While Meta and Alphabet raise concerns in the market over AI cash burn, Microsoft (MSFT.US) counters with over $10 billion in cash flow and leads AI commercialization from story to profit.

According to Zhitong Finance APP, Microsoft (MSFT.US) released its Q1 FY2027 results (ending September 30, 2026) after the US market closed on Wednesday Eastern Time. The company’s revenue, earnings per share, and Azure growth — its three core indicators — all exceeded market expectations. Notably, for the first time, Azure’s cloud revenue was disclosed in USD terms, providing investors with an observation anchor that had previously been missing.

The financial report shows Microsoft’s total revenue for the quarter was $92.3 billion, up 19% year-over-year; operating profit was $45.9 billion, rising 21% year-over-year (22% in constant currency); GAAP net profit was $36.6 billion, up 32% year-over-year, corresponding to diluted EPS of $4.92, also up 32%. Excluding the impact of the OpenAI investment, non-GAAP net profit was $36.4 billion (up 18% year-over-year), and non-GAAP diluted earnings per share were $4.89 (up 18% year-over-year).

In comparison, Wall Street analysts expected adjusted EPS of $4.73 and revenue of $90.69 billion. Microsoft CFO Amy Hood described the quarter in a statement as “a strong start with revenue, operating profit, and EPS all exceeding expectations.”

In detail, Azure was disclosed in USD for the first time, becoming Microsoft’s single largest business. Azure revenue for the quarter was $33.17 billion, a 48% year-over-year increase, accelerating from 42% last quarter and higher than the company's guidance of 44–45% (constant currency) provided in July and analysts’ estimates of about 44%. This marks the first time since IPO that Microsoft disclosed Azure’s USD revenue separately; previously, only growth rates were revealed. Microsoft Cloud’s total revenue was $63.9 billion, up 30% year-over-year (31% in constant currency), with its growth rate further accelerating from prior quarters.

The most direct support for revenue growth comes from Commercial Remaining Performance Obligation (Commercial RPO), which grew 73% year-over-year to $677 billion — this “signed but not yet recognized as revenue” backlog quantitatively supports the narrative of “demand still outstripping supply.”

Another noteworthy point: starting in FY2027, Microsoft will reorganize its three segments (Productivity & Business Processes, Intelligent Cloud, More Personal Computing) of the past 11 years into two operating units: Agents and Infra, and Devices and Consumer. Historical data have been restated accordingly.

Data shows Agents and Infra delivered revenue of $76.8 billion this quarter, an increase of 25% year-over-year, higher than the StreetAccount consensus of $75.63 billion. This segment includes Azure, Microsoft 365 Cloud, Productivity and Server Licensing, Industry Solutions, and Enterprise & Support Services, representing the main battlefield for Microsoft’s AI monetization.

Devices and Consumer revenue this quarter was $15.48 billion, down 3% year-over-year, but still above the $15.05 billion consensus. Windows OEM and device revenue declined 22% year-over-year, which Microsoft attributed to increased memory and storage costs pushing up prices, the high year-ago comparison due to the end of Windows 10 support, and weak PC demand.

At a product level, Microsoft 365 commercial cloud revenue increased 17% year-over-year (18% adjusted, 19% in constant currency, normalized for last year’s one-off revenue recognition benefit); Search and advertising revenue (excluding traffic acquisition costs) grew 12% year-over-year; Xbox content and services revenue declined 2% year-over-year.

Two Lines of AI Commercialization: OpenAI Revenue Sharing and Copilot Adoption

The report provided evidence for two directions of AI monetization, corresponding to “the relationship with OpenAI” and “the in-house Copilot ecosystem,” partly alleviating market concerns over “AI threatening Microsoft’s software empire” and “Microsoft’s AI business being excessively tied to the yet-unprofitable OpenAI.”

The report disclosed that income from commercial arrangements with OpenAI reached $10 billion. For the quarter, total income from commercial agreements with OpenAI (including revenue sharing) amounted to $10 billion, about 11% of Microsoft’s total sales, up from around 7% for the full FY2026. The report also noted that as of September 30, receivables from OpenAI were $7.2 billion (compared to $6 billion as of June 30); Microsoft’s total investment commitment to OpenAI is $13 billion, with $12 billion already funded, resulting in an equity-method based stake equivalent to about 24% of OpenAI.

