Microsoft Turns 50, But Its True Growth Cycle Is Just Beginning—Goldman Sachs Re-Initiates Coverage: AI Is a "Cross-50-Year" Business

On January 11, 2026, Goldman Sachs resumed coverage of Microsoft, giving it a Buy rating and a 12-month target price of $655, implying 37% upside potentialcompared to the current stock price.
In Goldman Sachs’ view, even in the context of Microsoft’s 50-year history, the current moment remains a“discovery value” point in time.

There’s only one reason:AI.

This is not a short-term earnings-driven report, but astudy of Microsoft’s “structural AI compounding” from a 5–10 year perspective.

1. The Market’s Biggest Misunderstanding: Treating Microsoft's AI as Just Azure

The current market debate about Microsoft focuses on two issues:

• Huge Capex, but slowing Azure growth

• Can AI investments really translate into profits, or are they just a “burn rate competition”?

  
  

Goldman Sachs’ view is clear:

 You can’t understand Microsoft using the linear framework of “Capex → Azure growth.”

Microsoft’s AI investments are not a single-point bet, but aredistributed across four directions:

1. Azure AI (direct cloud monetization)

2. First-party applications (Copilot / 365)

3. Internal AI (Microsoft AI / MAI)

4. Maintenance computing power (legacy businesses)

What does this mean?

Between “short-term certainty” and “long-term moat,” Microsoft has proactively chosen the latter.

2. Microsoft’s True Core Advantage: Not the Model, But the “Three-in-One”

Goldman Sachs repeatedly emphasizes that Microsoft’s core competitiveness in the AI era is not “having the strongest model,” but rather:

Vertical integration of infrastructure (Infra) + platform (Platform) + application (App)

This advantage is further amplified in the AI era.

1️⃣ In the Infrastructure Layer (Infra)

• Microsoft isone of the world’s largest AI infrastructure providers in terms of computing power

• With AI demand, Azure is not just the cloud but aglobally distributed inference-first computing network

• Microsoft intentionally reduces reliance on “training supercomputers,” shifting toward higher ROI inference workloads

2️⃣ In the Platform Layer

• Foundry: allows enterprises to flexibly call upon different models such as OpenAI, Anthropic, and MAI on Azure

• Essentially:the “control plane” for AI

Microsoft is not betting on a single model to win but isbetting on the long-term coexistence of multiple models.

3️⃣ In the Application Layer

• More than 400 million Microsoft 365 users

• Copilot is not just an add-on, it isa new work entry point

• Every 10% Copilot penetrationcan bring about 10% incremental growthto the Productivity & Business Processes (PBP) segment

 This is the most certain monetization path in AI commercialization.

3. Why is Microsoft’s AI Investment “Downside Risk Limited”?

This is a very key point in this report.

Goldman Sachs offers a very “financialized” explanation:

Microsoft is maximizing the “Sharpe Ratio” of its AI investments

What does this mean in practice?

• Upside leverage

• Holds about a 27% economic interest in OpenAI

• Copilot, Azure, and Agent 365 all benefit simultaneously

• Downside hedges

• In-house development of MAI reduces dependence on a single model

• Multi-GPU / multi-chip / in-house silicon

• Data centers are highly “reusable”

This means even if the AI industry path deviates,
it ishard for Microsoft to become a loser.

4. Profit Margin Is the Biggest Variable in the Medium to Long Term

Another market concern is:

Will AI permanently drag down cloud computing profit margins?

Goldman Sachs’ answer is no.

Currently, Azure AI’s gross margin is about 30%, significantly lower than traditional Azure’s 60%+.
But Microsoft management’s goal is very clear:

Within 5–7 years, return AI cloud service gross margins to historical highs

The implementation path includes:

• Increase in proportion of in-house chips (potentially >50%)

• Continual improvement in inference efficiency

• Maintaining GPU useful life at 6 years

• Algorithmic efficiency improvements (not just unlimited scaling of computing power)

Goldman Sachs believes:
The gross margin curve for AI cloud is “down first, then up,” rather than structurally deteriorating.

5. The Financial Conclusion: Why Goldman Sachs Dares to Give $655

In Goldman Sachs’ optimistic scenario:

• FY2030 EPS could exceed $35

• Compared with the current trillion-dollar market cap companies, which havemedian single- to mid-teen growth rates

• Microsoft can still maintain a20%+ long-term EPS compound growth rate

This is the core logic behind Goldman Sachs’ $655 target price.

My Understanding:

Goldman Sachs is not packaging Microsoft as “the next OpenAI,”
but instead repeatedly stresses its key characteristics:

• Systematic

• Patience

• Cross-cycle

  
  

Microsoft is not in a hurry to “prove whether AI investment is worth it” in a particular year,
what it wants to do is—

make AI the underlying operating system for enterprise software for the next 50 years.

From this perspective,
this is not a stock that needs to be traded based on single-quarter Azure growth,
but rather aclassic asset that accumulates value over time.