TL;DR
· Goldman Sachs maintains a Buy rating on Microsoft and a $610 price target, implying about 59% upside based on the stock price as of July 9.
· Azure growth remains the main focus of the earnings report. Goldman Sachs expects Q4 growth of 40%-41%, higher than the company’s prior guidance.
· Higher capital expenditures will amplify the debate over returns. Monetization of Copilot, the Maia chip, and new capacity ramp-up still need to be delivered.

Goldman Sachs maintained its Buy rating on Microsoft ahead of the company’s Q4 earnings report on July 29 and set a 12-month target of $610, while also raising its outlook for long-term capital expenditures. For investors, the earnings focus is not whether Microsoft is an AI winner, but whether Azure can sustain high growth while continuing to ramp up compute, turning higher spending on data centers, chips, and power into revenue rather than dragging on free cash flow and profit margins.

Behind the $610 price target, Azure still needs to keep beating expectations

Market data shows that as of July 9 UTC, Microsoft’s stock price is about $383.34. At this price, the $610 target implies approximately 59.1% potential upside.

This set of assumptions rests on several conditions: cloud demand keeps growing at a high rate, new data center capacity comes online as planned, Microsoft’s internal AI R&D and the allocation of compute to external customers don’t crowd each other out, and AI products such as Copilot start contributing clearer revenue and profits.

In the earnings report, Azure remains the first thing being watched.

Microsoft’s FY26 Q3 earnings call will show that Azure and other cloud services revenue grew 40% year over year, and 39% on a constant-currency basis. The company’s prior FY26 Q4 guidance was 39%-40% growth on a constant-currency basis, and it said customer demand still exceeds available capacity.

Goldman Sachs reported that Azure’s year-over-year growth on a constant-currency basis in Q4 could reach 40%-41%, and next-quarter guidance may also stay at 40%-41%. This forecast is slightly higher than the company’s prior guidance, but market expectations are already not low. If Microsoft merely delivers cloud growth in line with the already-high expectations, the stock may not continue to price in further AI investment.

Microsoft also needs to explain where the growth comes from. It could be from newly released data center capacity, from continued expansion in enterprise AI demand, or from smoother compute scheduling between internal applications and external customers.

In the past few quarters, the constraint on Microsoft’s AI business has not been insufficient demand, but tight supply. Azure must both serve external customers such as OpenAI and support Microsoft’s internal Copilot, MAI model R&D, and first-party applications. When compute is tight, cloud growth is limited by delivery capability. If capacity ramp is too slow, capital expenditures will first show up in cash flow and depreciation pressure.

Microsoft FY26 compute capacity spending broken down by use and external/internal compute allocation. AI compute, MAI, Copilot, and others have a high share. After internal compute spending increased over the past 12 months, it has moved toward a steady-state. This is key to judging whether Azure can support both customer demand and internal AI R&D at the same time.

Capital expenditures keep getting revised upward; the AI compute race hasn’t cooled

Microsoft has already sent signals of higher spending. FY26 Q3 capital expenditures were $31.9 billion. The company guided that Q4 capital expenditures will exceed $40 billion and expects 2026 calendar-year capital expenditures of about $190 billion, including about $25 billion from higher component prices.

Goldman Sachs said Microsoft’s capital expenditure outlook for fiscal years 2028-2030 has been raised by about 10%. Based on the report’s calculations, the adjusted portions of annual capital expenditures assume levels above market consensus, reflecting a more aggressive view of Microsoft’s future compute investment.

This isn’t a choice only Microsoft gets to make. Guidance from chipmakers such as Nvidia, Broadcom, and AMD, as well as capital moves by cloud and internet giants like Google and Meta, all show that AI compute demand has not clearly cooled. Hyperscale cloud providers are still preparing to expand data centers, chips, and power resources in the coming years.

For Microsoft, higher investment has two sides.

On one hand, Azure and the AI product cycle still provide valuation support. Goldman Sachs said that by mid-2030, Microsoft’s compute capacity could expand to around 40GW. On the other hand, the higher the capital expenditures, the more investors will ask whether new compute can be converted into cloud revenue, AI subscriptions, and higher-margin businesses—rather than only resulting in heavier depreciation and cash flow pressure.

Goldman Sachs also expects Microsoft’s FY26 revenue to be $329.4 billion and EPS to be $16.75. FY27 revenue is expected at $387.1 billion and EPS at $19.32. The implied assumption behind this forecast is that AI investment can both drive revenue and not continually slow the pace of profit release.

