During Microsoft's fiscal 2026 fourth-quarter earnings call, executives indicated that AI demand remains a strong tailwind for the company. Azure demand continues to outpace available capacity, a supply-demand imbalance that is not expected to ease in the near term.
Capital expenditure has been lowered to approximately $175 billion due to a change in accounting policy.
As reported, Microsoft released its fiscal Q4 earnings after the US market closed on the 29th. Revenue grew 18% year-over-year to $90.01 billion, exceeding analyst expectations by 2.6%.
For the full fiscal year, total revenue surpassed $331 billion, an 18% increase. Microsoft Cloud revenue exceeded $214 billion, growing 27%.
On the earnings call, Microsoft executives stated that capital expenditure for the entire 2026 calendar year is still expected to be the previously guided $190 billion. However, due to a change in accounting methodology that adjusted the estimated useful lives of data centers and office buildings, the expected expenditure for this year is reduced to approximately $175 billion, which is nearly 8% lower than the $190 billion figure.
Azure's growth rate accelerated to 43% this quarter, up from 40% previously, with guidance for the next quarter pointing to growth of about 45%, surpassing market expectations. CEO Satya Nadella stated that demand continues to outpace available capacity, and this supply-demand imbalance is not expected to ease in the short term.
CFO Amy Hood emphasized that the engineering team has made significant progress in improving CPU and GPU efficiency. The newly freed-up capacity was quickly monetized within the quarter, becoming a key driver for the quarter's better-than-expected performance.
The AI Infrastructure Surge and Effectively Reduced CapEx
The market's most-watched cloud business and AI infrastructure orders showed strong performance in Q4. Azure and other cloud services revenue achieved 43% growth, exceeding expectations.
Microsoft CFO Amy Hood directly addressed the current supply-demand contradiction, stating, "The situation is clearly that demand still exceeds our available supply. You can even see this in asset pricing in the spot market."
To address this, Microsoft has significantly accelerated its infrastructure buildout.
Microsoft CEO Satya Nadella revealed the remarkable pace of expansion, noting, "We added 31 new data centers across five continents this quarter, bringing the total to 88 this year. We added another gigawatt of compute capacity this quarter and remain on track to roughly double our total compute capacity in just two years."
Regarding future guidance, Amy Hood indicated that despite a larger capacity base, Azure's constant currency growth rate is expected to further accelerate to around 45% in the first quarter of fiscal 2027.
Simultaneously, due to an accounting policy change extending the useful life of data centers from 15 to 25 years, some leases have been reclassified from finance leases to operating leases. As a result, total capital expenditure for fiscal 2027 is expected to be around $175 billion.
Defending Core Enterprise IP with Model Choice and "Agent Architecture"
When answering an analyst's question about enterprise concerns regarding open-source models and protecting intellectual property, Satya Nadella provided market insight, which is the core logic of Microsoft's future AI architecture design.
Satya Nadella emphasized, "This is not going to be a model where you give away all your knowledge and get nothing in return. Every enterprise will evaluate which providers are genuinely helping them create value and knowledge."
Based on this market understanding, Microsoft has introduced an underlying architecture that decouples the model from context memory and scheduling tools, making any model replaceable at any time.
He stated that Copilot, GitHub Copilot, and Security Copilot are all built on this architecture, and Microsoft is working to promote this design pattern to all enterprise customers. Nadella said, "You can use frontier models, you can use multiple models simultaneously, or you can train your own proprietary model using your own accumulated outputs, track records, and context when you don't want to rely on any external model."
He also cited the recent Hugging Face security incident as an example of the systemic risk of relying on a single model, stating, "You cannot be at the mercy of a denial of service from one model." Satya Nadella added, "When we talk about frontier models, we often think of it as a single thing. But the true frontier is that every company can have its own frontier, with the model choice, cost control, and business capabilities needed to control its own destiny."
Hood added from a financial perspective, stating that regardless of which model provider a client chooses or if they run their own models, Azure remains the underlying infrastructure. Therefore, demand growth has considerable resilience for Microsoft. She stated, "This infrastructure is quite fungible."
Business Model Reinvention: A Dual "Subscription + Consumption" Approach
Beyond infrastructure, the commercialization of the AI application layer and the sustainability of revenue are other major concerns for investors.
The earnings report shows that M365 Copilot now has over 30 million paid seats, with net new seat additions doubling quarter-over-quarter. GitHub Copilot, aimed at developers, has reached 50 million users, and its revenue growth accelerated by over 60% quarter-over-quarter.
More critically, Microsoft is changing the billing logic for its AI products, which will significantly expand its total addressable market.
Satya Nadella dropped this significant signal during the call, stating, "We are evolving our business model from a pure 'per seat' model to a 'per seat plus consumption' model. This will further expand our total addressable market and deliver more customer value."
This means that as AI becomes deeply integrated into all aspects of enterprise business processes, Microsoft's revenue will no longer be limited solely by the number of enterprise employees but will be directly linked to the actual consumption volume of enterprise AI model calls.
Customer service and business applications are leading this transformation. Satya Nadella mentioned, "Customer service is at the forefront of this transition, and the consumption-based credit usage in this category grew fourfold quarter-over-quarter."
Supply-Demand Dynamics and Cost Pressures: Flexible Management and Long-Term Confidence
Analyst Mark Moerdler raised dual concerns about potential oversupply of AI infrastructure and rising hardware prices. In response, Hood emphasized the inherent flexibility in Microsoft's capital expenditure structure.
Hood stated, "Our capital expenditure has shifted significantly towards short-lived assets, primarily CPU and GPU. If the demand environment changes, we can adjust procurement pace relatively quickly, and the timing for land and data center construction also has considerable flexibility."
She also noted that Microsoft's highly diversified business portfolio across geographies, customer segments, and industries, along with its first-party application business's ability to absorb its own compute capacity, provides a significant buffer against demand fluctuations.
Regarding the impact of rising hardware prices on profit margins, Hood said that cloud computing still offers significant cost advantages compared to customers building their own servers. New contracts already reflect adjusted pricing. Hood stated, "For both customers and for us, we want pricing to be effective over the long term."
Nadella added that getting the product portfolio right, having a diverse customer base, and continuously improving efficiency are the three key factors for navigating industry cycles. Nadella said, "There will be ups and downs, but the long-term structural shift is clear. We are very confident in our business portfolio, our margin structure, and most importantly, the value we create for our customers."
Comments