addressable seats. That gap is what he’s buying.
Bill Ackman has spent nearly 34 years navigating the market’s ups and downs. He cofounded his first investment firm, Gotham Partners, back in 1992 before founding Pershing Square Capital Management in 2004 with just $54 million. Today, Pershing oversees billions in assets (10).
But it’s important to remember that Ackman isn’t making these calls alone from a laptop at home. Billion-dollar hedge funds operate with teams of analysts and industry experts who have access to institutional-grade research and sophisticated market data.
Everyday investors often don’t have access to that same level of research — or the ability to absorb losses if a trade goes wrong. That’s why following billionaire investors can be risky.
For investors who want exposure to the market but aren’t confident picking individual winners, platforms like Moby can help. Their platform offers expert research and recommendations to help you identify strong, long-term investments backed by advice from former hedge fund analysts.
In four years, and across almost 400 stock picks, their recommendations have beaten the S&P 500 by almost 12% on average. They also offer a 30-day money-back guarantee.
Moby’s team spends hundreds of hours sifting through financial news and data to provide you with stock and crypto reports delivered straight to you. Their research keeps you up-to-the-minute on market shifts and can help you reduce the guesswork behind choosing stocks and ETFs.
Plus, their reports are easy to understand for beginners, so you can become a smarter investor in just five minutes.
At the same time, keeping investing costs low can be just as important as choosing the right stocks. High fees, trading commissions and unnecessary account costs can quietly eat away at long-term gains over time.
Using a discount broker with low or zero commissions on trades can help investors preserve more of their returns. Lower investing costs mean more of your money stays in the market, compounding over the long run instead of being chipped away by fees every time you make a move. Even saving a small percentage annually can add up to thousands of dollars over decades of investing.
SoFi’s easy-to-use DIY investing platform lets you<a href="https://moneywise.com/c/1/236/1718?placement=4"> buy stocks, ETFs and more</a> with no commission fees and no account minimums.
SoFi is designed for both beginners and seasoned investors, with real-time investing news, curated content and the data you need to make smart decisions about the stocks that matter most to you.
Plus, for a limited time you can get up to $1,000 in stock when you fund a new account.
For investors managing larger portfolios, professional guidance can play an essential role in protecting and growing wealth. Those with portfolios of $250,000 or more may in particular benefit from working with a financial advisor to help avoid costly mistakes and build a plan aligned with long-term goals like retirement.
So, if you have a portfolio of $250,000 or more, platforms like WiserAdvisor can connect you with vetted professionals who specialize in this kind of planning.
Simply answer a few questions about your savings, retirement timeline and overall investment portfolio.
From there, WiserAdvisor reviews its network to match you — for free — with up to three vetted, reputable advisors aligned with your specific needs.
You can then schedule no-obligation consultations with your matches to determine who is the best fit for your long-term goals.
WiserAdvisor is a matching service and does not provide financial advice directly. All matched advisors are third parties, and specific financial results are not guaranteed.
— With files from Rudro Chakrabarti
Join 250,000+ readers and get Moneywise’s best stories and exclusive interviews first — clear insights curated and delivered weekly. Subscribe now.
We rely only on vetted sources and credible third-party reporting. For details, see our editorial ethics and guidelines.
Securities and Exchange Commission (1),(3); @BillAckman (2),(4); Gates Foundation (5); Recon Analytics (6); Microsoft (7),(8); Reuters (9); Forbes (10)
This article provides information only and should not be construed as advice. It is provided without warranty of any kind.
Four leading AI models discuss this article
"Microsoft’s $190B 2026 capex risks sustained margin pressure unless Copilot monetization scales far faster than current 3% seat penetration implies."
The article frames Gates’ exit as purely philanthropic and Ackman’s MSFT purchase at 21x forward earnings as a high-conviction correction of market mispricing on Azure growth and the OpenAI stake. Yet Pershing liquidated a 3-year Alphabet winner to fund it, signaling finite capital is being redeployed into a name already carrying $190B 2026 capex against only 15M paying Copilot seats. If Azure’s 39% growth decelerates or enterprise ARPU stays capped near $20, the re-rating case collapses quickly. The structural seller dynamic does not change the fact that MSFT’s AI spend must deliver visible margin expansion by late 2027 or the trade becomes a crowded, expensive bet.
Ackman’s thesis could still hold if Azure constant-currency growth sustains above 30% and the 27% OpenAI economic interest is eventually marked up, turning the $190B capex into genuine growth spend rather than a margin trap.
"Gates' complete exit from a company he founded, coinciding with Copilot's market share collapse and a $190B capex program with no clear payback timeline, is a stronger signal than Ackman's rebalance from a 4x-gain position in GOOG."
The article frames Gates' exit as irrelevant (forced liquidation for charitable spending) while positioning Ackman's entry as conviction. But this inverts the signal hierarchy. Gates built Microsoft; his foundation held the stock for decades through multiple cycles. A founder-backed entity exiting entirely—not trimming, but zeroing out—while the company faces Copilot adoption collapse (15% penetration, market share halving YoY) and a $190B capex bet with unproven ROI deserves more weight than one activist's contrarian call. Ackman bought at 21x forward earnings after a selloff; that's not cheap on a normalized basis. The article also glosses over execution risk: Nadella just sidelined his $650M AI hire, suggesting internal conviction issues on the AI roadmap itself.