Hyperscale cloud providers’ expected capital expenditure on the street for 2026/2027. Since January, AMZN, META, GOOGL, MSFT, and ORCL’s capex expectations have all been raised meaningfully. MSFT’s expected capex growth for 2027 reaches 55%.

Copilot needs to be charged; Maia should reduce reliance on GPUs

Whether Microsoft’s AI investment can “run” ultimately comes down to two levers: the commercialization of Copilot and the maturity of in-house and replacement chip supply.

Copilot’s logic is relatively straightforward. Long-term, rising usage benefits the expansion of software revenue and may also improve the profit structure. But in the short term, the issue is that usage itself does not equal revenue being realized.

Microsoft’s FY26 Q3 disclosures show that paid seats for M365 Copilot have surpassed 20 million. GitHub Copilot is also shifting toward more usage- and value-based pricing. The company has also introduced fair-use terms for high-usage scenarios, trying to bind higher inference costs and paid mechanisms more tightly together.

What the market will look at is not just continued growth in seats, but also user engagement, renewal willingness, and actual paid expansion on the enterprise side. If Copilot’s user experience and commercialization cadence can’t improve in sync, the timing for realizing high-margin AI software could be pushed out.

Chips and the supply chain are another track. Microsoft’s in-house AI chip Maia is still in a catch-up stage, lagging behind some peers in maturity. Improvements to Maia 300, AMD’s production progress as a second source, and memory procurement costs will all affect Microsoft’s ability to reduce reliance on external GPU supply chains.

The company has also previously mentioned that new supply needs to be balanced across Azure, first-party applications, R&D, and server replacements. If new supply ramps smoothly, Microsoft can deliver more compute to external Azure customers while continuing to invest in internal AI R&D. If the ramp is uneven, Azure growth, internal model training, and Copilot inference demand will still crowd each other out.

The Xbox reorganization is just a valuation side note

Beyond the AI main thesis, Goldman Sachs also used an SOTP approach to estimate the value of Microsoft’s gaming business at about $30 billion.

On July 6, Microsoft announced the reorganization of its Xbox business. Multiple media outlets reported that Microsoft laid off about 4,800 employees, including about 1,600 immediately eliminated in Xbox, with another approximately 3,200 within FY27. Four studios—Compulsion, Double Fine, Ninja Theory, and Undead Labs—left Xbox’s management structure, and the company also reportedly cut part of its management layer.

This part is more of a business-structure adjustment, not the main line of the earnings-report narrative. Microsoft’s gaming business still has value, and the reorganization shows the company is cleaning up low-efficiency assets and scaling back some non-core spending. However, in the short term it can’t replace the returns from Azure, Copilot, and AI capital expenditures, making it unlikely to be the main factor explaining the stock direction.

Based on Goldman Sachs’ SOTP valuation, Intelligent Cloud remains the largest contributor to Microsoft’s enterprise value. M365 commercial and consumer business implies an enterprise value of about $492 billion, corresponding to roughly 4x EV/sales or 6x GAAP EBIT for 2027. This includes some de-verticalization risk assumptions.

Whether $610 can be delivered depends on three things

The forward-looking direction in this earnings preview still leans optimistic: Microsoft is in a favorable position across AI compute, Copilot, and agent orchestration layers, and could continue to benefit from the AI product cycle. But whether the $610 price target can be achieved depends on whether the earnings report and the conference call provide more verifiable progress.

Azure needs to keep delivering high growth and explain whether new capacity coming online can sustain external customer demand. If growth only meets market expectations that are already high, higher capital expenditures could instead become a point of controversy.

Maia 300 and AMD as a second source need to provide clearer progress. Tight supply, rising memory costs, and insufficient chip maturity will all affect the unit economics of Microsoft’s AI investments.

Copilot also needs to prove true monetization capability. Over 20 million paid seats is just the starting point; enterprise-side paid expansion, usage-based billing, and user feedback will determine whether it can turn from an AI entry point into a profit source.

The key question for Microsoft’s earnings isn’t whether AI investment will continue, but whether higher investment can turn into faster Azure growth, AI software revenue, and sustainable profit margins. If this evidence is still insufficient, the debate over capital expenditure return rates will continue to weigh on the stock.

Click to learn about LyDing BlockBeats recruiting for positions

Welcome to join LyDing BlockBeats’ official community:

Telegram subscription group: https://t.me/theblockbeats

Telegram group chat: https://t.me/BlockBeats_App

Official Twitter account: https://twitter.com/BlockBeatsAsia