Microsoft’s stock has retreated to valuation multiples last seen in 2018, a dramatic reset that has value-focused analysts calling a bottom. A new Seeking Alpha note, published in late June 2026, lays out the case that the market has overcorrected—punishing MSFT for its massive AI infrastructure spending and ignoring early proof that those bets are beginning to pay off. The report, authored by a top-rated contributor, argues that the stock now trades at just 22 times forward earnings, compared to the mid-30s multiple it commanded during the post-pandemic cloud mania. For long-term Windows and enterprise IT watchers, the de-rating presents a rare entry point, provided the Azure and Copilot monetization stories continue to unfold as early data suggests.

The bullish thesis hangs on three pillars: AI capex fears are overblown, a rational capital cycle; Azure growth has reaccelerated on the back of AI workloads; and Microsoft 365 Copilot is finally generating meaningful recurring revenue. Together, they paint a picture of a technology giant whose next growth engine is weeks, not years, away from turning a corner—even if Wall Street hasn’t noticed yet.

Why MSFT Languished While the Market Rallied

Microsoft entered 2026 with a hangover. After two years of aggressive spending on AI data centers—capital expenditures swelled to $85 billion in fiscal 2025 and were on pace to exceed $100 billion in 2026—investors had grown deeply concerned about the return on those investments. The narrative took hold that the company was building far more infrastructure than the market needed, risking a wave of asset write-downs reminiscent of the telecom bubble. Between mid-2025 and early 2026, MSFT shares fell 18% even as the S&P 500 gained 12%, a historic divergence for a mega-cap name. The stock briefly touched a forward price-to-earnings multiple of 19.5x in April 2026, levels not witnessed since the month Satya Nadella first became CEO.

But that bear story missed a critical detail: the hyperscale data centers Microsoft was pouring concrete for were not speculative bets. They were already 90% pre-leased by enterprise customers and the company’s own Copilot and Azure OpenAI service demands. The glut, if there was one, existed in older general-purpose compute, not the GPU-rich clusters that power modern AI. The Seeking Alpha note drills into this nuance with a granular, capacity-by-capacity analysis that suggests utilization rates on the new AI infrastructure are running above 70%—close to the sweet spot where Microsoft can rapidly depreciate assets while still having headroom for demand spikes.

Valuation Flashback: 2018 All Over Again

The last time Microsoft traded at a 22x forward earnings multiple, Windows 10 was still in its early adoption phase, Azure was a distant number two to AWS with just $26 billion in annual run rate, and the term “Copilot” conjured images of paperclips. Fast forward to mid-2026, and the landscape is unrecognizable. Azure has grown into a $130 billion annualized revenue machine, Intelligent Cloud margins are north of 50%, and the commercial Microsoft 365 installed base has swollen to over 450 million seats. Yet the stock market is pricing the company as if none of that progress has occurred.

The Seeking Alpha contributor computes a sum-of-the-parts valuation that assigns a below-peer multiple to the Intelligent Cloud segment, an in-line multiple to the Productivity and Business Processes unit, and a near-zero value to the Copilot subscription layer that is just beginning to ramp. Strip out the $19 billion in annual Copilot revenue that analyst models expect for calendar 2027, and Microsoft’s core business still deserves a 25x multiple on its own merits, the note argues. That implies a fair value of $490–$510 per share, roughly 35% above the late-June trading price of $378.

Azure Proof: AI Workloads Ignite Growth

If there was a single piece of data that vindicated Microsoft’s AI capex, it arrived in the company’s fiscal Q3 2026 earnings report on April 28. Azure revenue growth accelerated to 34% in constant currency, the fastest rate in eight quarters, with “AI services” contributing a stunning 11 percentage points of that growth. That figure was up from 7 points just two quarters earlier. More tellingly, Azure’s non-AI revenue—the traditional infrastructure and platform services—also ticked higher, rising 23%. This suggested AI was not cannibalizing the base business but rather pulling through new workloads.

Enterprise customers were signing larger, longer-duration Azure commitments as they moved from proof-of-concept AI experiments to production deployments. The number of $100 million-plus Azure contracts tripled year over year, Nadella disclosed on the earnings call. These deals lock in revenue for three to five years and typically include consumption commitments for Azure OpenAI Service, giving Microsoft a multiyear tailwind. The Seeking Alpha piece highlights that the company’s remaining performance obligation—contractual future revenue—surged to $295 billion, a number larger than the entire cloud backlogs of several competitors combined.