It should be noted that a significant portion of this $10 billion comprises Microsoft's sale of Azure compute to OpenAI and does not entirely represent net new external demand.

Microsoft also revealed growth in Copilot users. CEO Satya Nadella said on the conference call that “Microsoft 365 Copilot users increased by 50% quarter-over-quarter,” but emphasized “paid seats were not disclosed.” While the call’s transcript records CFO Hood clarifying, “This quarter, we did not update Copilot numbers; we’re talking about revenue and usage momentum,” the two statements are not contradictory — the former refers to total user growth (including free/trial users), while the latter refers to undisclosed paid seats.

In addition, the newly launched premium productivity suite Microsoft 365 E7 has already garnered “millions of seats.” Nadella’s “low-cost strategy” — including Copilot Cowork multi-model agents and a range of lower-priced models — is seen as targeting enterprise customers who are increasingly sensitive to AI deployment costs. All these data points indicate “Microsoft is profiting from enterprise AI demand, alleviating fears that this technology threatens its software empire or that its fate is overly tied to unprofitable OpenAI.”

Divergent AI Returns: Azure Growth and the Rarity of Positive Free Cash Flow

Against a backdrop where tech giants like Alphabet (GOOGL.US) and Meta (META.US) have seen their free cash flow turn negative or shrink sharply due to massive AI infrastructure spending, Microsoft still maintained positive free cash flow this quarter, representing the clearest differentiation among individual stocks in this phase of AI capital expenditure. Following the earnings report, Microsoft’s stock price rose more than 6% after hours, in stark contrast to Alphabet’s more than 9% decline.

Microsoft’s free cash flow for the quarter was $18.6 billion, down 27% year-over-year, but significantly higher than analysts’ prior estimate of about $8 billion. The key to this “beat” lies in capex coming in below the company’s own guidance: this quarter’s capex (including finance leases) was $45.2 billion, a 30% year-over-year increase, but below the “over $50 billion” guidance provided in July, and less than the StreetAccount consensus of $50.67 billion. CFO Hood attributed the shortfall on the call to delays in third-party capacity (GPU as a Service, GPUaaS) deliveries, not a contraction in spending plans: “Some third-party capacity expected to be delivered in Q1 was in fact delayed.”

By contrast, Alphabet’s Q3 free cash flow was -$9.07 billion, Meta last week reported negative free cash flow for the first time, and Oracle (ORCL.US) said on the same day that capex would peak this and next fiscal years. Investing.com senior analyst Jesse Cohen described Microsoft’s report as “about as clean as it gets” for a company of its size, serving as a “counterweight” after Meta and Alphabet data reignited AI spending anxiety: “Demand still exceeds supply, commercial backlog is massive, and the returns have finally flowed through to profits and overall revenue.”

Chad Morganlander, senior portfolio manager at Washington Crossing Advisors, also pointed out that Microsoft is “the only one among the four major compute giants not to see annual free cash flow turn negative,” adding, “As a company that is still investing yet not falling into the red on the books, it looks like the adult in the room.”

However, it’s worth noting that “better-than-expected” free cash flow is to a large extent passively due to lower capex resulting from third-party capacity delivery delays — this is essentially a timing mismatch, not an efficiency leap. To conclude that “AI is already self-sustaining” would muddle accounting timing with true investment returns.

A recent research note from Goldman Sachs maintained a “buy” rating on Microsoft but also flagged that “continuing expansion of AI infrastructure investment will keep pressuring Microsoft’s free cash flow,” and projected FCF will fall from about $67 billion in FY2026 to around $40.1 billion in FY2027, and further to about $14.3 billion in FY2028. In other words, positive FCF is a “relative advantage” for now, not “absolute and stress-free.”

Capital Expenditures: Below Expectations Due to Delay, Not Contraction

The “below expectation” capex this quarter is the most easily misunderstood signal in this report. Proper understanding requires distinguishing between accounting/delivery timing and true investment intent.

