Microsoft's 4QFY26 results reinforce the resilience of AI infrastructure demand

Overview

This report is Nomura's quick review of Microsoft's fourth-quarter FY2026 results and management's earnings call, with a primary focus on AI infrastructure, AI platforms, applications and agent commercialization, as well as the 1QFY27E and FY27E outlook. The report believes that Microsoft's commentary on cloud revenue, Azure growth, paid seats for AI applications, and capital expenditure guidance has positive read-through implications for recently pressured AI data center infrastructure stocks.

Core views

The report's core views are as follows: First, demand for Microsoft's cloud and AI offerings remains strong, with Microsoft Cloud revenue up 27% year over year and Azure and other cloud services revenue up 43%, while demand continues to exceed supply; second, capital expenditure has not shown a substantive slowdown, with the actual CY26E guidance unchanged, 1QFY27E capital expenditure expected to exceed USD50bn, and FY27E capital expenditure still expected to grow year over year; third, free cash flow is expected to remain positive in FY27E, helping alleviate concerns about cash flow pressure at hyperscale AI platforms; fourth, AI commercialization is expanding from infrastructure into platforms, applications, and agents, with Copilot, Fabric, Foundry, and GitHub Copilot all showing user or revenue growth.

Analysis framework

The report applies an earnings read-through framework, mapping Microsoft's cloud revenue, bookings, RPO, capital expenditure, data center deployment, AI chip performance, and AI application user numbers to demand across the upstream AI infrastructure supply chain, particularly optical modules, PCBs, and CCLs.

Methodology notes

• Fundamental analysisEarnings review

  Validation of cloud revenue and AI demand

  Enterprise AI and cloud demand intensity is assessed through growth rates for Microsoft Cloud, Azure, M365, and Dynamics 365.

• Supply chain mappingAI capital expenditure read-through

  Transmission of hyperscaler capital expenditure to the hardware supply chain

  Microsoft's continued capital expenditure growth, investment in short-lived GPU/CPU assets, and data center construction are mapped to demand expectations for AI infrastructure names in optical modules, PCBs, CCLs, and related areas.

• Valuation methodologyPrice-to-earnings multiple method

  Target price valuation

  Some of the mentioned stocks are valued using forecast EPS and target P/E multiples. For example, Zhongji InnoLight's target price is based on 20x 2027F EPS, while Shengyi Technology's target price is based on 43x 2027F EPS.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Microsoft (MSFT US)

  Subject of the report and source of AI demand validation

  Strengths

  Cloud revenue, Azure, M365 Copilot, GitHub Copilot, Fabric, and Foundry all showed strong growth; capital expenditure continues to support AI infrastructure expansion.

  Weaknesses

  The report did not assign Microsoft a formal rating, and capital expenditure intensity may still raise market concerns about cash flow and the investment return cycle.

  Comparison

  Relative to AI hardware supply chain companies, Microsoft is a bellwether for hyperscale AI platforms and the demand side.

  Risks

  Capital expenditure returns falling short of expectations, a slowdown in AI demand, and changes in the pace of cloud capacity expansion.

• Zhongji InnoLight (300308 CH)

  AI optical module supply chain beneficiary

  Strengths

  Described by the report as a global optical module leader, benefiting from demand for high-end optical modules in AI data centers.

  Weaknesses

  Faces risks from weaker-than-expected demand for high-end optical modules, intense 400G/800G competition, slower-than-expected product upgrades, and price wars.

  Comparison

  Compared with PCB/CCL suppliers, it has a more direct exposure to high-speed interconnect demand from AI data centers.

  Risks

  High-end optical module demand weaker than expected, slower 800G/1.6T upgrades, and pricing pressure from export customers.

• Shengyi Technology (600183 CH)

  PCB/CCL supply chain beneficiary

  Strengths

  Demand for AI servers, data centers, and high-performance electronic materials can support growth; the report disclosed a Buy rating.

  Weaknesses

  Intensifying competition in the CCL market may pressure margins.

