Microsoft Stock Is Down 30% From Its Peak — Is This the Buying Opportunity of the Decade or a Broken Story?

Microsoft’s engineers and product teams are hard at work building data centers, training AI models, and integrating Copilot features into software that hundreds of millions of people use on a daily basis on the expansive campus in Redmond, Washington, where the company’s buildings are connected by a network of peaceful walking paths and the Cascade foothills are visible on clear days. That work hasn’t slowed down at all. The pace has actually accelerated by the majority of operational measures. Nevertheless, Microsoft’s stock has experienced its worst decline since the 2008 financial crisis, losing about 30% of its value from its peak of $555.45 to close at $372 on Tuesday. Strong opinions are typically generated on both sides when there is a discrepancy between a company’s stated goals and its stock price.

It is fairly simple to make the argument against Microsoft stock in 2026. As the company builds out data centers and custom AI chip infrastructure at a scale that carries no guarantee of commensurate near-term revenue, capital expenditures are rising sharply. Although Azure’s cloud growth has been steady, it has been only marginally slower than Wall Street’s most optimistic projections.

NASDAQ: MSFT  ·  Technology / Cloud / AI  ·  Redmond, Washington

microsoft.com/investor ↗

This seems to be sufficient to draw persistent selling pressure in a market that has very little tolerance for “almost.” And then there is the OpenAI partnership, which has grown to be a real source of concern for investors. Microsoft has invested a significant amount of resources in a partnership with an unprofitable company, and a significant amount of the cloud demand backlog is linked to that one partnership in ways that are challenging to fully understand from the outside.

The contrast with Meta Platforms, which led to similarly high infrastructure spending around the same earnings cycle and was rewarded rather than penalized, is what makes the sell-off intellectually fascinating. Both businesses are investing heavily in AI infrastructure. Both are wagering that the short-term margin pressure is justified by the buildout.

However, Meta showed how its AI investments are already enhancing advertising monetization across its current platforms in a clear, succinct manner in the language Wall Street prefers. Microsoft’s narrative focused more on laying the groundwork than on demonstrating the profits. The stock prices almost instantly reflected the market’s conclusion that one story was convincing and the other was lacking. This might be a reasonable distinction. In two or three years, it might appear embarrassing if it’s a rather myopic one.

The Copilot criticism is genuinely conflicting and should be carefully read. Consumer adoption rates have lagged behind the initial wave of enthusiasm, and competitors’ frequent releases of new large language models have created legitimate optionality that wasn’t present when Microsoft first integrated OpenAI’s technology into its products.

That’s all true. However, the Copilot bear case mainly relies on consumer behavior patterns that don’t really explain how enterprise software functions. Based on quarterly comparisons of AI benchmarks, Fortune 500 companies do not change their productivity suites or ERP systems. They execute procurement procedures that take months, they sign multi-year contracts, and once a platform is integrated into a business’s workflows—managing contracts, analyzing financial data, and coordinating supply chains, for example—it doesn’t simply disappear. In just one year, the number of business clients with more than 35,000 Copilot seats is said to have tripled. A failing product does not follow that course. It is quite different from the trajectory of a product with a lengthy sales cycle.

By any measure, Microsoft’s Q2 FY2026 results were impressive. Revenue exceeded analyst projections with $81.27 billion, up 16.7% year over year. Additionally, earnings per share improved. Over the past ten years, the company’s revenue has increased by about 300%. This growth includes several economic cycles, a pandemic, and the full cloud transition that Satya Nadella oversaw from the time he took over as CEO in 2014. For Microsoft, the forward P/E ratio is currently at its lowest point in about ten years, at about 22 times. Ten years ago, the company was regarded as a mature, slow-growing legacy software company that traded at modest multiples because it appeared to have outlived its best days. One of the most amazing corporate reinventions in the history of technology ensued.

Observing this specific sell-off gives the impression that the market is holding Microsoft to a standard that it hasn’t fully expressed and isn’t applying uniformly throughout the industry. The infrastructure expenditures that investors are currently penalizing are similar to those made by AWS during the 2010s, when the company was losing money every quarter but developing what would eventually become a key component of international trade.

Microsoft’s balance sheet can withstand multi-year buildouts without experiencing financial strain, its gross margins stay at roughly 69%, and its enterprise software division consistently produces cash throughout all economic cycles. The Iran ceasefire and the ensuing rate relief were major factors in Wednesday morning’s 3.25% pre-market rally, which demonstrated how quickly sentiment can change when a macro headwind shifts.

When the market will reconcile Microsoft’s share price with its operational reality is still up in the air. That kind of reconnection can take longer than seems reasonable, as the patient investors who purchased Apple during its numerous corrections or Amazon during its years of thin margins are aware. The stock of Microsoft is not damaged. It is just being asked the wrong question at the moment.