By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
The panel is bearish on Microsoft due to near-term margin compression from a $190B capex commitment and dependency on OpenAI, which could shift workloads or face regulatory pressure. The consensus is that the stock will likely remain range-bound until operating leverage returns to the cloud segment.
Risk: Near-term margin compression from $190B capex and OpenAI dependency
Opportunity: Sovereign AI pivot and locking in enterprise Azure spend via long-term, high-margin compliance contracts
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Microsoft is the worst-performing "Magnificent Seven" stock in 2026, down about 13% year to date.
Revenue growth accelerated in the company's fiscal third quarter, and its annual AI business revenue run rate more than doubled.
The software and cloud giant is now one of the cheapest stocks in the group, but its spending plans keep climbing.
The "Magnificent Seven" began 2026 in a hole. Every member slid in the year's first few months as investors started questioning how much they had been paying for promises tied to artificial intelligence (AI). Since then, however, most of the group has climbed back. As of this writing, the seven are collectively higher on the year, the S&P 500 is up more than 8%, and Alphabet has jumped more than 20%.
But one name has been left out of the rebound. Microsoft (NASDAQ: MSFT) is down about 13% so far in 2026, the worst showing in the group. Even Tesla, which had been vying with Microsoft for last place earlier in the year, has since pulled ahead -- as have chipmaker Nvidia, fresh off another strong quarter, and iPhone maker Apple.
Will AI create the world's first trillionaire? Our team just released a report on the one little-known company, called an "Indispensable Monopoly" providing the critical technology Nvidia and Intel both need. Continue »
What makes the gap unusual is that the business itself hasn't stumbled. So, with the stock sitting at the back of the pack, is the software and cloud giant's stock finally worth buying?
Microsoft's fiscal third quarter of 2026 (the period ended March 31, 2026) didn't look like a company in trouble. Revenue rose 18% year over year to $82.9 billion -- an acceleration from 17% growth the prior quarter, and operating income climbed 20% to $38.4 billion. Further, the software giant's non-GAAP (adjusted) earnings per share rose about 21%.
Even more, Microsoft said its AI products now carry an annual revenue run rate of more than $37 billion -- up 123% from a year earlier. Behind it is everything from outside developers building on Azure, the company's cloud computing business, to Microsoft's own Copilot assistant, which crossed 20 million paid seats after adding 5 million in a single quarter.
Management also signaled a change in how it plans to charge for all of this.
"Any per-user business of ours, whether it's productivity, coding, security, will become a per-user and usage business," CEO Satya Nadella said during the company's fiscal third-quarter earnings call. In plain terms, Microsoft wants to keep collecting its familiar per-seat fees while adding charges based on how much customers actually lean on its AI tools -- a model it's already rolling out, starting with usage-based pricing for its GitHub coding assistant.
And don't forget that Microsoft has a roughly 27% stake in OpenAI and a non-exclusive license to its technology through 2032.
Further, after the stock's slide, Microsoft trades at a forward price-to-earnings ratio of about 22 as of this writing -- among the lowest in the "Magnificent Seven."
It also offers the group's most generous dividend yield, though it's still modest at 0.9%.
For a business compounding at these rates, this doesn't look like a demanding price.
The catch is what the growth now costs. Microsoft expects to spend roughly $190 billion on capital expenditures in calendar 2026, up about 61% from the prior year, as it races to build data centers.
That spending is starting to show.
Microsoft's gross margin in its fiscal third quarter was down year over year as depreciation from those data centers piled up, and management expects to stay capacity-constrained at least through 2026. Additionally, if demand for AI computing cools before the build-out pays off, the margin pressure could worsen before it improves.
And the OpenAI relationship cuts both ways. A single partner still anchors a large share of Microsoft's contracted commercial work. But that means the company's backlog could suffer if OpenAI's business unexpectedly declines or if the AI company shifts more of its future business to other cloud providers.
Even so, the discount may be doing too much work. Microsoft is among the cheapest members of the group, yet it's growing faster than it was a year ago, with a credible path to charging more as AI usage climbs. The heavy spending, of course, is a risk worth watching. And investors who can't stomach a long, costly build-out may prefer to wait for clearer evidence that it's paying off. But for those willing to look past a year of underperformance, the market's least-loved "Magnificent Seven" stock could finally be worth a fresh look, but ideally only as a small position given the high risks of the company's costly AI build-out.
Before you buy stock in Microsoft, consider this:
The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Microsoft wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you’d have $477,813! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,320,088!
Now, it’s worth noting Stock Advisor’s total average return is 986% — a market-crushing outperformance compared to 208% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors.
**Stock Advisor returns as of May 23, 2026. *
Daniel Sparks and his clients have positions in Apple. Daniel has clients with positions in Tesla. The Motley Fool has positions in and recommends Alphabet, Apple, Microsoft, Nvidia, and Tesla. The Motley Fool has a disclosure policy.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
Four leading AI models discuss this article
"Microsoft's accelerating AI revenue is real but insufficient to offset the margin and concentration risks from its outsized 2026 capex plan."
Microsoft's 18% revenue acceleration and $37B AI run rate look strong on paper, yet the $190B 2026 capex commitment (up 61% YoY) is already pressuring gross margins via depreciation while capacity stays constrained through year-end. Forward P/E of 22x is the group's lowest, but that discount embeds execution risk on usage-based pricing and a 27% OpenAI stake that could swing either way. Investors buying the dip overlook that AI spend must deliver sustained margin expansion before the stock re-rates higher; otherwise, 2026 becomes a multi-year drag rather than a setup.
