Bill Ackman, chief executive of Pershing Square Capital Management, drew a rare burst of megacap attention on 15 May 2026 when he flagged a new Microsoft position on X minutes before the fund’s quarterly Form 13F-HR hit the U.S. Securities and Exchange Commission wire. The regulatory snapshot, covering positions as of 31 March 2026, is blunt arithmetic: 5,654,078 shares of Microsoft common stock carried a mark-to-market value of $2,092,970,053 in Pershing’s information table—roughly $2.09 billion—making Redmond one of the manager’s largest disclosed U.S. equity lines overnight.
The filing is more than a trophy headline for activist-watchers. Compared with Pershing’s prior 13F (filed 17 February 2026 for the previous quarter-end), Microsoft did not appear at all, while Alphabet appeared twice with Class A and Class C stakes in the millions of shares. In the 15 May table, those Alphabet lines collapse to 32,376 Class A shares and 311,726 Class C shares—still present, but economically residual next to the new Microsoft block. Newsorga reads that pattern as a portfolio-level rotation into Seattle’s bundled cloud/productivity franchise rather than a cosmetic trim; the residual Alphabet stubs may reflect liquidity, tax, or transition mechanics rather than a renewed search bet at scale.
What Ackman argued publicly before the file landed
In a long X thread that circulated widely the same afternoon—and is excerpted in the desk’s linked materials—Ackman framed the purchase as a valuation exercise in a market he thinks hands long-horizon owners occasional gifts. He wrote that converging forces—passive index growth and levered, short-horizon capital—can periodically price “dominant long-term compounding franchises” too cheaply. On Microsoft specifically, he previewed the 13F disclosure and added that Pershing Square USA, a newer closed-end vehicle, had also made Microsoft a core holding even though that pocket will not mirror the 13F cadence—useful context for readers who watch only the flagship file.
His operating bullets match how institutional analysts already model the company: Microsoft 365 as the default fabric for knowledge work, and Azure as the number-two hyperscaler platform riding both enterprise workload migration and AI inference demand. Ackman claimed M365 and Azure contribute on the order of seven-tenths of company-wide profit—a figure readers should treat as portfolio-manager math, not an audited segment allocation, but one that explains why he is buying embedding and compute rather than a single gadget cycle.
OpenAI, exclusivity headlines, and why this story sits at that intersection
Ackman’s thread is explicit that he began accumulating Microsoft in February after a “meaningful” share-price decline following fiscal Q2 2026 results, at what he described as about 21 times forward earnings—a multiple he portrayed as out-of-line with history relative to the durability of the underlying cash flows. The same downstream summaries also carry a line that matters for the OpenAI subplot: Ackman argued Azure competition worries and press narratives about OpenAI partnership changes—particularly language about Microsoft losing exclusive rights to resell the startup’s technology on its cloud—were “overblown.” That is a direct answer to the bear case that Redmond is merely a reseller with a shrinking moat as the model supplier courts multi-cloud routes.
Even readers who discount any one manager’s adjectives should separate three layers: commercial economics between Microsoft and OpenAI, capital intensity as hyperscalers race to add GPU capacity, and sentiment beta when AI capex headlines whipsaw mega-cap multiples. Ackman is effectively betting that the middle layer—spend—will convert into durable revenue attach through Azure consumption and Copilot-style SKU expansion rather than permanently impair returns on invested capital. The 13F cannot prove that thesis; it only proves he bought the ticker.
How to read the numbers—and what still moves without appearing here
13F files are 45-day-lagged snapshots of long equities at quarter-end. They do not capture shorts, options, private deals, or April trading unless a later amendment arrives. The May 15 Microsoft line therefore encodes March 31 positioning; any April Alphabet sales or May adds sit off-stage until the next window. Likewise, Pershing Square USA’s parallel stake—flagged by Ackman—will not show up on the same grid, which matters when benchmarking his true economic exposure.
Cross-currents other desks are watching in the same tape
The same week’s AI narrative is hardly uniform: other high-profile managers have been adding bearish derivatives on semiconductor and software leaders even as cloud hyperscalers post record backlogs. Ackman’s move is a relative-value statement inside that noise—Microsoft as infrastructure and default workflow rather than a pure model lottery ticket. Whether that distinction survives the next GPU supply shock or antitrust remedy is the open empirical question the Markets desk will track through Azure growth rates, net margin bridges, and the next OpenAI-related contract disclosures.
Bottom line
Regulators now show Pershing Square with a fresh, multi-billion-dollar Microsoft stake and a collapsed Alphabet footprint on the 31 March snapshot—hard evidence of a rotation, not a rumor. Ackman’s same-day X narrative supplies the why: forward earnings discipline, M365 scale, Azure AI inference optionality, and a skeptical read on partnership doom loops. The trade will be scored, as always, in free cash flow per share, not in thread length.