Official data show capex (including finance leases) this quarter was $45.2 billion, up 30% year-over-year (from $34.9 billion); cash-based "purchases of property and equipment" were $39.367 billion — nearly doubled year-over-year (previously $19.394 billion). Hood directly linked the shortfall to third-party GPUaaS capacity delivery delays and stressed that “demand still exceeds supply.”

Capex will accelerate in the second half. On the call, Hood guided: Q2 capex (including finance leases) will exceed $50 billion, and “due to the above-mentioned third-party capacity delays, capex in the second half of the fiscal year is expected to accelerate, mainly driven by CPUs (with higher memory and storage costs), GPUs, and delayed third-party capacity deliveries now scheduled for H2”; full-year 2027 calendar year capex guidance is about $170 billion.

A detail easily overlooked: about 75% of capex is currently going to short-cycle assets (such as Intel (INTC.US) CPUs, Nvidia (NVDA.US) GPUs), up from about 66% in the previous three quarters and about half last year. This shift shows Microsoft is prioritizing infrastructure that can quickly be converted to revenue — “using faster asset turnover for faster revenue recognition,” rather than sheer scale expansion.

Another accounting change worth noting: In July, Microsoft proactively extended its datacenter building useful life from 15 to 25 years and reclassified some leases from finance leases to operating leases, causing reported capex to drop from $190 billion to $175 billion — but actual construction spending was unchanged. This means a decline in reported capex does not imply a slowdown in actual construction; tracking Microsoft’s AI infrastructure intensity requires monitoring cash purchases, finance leases, and operating lease commitments together.

Capital Efficiency Shift and "Responsible Spending" Narrative

Despite the massive investments, Microsoft is signaling “responsible spending” to the market through a series of capital efficiency measures, underpinning its valuation premium relative to peers.

D.A. Davidson analyst Gil Luria stated that Microsoft “continues to have the most responsible approach” to spending. Supporting this judgment are three parallel factors: first, the aforementioned shift of about 75% of capex toward short-cycle assets; second, Microsoft’s layoff of 4,800 employees this quarter (about 2.1% of the total workforce) and an October 26 offer of voluntary buyouts to around 60,000 employees (about one-fourth of the workforce) to balance costly AI buildout and capital efficiency; third, the launch of low-priced models and Copilot Cowork multi-model switching, to capture broader demand with lower-cost product portfolios.

However, “responsible spending” is essentially a narrative and not a financial outcome, and should be presented alongside “second-half capex acceleration.” In fact, Bank of America’s September 2026 global fund manager survey found that 42% of respondents see “AI infrastructure spending” as the most likely source of a systemic credit event (up from 38% in August), and 53% characterized “buying semiconductor stocks” as the “most crowded trade” — market skepticism over AI capex is systemic and not limited to individual companies.

Still, as “debt-driven spending” increasingly raises market concerns, Microsoft’s credit quality remains the key buffer enabling it to “trade time for space.”

According to Truist’s statistics, Microsoft’s five-year credit default swap (CDS, a measure of perceived default risk) remains the tightest among major tech firms, even as spreads on most AI-related names have widened. Hood also acknowledged on the call: “Given our rating, we’re fortunate to be able to fund in the debt markets… I have to say, we haven’t yet tapped (the debt market).”

Looking ahead, Microsoft offered upbeat guidance on the call: Q2 revenue is expected to be $96.0–$97.1 billion (midpoint implying about 18.8% year-over-year growth), higher than the $95.6 billion LSEG consensus; Q2 Azure is expected to grow 49–50% in constant currency, above the StreetAccount consensus of 45.1%. Hood stated: “We expect demand to continue to outpace supply.”

This guidance extends the “demand outstripping supply” theme, but also means the market’s verification window for Microsoft’s AI investments will lengthen further — the real question is not this quarter’s or next quarter’s growth, but whether, as capex of about $170 billion in calendar 2027 is gradually rolled out and third-party capacity delays are made up in H2, free cash flow can remain positive “without passively dropping below guidance,” and whether Azure’s high growth can sustain on a larger base.