  Comparison

  Compared with optical modules, it more directly reflects the transmission of AI hardware expansion to the materials and board-level supply chain.

  Risks

  Downstream PCB demand from 5G, servers, and automotive electronics falling short of expectations, and intensifying CCL competition.

• WUS PCB (002463 CH)

  PCB supply chain beneficiary

  Strengths

  Mentioned positively as a major PCB supplier, benefiting from upgrade demand for data centers and network platforms.

  Weaknesses

  Growth depends on the realization of downstream demand for 5G base stations, data centers, network products, and automobiles.

  Comparison

  It is in the same PCB/CCL chain as Shengyi Technology, but the specific analyst coverage and valuation disclosures differ.

  Risks

  PCB growth falling short of expectations, weaker-than-expected data center demand, slower-than-expected upgrades of Intel and AMD network products, and automotive demand below expectations.

Key data

• Microsoft Cloud revenueUSD59.3bn+, up 27% year over year4QFY26, reflecting strong demand for Azure and AI applications.
• Intelligent Cloud revenueUSD39.3bn, up 32% year over yearUp 31% year over year at constant exchange rates, driven by Azure.
• Azure and other cloud services revenueUp 43% year over yearCustomer demand continued to exceed available capacity.
• 4QFY26 capital expenditureUSD41.0bn, up 69% year over year and down 29% quarter over quarterIncludes finance leases; approximately two-thirds was allocated to short-lived assets such as GPUs and CPUs.
• 1QFY27E capital expenditure guidanceUSD50bn+Includes the impact of lease reclassification resulting from updated useful-life assumptions.
• CY26E capital expenditure accounting figureUSD175bn, versus USD190bn previouslyThe adjustment resulted from some data centers shifting from finance leases to operating leases; the underlying capital expenditure guidance was unchanged.
• 1QFY27E revenue guidanceUSD89.85-90.95bn, up 18%-19% year over yearManagement outlook.
• Paid Microsoft 365 Copilot seats30mn+Net additions more than doubled sequentially.
• GitHub and GitHub Copilot usersGitHub 225mn; GitHub Copilot 50mnGitHub Copilot revenue growth accelerated to more than 60% quarter over quarter.
• Fabric and Foundry customers40,000+ paying Fabric customers; 100,000 Foundry customersFabric customers grew more than 60% year over year, while Foundry revenue more than doubled year over year in 4QFY26.

Impact & implications

Microsoft's results and outlook provide positive signals for the AI infrastructure chain: cloud and AI demand remain strong, capital expenditure continues to grow, and free cash flow is expected to remain positive, potentially easing concerns about the sustainability of AI data center investment. For China's AI hardware supply chain, optical module leaders and PCB/CCL suppliers may benefit from continued expansion of AI computing infrastructure by global cloud providers.

Risks

• The sustainability of AI infrastructure capital expenditure growth falls short of expectations.
• Free cash flow pressure at hyperscale AI platforms increases.
• Azure and AI application demand slows, or available capacity expansion falls short of expectations.
• Intensifying competition in high-end optical modules, PCBs, CCLs, and other supply chain segments pressures prices and margins.
• Demand for data centers, 5G, servers, automotive electronics, and network platform upgrades falls short of expectations.
• Achievement of certain target prices may be affected by macro market conditions, industry demand, and deviations in company earnings forecasts.

What to watch

• Whether Microsoft's FY27E capital expenditure continues to grow year over year and whether free cash flow remains positive.
• The growth rate of Azure and other cloud services revenue and whether customer demand continues to exceed capacity.
• The pace of growth in paid Microsoft 365 Copilot seats, enterprise penetration, and ARPU.
• GitHub Copilot revenue growth, the rollout of usage-based billing, and growth in Business/Enterprise seats.
• Growth in Fabric and Foundry customer numbers and token annualized run-rate customers.
• The transmission of AI data center demand into orders for 800G and 1.6T optical modules and high-end PCBs/CCLs.