Ackman's thesis on M365 pricing power and OpenAI optionality is legitimate; a 3% Copilot adoption rate on 450M seats is genuinely early, not broken, and the $37B AI run rate growing 123% YoY is real revenue, not vaporware.
"Microsoft's massive capital expenditure is currently failing to translate into meaningful incremental revenue growth, leaving the stock vulnerable to further valuation multiple contraction."
Ackman’s pivot into MSFT at 21x forward earnings is a classic value-trap play disguised as growth. While he highlights the 3% Copilot penetration as 'upside,' the market is clearly pricing in a structural slowdown in AI ROI. Spending $190 billion in capex to defend a 365 franchise that is already saturating is a margin-dilutive strategy, not a growth one. The Gates exit is noise, but the market's skepticism toward Microsoft's AI-driven margin expansion is signal. Unless MSFT can prove that Copilot justifies a permanent price hike rather than just cannibalizing existing seat revenue, the valuation compression will continue.
If Microsoft's Azure growth maintains its 39% trajectory, the $190 billion capex will be viewed as a necessary moat-building exercise that cements their dominance, making 21x forward earnings look like a massive entry point bargain.
"If Copilot monetizes faster and the AI capex ROI proves out, Microsoft could command a higher multiple than today’s ~21x forward earnings."
Big takeaway: Gates selling to fund philanthropy isn't a bearish signal about Microsoft; Ackman's MSFT purchase suggests he sees durable AI-driven growth, but the catalysts carry execution risk. The article highlights Copilot adoption (only ~3% conversion of paid seats) and a $190B 2026 capex line—both could pressure near-term margins if ROI isn't as strong as hoped. Yet Azure+M365 generate the bulk of profits (~70%), and OpenAI stake isn't currently priced into the stock. A key missing context is how quickly Copilot and OpenAI-based offerings translate into revenue per seat and overall ARR growth; if that accelerates, multiple re-rating is plausible. If not, risk to downside remains.
The Gates exit could be interpreted as a signal of resource reallocation away from long-hold equities, not a validation of MSFT AI strength; Copilot monetization risk remains a real headwind if enterprise pricing and adoption stall.
"Pershing's Alphabet-to-MSFT rotation creates direct pressure on Google Cloud that the Gates exit alone does not explain."
Claude weights the Gates foundation exit heavily but misses how Pershing's simultaneous Alphabet sale to fund the MSFT position creates an explicit rotation from one AI infrastructure name into another. If Azure's 39% growth slows, that capital shift could accelerate Google Cloud share losses without any change in MSFT fundamentals. The $190B capex then becomes a defensive spend against both OpenAI alternatives and hyperscaler peers rather than pure growth.
"Hyperscaler capex arms race could compress margins across the sector if enterprise AI ROI remains unproven, making Ackman's conviction bet a sector-wide bet disguised as stock-picking."
Grok's rotation thesis is sharp but assumes Alphabet's cloud trajectory is independent of MSFT's capex intensity. If Azure's 39% growth persists *because* of that $190B spend moat-building, Google Cloud doesn't automatically gain share—it faces the same ROI pressure. The real risk: both hyperscalers overspend on inference capacity while enterprise customers delay adoption pending clearer unit economics. That's a sector-wide margin trap, not a MSFT-specific one. Ackman's bet assumes MSFT wins that race; nobody's priced the scenario where both lose.
"The $190B capex strategy invites aggressive regulatory scrutiny that will likely force a valuation multiple compression, regardless of Azure's growth metrics."
Claude is right about the sector-wide margin trap, but everyone is ignoring the regulatory shadow. Microsoft’s $190B capex isn't just a financial risk; it’s a massive antitrust target. If the FTC or EU views this spend as a weaponized barrier to entry for smaller LLM competitors, the valuation multiple will compress regardless of Azure growth. Ackman is betting on a winner-take-all AI outcome that regulators are actively trying to dismantle. This isn't just a capex risk; it's a structural political risk.
"ROI realism on MSFT's 190B AI capex is the real risk; regulatory headlines aside, unit economics determine whether Copilot monetization actually expands margins."
Gemini's regulatory risk angle is valid, but the bigger, underappreciated risk is ROI realism: 190B capex funded by productivity gains must translate into margin expansion; if Copilot pricing power is weaker than modeled, and if data privacy/localization adds compliance costs, the implied ROIC may never exceed corporate hurdle rates. Regulators may slow things, but the spend's success hinges on unit economics, not antitrust headlines.
The panel is largely bearish on Microsoft's AI-driven growth prospects, citing slow Copilot adoption, high capex, and potential regulatory risks. They agree that the Gates foundation exit is not a bearish signal, but the market's skepticism towards Microsoft's AI-driven margin expansion is a significant concern.
The single biggest opportunity flagged is the potential for Copilot and OpenAI-based offerings to translate into revenue per seat and overall ARR growth, which could lead to multiple re-rating (ChatGPT).
The single biggest risk flagged is the potential for both hyperscalers to overspend on inference capacity while enterprise customers delay adoption, leading to a sector-wide margin trap (Claude).