The article underplays how quickly hyperscaler capex cycles have reversed in prior tech waves; if enterprise AI adoption slows, the $190B build-out could leave Microsoft with stranded assets and permanently lower returns on capital.
"MSFT is cheap on growth metrics but expensive on execution risk—the $190B capex spend is a bet that AI ROI materializes before competitive pressure or demand destruction forces a writedown."
Microsoft's 13% YTD underperformance masks a business accelerating into 18% revenue growth and 123% AI revenue run-rate expansion. At 22x forward P/E against 21% EPS growth, the valuation discount is real—but the article buries the actual problem: $190B capex (61% YoY increase) is compressing gross margins NOW, not hypothetically. The OpenAI dependency is also understated; if OpenAI shifts workloads to competitors or faces regulatory pressure, Microsoft's backlog evaporates. The 'small position' caveat in the conclusion is doing heavy lifting—it's admission that execution risk is material.
If AI capex payoff accelerates faster than expected and usage-based pricing sticks (Copilot already at 20M paid seats), MSFT could re-rate to 25-27x within 12 months, making today's 22x look cheap in hindsight; the margin compression may be temporary.
"Microsoft's transition to a capital-intensive infrastructure model is permanently resetting its valuation floor, making the historical P/E comparison misleading."
Microsoft’s 13% drawdown in 2026, despite 18% revenue growth, signals a fundamental shift in market sentiment regarding capital intensity. A forward P/E of 22x is historically attractive for a company of this quality, but the $190 billion capex guidance is the real story. We are seeing a transition from 'software-as-a-service' (SaaS) margins to 'infrastructure-as-a-service' (IaaS) capital burdens. While the usage-based pricing model is a brilliant lever to capture AI value, it introduces volatility into cash flow projections. I am neutral because the market is rightly punishing the margin compression; until we see operating leverage return to the cloud segment, the stock will likely remain range-bound.
The bearish case is that Microsoft is essentially becoming a utility company with massive, non-discretionary infrastructure costs that will permanently erode the high-margin software business model investors once prized.
"MSFT's AI-driven growth hinges on expensive data-center expansion and a shaky mix-risk monetization; without clear margin recovery and durable AI RPM, a 22x forward multiple may re-rate lower as earnings quality proves more volatile than the headline AI growth suggests."
Microsoft is the standout in the Magnificent Seven for being cheap and still growing, but the core risk is the cost of the AI build-out. The 190B capex in 2026 (+61%) and data-center depreciation imply near-term gross margin pressure, while AI backlogs and OpenAI exposure create a dependency risk if demand or licensing shifts. The article glosses over monetization: a shift to per-user plus usage pricing could compress ARPU if usage surges but seat licenses lag; a slowdown in AI computing demand or a competitive AWS/Google response could cap the earnings acceleration. A small position is prudent until margins stabilize and AI RPM (revenue per unit of AI usage) proves durable.
The bull case is that Microsoft remains the AI platform winner with Azure, copious spillovers from OpenAI, and capital efficiency will recover; capex drives scale and cost per unit falls; this should push margins higher once utilization peaks.
"OpenAI regulatory or structural shifts would leave Microsoft with stranded $190B assets and no offsetting economics."
Claude correctly flags OpenAI dependency risk, yet the $190B capex already committed creates a harder problem: even if OpenAI shifts workloads, the depreciation and capacity lock-in persist through 2026 regardless. That turns the 27% stake from optional upside into a correlated liability. If regulators later force partial divestiture, Microsoft loses both the AI moat and any offsetting economics while still carrying the full infrastructure burden.
"Microsoft's capex thesis collapses if OpenAI regulatory risk forces divestiture, not because of stranded assets alone, but because the partnership itself justifies the infrastructure spend."
Grok's divestiture scenario is plausible but underweights Microsoft's optionality. The $190B capex locks in depreciation, yes—but if OpenAI faces regulatory pressure and Microsoft must divest, the real damage isn't the stranded assets; it's losing the *exclusive* training and inference partnership that justifies the capex in the first place. That's the correlation nobody's quantified: capex value and OpenAI stake are inseparable. Lose one, the other becomes a sunk cost.
"Microsoft's massive capex is a strategic play for sovereign, localized cloud dominance that renders the OpenAI dependency risk secondary to long-term enterprise lock-in."
Claude and Grok are fixated on the OpenAI partnership as a binary risk, but you are all ignoring the 'sovereign AI' pivot. Microsoft is aggressively selling private, localized cloud instances to governments and regulated industries to bypass the latency and data-privacy issues of centralized OpenAI models. This isn't just about OpenAI; it’s about locking in enterprise Azure spend via long-term, high-margin compliance contracts. The capex isn't just for OpenAI—it's for building a proprietary, moat-protected, distributed infrastructure.
"The sovereign-AI pivot won’t cure near-term margin compression; public-sector deals add visibility but slow cycles and compliance costs leave capex drag and OpenAI exposure intact."
Gemini's sovereign-AI pivot is a plausible tailwind, but it doesn't solve the near-term margin compression from $190B capex and depreciation. Public-sector deals bring revenue visibility but slower procurement cycles and strict compliance costs can flatten cash flow upside. If governments delay contracts or impose localization rules, OpenAI dependency and capex drag remain. The real test is operating leverage in cloud, not just new customer adds.
The panel is bearish on Microsoft due to near-term margin compression from a $190B capex commitment and dependency on OpenAI, which could shift workloads or face regulatory pressure. The consensus is that the stock will likely remain range-bound until operating leverage returns to the cloud segment.
Sovereign AI pivot and locking in enterprise Azure spend via long-term, high-margin compliance contracts
Near-term margin compression from $190B capex and OpenAI dependency