Microsoft shares are down about a third from their late‑October peak. Investors who think the selloff is over are pointing to valuation and AI traction as the reason to buy.
Why one Wall Street analyst is bullish
Focus on the numbers. Bank of America analyst Tal Liani restarted coverage on Microsoft with a buy rating and a $500 price target, arguing the pullback offers a rare buying opportunity for long‑term investors. Liani notes Microsoft sits in a unique spot: it's both the provider of the cloud infrastructure that powers AI and the maker of the software that customers use for daily work. "At the center of the AI supercycle," he said, and "a primary beneficiary of AI monetization," phrases that sum up his thesis.
Microsoft reported fiscal second‑quarter revenue of $81.3 billion and adjusted net income of $30.9 billion, or $4.14 a share — numbers many investors point to when arguing the core business still grows even as AI investments heat up. By GAAP earnings, the stock trades at about a 23 P/E — analysts say that's the lowest it's been in roughly ten years. That cheaper multiple is the core of Liani's argument: downside looks limited if the company continues to grow revenue and monetizes AI over time.
Liani argues Azure could be the backbone for AI while Microsoft’s enterprise apps drive demand for that compute, creating a mutual lift. If customers rely on Azure for compute and then pay for Microsoft applications that embed AI, both parts of the business can lift each other — and that feedback loop is what the Bank of America note calls an "ideal outcome."
Headwinds investors cite
Not everyone agrees the bottom is in.
Ben Reitzes, analyst at Melius Research, catalogued problems he sees at Microsoft that help explain recent selling pressure. Reitzes said Microsoft’s Copilot efforts have disappointed and that a reorganization announced in March "doesn't seem like it was done in strength." He also flagged tension in the company’s relationship with OpenAI, calling OpenAI "the most important partner it has ever had."
Reitzes pointed to a practical cost: OpenAI accounts for roughly 45 percent of the Azure backlog, he wrote, and Microsoft was expecting shared intellectual property from that partnership to accelerate its AI progress. That shortcut hasn't materialized in the way investors hoped, he argued, forcing Microsoft to spend more on research and development and to use Azure computing capacity internally to train its own models.
Azure capacity is finite. The company has said customer demand has at times outpaced available supply, and the trade‑off is clear — compute used for internal model training can't be simultaneously sold to customers. That dynamic showed up in Microsoft's fiscal second‑quarter cloud growth slowing slightly: Azure and other cloud services revenue grew 39 percent year‑over‑year, down from 40 percent the prior quarter. Management guided to about 37 percent growth for the next quarter, which stoked questions about whether AI demand will translate into steadier, long‑term cloud revenue growth.
Valuation versus execution
Investors are split: the stock looks cheaper today, but doubts about execution keep many on the sidelines. The valuation argument is straightforward: the stock has fallen materially from its peak, and at current multiples it looks cheaper than it has in years. That’s why Liani and other bulls believe the downside is limited.
Execution problems are messier and harder to quantify than a single valuation multiple. Critics say Microsoft has spent heavily on AI and has yet to demonstrate consistent monetization of those investments. The Copilot rollout, in particular, has drawn criticism for failing to meet expectations, and anecdotal reports of companies building bespoke AI tools rather than licensing enterprise software add to the uncertainty. Layoffs and hiring freezes at some tech customers don't help either — they reduce near‑term cloud demand.
History matters here: big tech firms have been punished even while sales climbed, when future growth expectations slipped. Large tech firms have weathered periods where investors punished shares even as revenue grew, often because market prices were tied tightly to future growth expectations. Microsoft is no stranger to that cycle — it's steered through cloud transitions before — but AI brings a different kind of capital intensity and competitive pressure. If Microsoft monetizes AI tools inside Office, GitHub, and other apps while keeping Azure capacity working for customers, that could validate Liani's view. If not, the stock could languish.
What investors are watching next
Investors will pore over the next quarterly report for signs of Azure demand and AI monetization. Management has already signalled that customer demand for Azure outstripped supply at times, an operational detail that matters in a business where margins and growth depend on selling compute capacity. The outlook for the current quarter — the one management said was expected to show roughly 37 percent cloud growth — will be one of the clearest short‑term tests of the bull case.
Other near‑term indicators include how Microsoft describes its OpenAI partnership going forward and what it says about internal versus customer compute usage. Any sign the partnership is moving toward clearer IP sharing or that Microsoft is easing internal demand on Azure would be read as positive. Conversely, further disclosures that Microsoft must materially divert Azure capacity to internal model training could reinforce the bears' argument that AI spending is crowding out revenue.
Risks and the path to upside
The risks are obvious: Copilot missteps, Azure scaling problems, or failure to monetize AI would all keep the shares under pressure. Continued execution issues with Copilot, a drawn‑out scaling problem for Azure, or a failure to monetize AI across Microsoft’s software stack would keep pressure on the share price. Melius Research's Reitzes points specifically to execution and partnership frictions as reasons the stock can't sustain momentum right now.
But the upside scenario is straightforward too: if Microsoft starts turning AI features into paid products at scale, and if Azure remains the default cloud for AI workloads, revenue and margins could reaccelerate. That's the case Liani lays out — one where being both the infrastructure provider and the software vendor pays off in the form of recurring revenue tied to AI workloads. Bank of America’s $500 target implies roughly 34 percent upside from current levels at the time of the note.
Bottom line for investors
Investors who believe the selloff has overshot the fundamentals point to current earnings strength and a lower valuation. Those who worry the AI transition is costing the company soon focus on Copilot’s rollout, higher R&D and compute usage, and strained ties with key partners like OpenAI. The debate is fundamentally about timeframes: how quickly Microsoft can turn AI investment into steady, monetizable growth.
For now, the market is split — and that split is what's setting up the current trade: cheap valuation on one side, execution risk on the other.
Related Articles
Tal Liani, Bank of America analyst, said Microsoft is "at the center of the AI supercycle" and stands to be "a primary beneficiary of AI monetization."
This article was created with AI assistance.