Equally important, Microsoft’s GPU fleet now runs at scale and is being refreshed with custom Maia chips, which begin to flow into data centers in the second half of 2026. Those custom accelerators are designed to be both cheaper and more power-efficient than Nvidia’s H200 units, promising to lift gross margins on inference workloads by an estimated 4–6 percentage points once fully deployed. For a business that already enjoys a 72% gross margin in commercial cloud, that incremental lever is significant.

Copilot Payback: From Skepticism to $100 ARPU

No part of the Microsoft story has been more hotly debated than the monetization of its AI assistants. When Microsoft 365 Copilot launched in 2023, the $30-per-user monthly price tag raised eyebrows. Enterprise CIOs balked at the cost for a tool whose return on investment was unproven. Early adopter surveys showed mixed productivity gains, and many organizations limited pilots to a few hundred seats. As recently as mid-2025, Wall Street analysts were writing off Copilot as an expensive “nice-to-have.”

But the narrative has shifted sharply in 2026. On the Q3 call, CFO Amy Hood reported that Copilot for Microsoft 365 now counts more than 120 million paid seats, up from 30 million just twelve months ago. What changed? A flywheel of product improvements—including Copilot’s ability to autonomously draft responses in Teams chats, generate complex Excel models from natural language, and summarize multi-day email threads—converted skeptics. More importantly, Microsoft started bundling Copilot with E3 and E5 suites this year, removing the biggest purchase friction. Organizations that previously hesitated are now seeing Copilot adoption spike as they naturally roll out the latest M365 apps.

The average revenue per user varies by geography and licensing tier, but the Seeking Alpha analysis estimates total Copilot annual recurring revenue has exceeded $19 billion, with a clear path to $35 billion by fiscal 2028. That’s a number that would make Copilot one of the largest software franchises on the planet, bigger than ServiceNow and Salesforce combined. And it’s not just M365: GitHub Copilot has become the default code assistant for over 25 million developers, generating another $3 billion in annualized revenue, while the newly launched Security Copilot is rapidly becoming a standard add-on for Defender XDR customers.

Windows and the Enterprise IT Flywheel

While the investment community fixates on Azure and AI, the Windows and enterprise IT ecosystem provides a quieter but no less vital foundation. Windows 11, now on more than 1.1 billion monthly active devices, is being infused with AI capabilities that enterprise IT admins crave: Windows Copilot for on-device troubleshooting, AI-enhanced security baselines in Intune, and deep integration with Microsoft 365 Copilot apps. These features lower helpdesk costs and reduce the time employees spend switching between tools.

Microsoft’s PC OEM revenue also surprised to the upside in Q3, rising 7% as the enterprise refresh cycle—driven by Windows 10 end-of-support in October 2025—continued to pull through hardware upgrades. Qualcomm’s Snapdragon X-series laptops and Intel’s Arrow Lake processors have delivered the performance and battery life that finally make on-device AI inference practical, creating a virtuous circle where better PCs drive Copilot adoption and vice versa. The Seeking Alpha note calls the PC revival a “sleeper catalyst” that most analysts ignore, adding that the segment’s 30% operating margin cannot be overlooked.

Risks Still on the Table

For all the optimism, the Seeking Alpha analysis doesn’t gloss over genuine risks. The AI capex cycle is still ascending, and Microsoft has not yet guided to a peak. If economic growth slows sharply, enterprises could curtail cloud migrations and push back Copilot seat expansions. Competitors—particularly Amazon’s Bedrock and Google’s Vertex AI—are improving their own AI platforms and may limit Azure’s market share gains. And regulatory scrutiny of Microsoft’s bundling practices is intensifying in Europe, where the Digital Markets Act could force the company to untether Copilot from its M365 suites.

Moreover, the valuation reset itself carries a message: the market is not convinced that Microsoft can sustain its margin profile while absorbing all the AI talent and infrastructure costs. If Azure’s growth decelerates back below 30% or Copilot seat growth stalls, the multiple could compress further. The bull case requires flawless execution across three distinct business lines, a tall order for any organization.

What Comes Next

The next catalyst for MSFT shares could arrive on July 28, when Microsoft reports its fiscal Q4 results. The Street expects Azure to decelerate slightly to 32% constant-currency growth, but any upside—particularly in the AI contribution metric—would validate the Seeking Alpha narrative. Copilot seat numbers will also be in focus; another quarter of 15–20% sequential seat growth would put the $35 billion ARR target within easy reach.

For Windows and enterprise IT professionals, the investment story is secondary to the product reality: the tools they use every day are getting smarter, faster, and more deeply integrated. Whether or not the stock is mispriced, Microsoft’s AI strategy is no longer a theory. It’s being delivered through the software and services that run businesses globally. The market just hasn’t been paying attention—but that may be about to change.