![\"Microsoft](\"https:\/\/www.europesays.com\/ai\/wp-content\/uploads\/2026\/08\/microsoft-chairman-chief-executive-officer-satya-nadella.jpg\")
<\/p>\n<\/p>\n
Microsoft Chairman and Chief Executive Officer Satya Nadella speaks during the Microsoft Build conference opening keynote in Seattle, Washington on May 19, 2025.
\nJASON REDMOND\/AFP via Getty Images<\/p>\n
Microsoft’s stock rose roughly 25% in the week following its July 29 fiscal fourth-quarter earnings release, carrying shares from around $389 to approximately $500 \u2014 a post-earnings surge that analysts at 24\/7 Wall St. described as one of the most dramatic<\/a> in recent mega-cap history. That move is already history. The question for investors evaluating MSFT today, eleven days later, is a different one: what does buying at $500 actually require you to believe, and has a development reported this morning changed any of those beliefs?<\/p>\n The short answer is yes. The Information reported Monday morning that Microsoft is targeting a September 2026 public unveiling of its next-generation Maia 300 AI accelerator and has entered talks with TSMC to secure manufacturing capacity for more than 300,000 units for delivery in 2027. That disclosure reframes what Microsoft’s capital expenditure guidance for fiscal year 2027 actually funds \u2014 and it is the kind of new variable that prior earnings coverage could not have incorporated.<\/p>\n What the Record Earnings Actually Delivered<\/p>\n Microsoft’s fiscal fourth quarter ended June 30, 2026, produced results that cleared analyst expectations across every major line. Revenue reached $90.0 billion<\/a>, up 18% year-over-year against a consensus of roughly $87.6 billion. Adjusted earnings per share came in at $4.74 versus the $4.24 estimate, an 11.8% beat. Net income surged 31% to $35.8 billion.<\/p>\n Azure, the metric the market was actually watching, grew 43% year-over-year, accelerating from 40% in the prior quarter and clearing the informal floor of roughly 36% that analysts had identified as the level below which a sell-off would be near-certain. For fiscal year 2026 in full, Azure crossed $100 billion<\/a> in annual revenue for the first time, growing 41% over the year. That places Azure behind only Amazon Web Services globally, ahead of Alphabet’s Google Cloud.<\/p>\n Microsoft 365 Copilot added roughly 10 million paid seats in a single quarter, bringing the total to more than 30 million. This matters specifically because TechTimes documented the paid-adoption crisis in early July \u2014 fewer than 4.5% of Microsoft’s 450 million commercial Microsoft 365 users had then converted to paid Copilot \u2014 and the 30 million figure represents a direct answer to that concern. One caveat that prior coverage has established remains valid: seat count is not the same as recognized revenue. Enterprise Copilot deals have involved significant discounts in competitive displacement scenarios, and Microsoft does not yet disclose average revenue per seat or weekly active use rates alongside its seat count.<\/p>\n The full-year scorecard reinforced the trend: revenue of $331.8 billion, up 18%; operating income of $155.2 billion, up 21%; net income of $133.7 billion, up 31%. For fiscal year 2027, CFO Amy Hood guided Azure growth of approximately 45% in constant currency for the first quarter \u2014 above the StreetAccount consensus of 41.4% and above even the just-delivered 43%.<\/p>\n Why the $678 Billion Backlog Is the Bull Case’s Foundation<\/p>\n The single most consequential number in Wednesday’s report for investors evaluating MSFT today is not the Azure growth rate. It is the commercial remaining performance obligations figure: $678 billion, up 84% year-over-year<\/a>. This represents contracted but not yet recognized revenue \u2014 enterprise customers who have signed agreements for future Azure and AI services, committing payment before Microsoft has delivered the capacity to serve them.<\/p>\n That $678 billion exceeds Microsoft’s entire fiscal year 2026 revenue of $331.8 billion by roughly two to one. When a company’s signed future contracts are twice its current annual revenue, the conventional concern about whether demand exists to justify infrastructure spending is structurally answered: the demand exists, it is committed, and it is under contract. What remains unanswered is the conversion timeline \u2014 at what rate that backlog turns into recognized quarterly revenue \u2014 and the concentration question. Prior reporting established that OpenAI alone had accounted for approximately 45% of Microsoft’s contracted Azure backlog figure<\/a>. Microsoft disclosed in its fiscal year 2026 Form 10-K that it generated $24.1 billion in revenue from commercial arrangements with OpenAI and held $6 billion in accounts receivable from OpenAI as of June 30, 2026. An investor treating the $678 billion as a simple floor for future revenue needs to understand that OpenAI’s own commercial trajectory \u2014 now complicated by its April 2026 partnership restructuring, which ended Azure exclusivity and let OpenAI sell models through AWS \u2014 is a meaningful uncertainty embedded in that number.<\/p>\n Hood’s response to the free cash flow concern is precisely this backlog. Capital expenditure for Q4 reached $41 billion including finance leases \u2014 more than double the $17 billion spent in the same quarter a year earlier \u2014 while free cash flow came in at approximately $19.6 billion after capital expenditure. In a single quarter, capex including finance leases exceeded free cash flow for the first time in the company’s recent history. For calendar year 2026, Microsoft’s capital expenditure guidance (including the effect of a lease reclassification resulting from extending estimated useful lives of data center buildings from 15 to 25 years) stands at approximately $175 billion, reduced from a prior guidance of $190 billion due to the accounting reclassification alone; the underlying physical investment is unchanged. Analysts have estimated fiscal year 2027 capital expenditure in the range of $255 billion to $260 billion based on Hood’s confirmation that spending will grow year-over-year from the calendar 2026 level, with Q1 FY2027 capex expected to exceed $50 billion.<\/p>\n Hood told analysts that the company expects to remain free cash flow positive in fiscal year 2027. That commitment is a directional assurance, not a specific dollar figure.<\/p>\n Maia 300: Today’s Report Changes the Capex Story<\/p>\n The capex debate changed materially this morning. The Information reported that Microsoft is targeting a September 2026 public unveiling of its Maia 300 AI accelerator and has entered talks with TSMC<\/a> to secure manufacturing capacity for more than 300,000 units \u2014 a figure that dwarfs the tens of thousands of Maia 200 chips produced to date. The company’s longer-term ambition is capacity for more than one million units, though component supplies and packaging negotiations could constrain that target.<\/p>\n Understanding why this matters requires understanding the economics of AI inference at Microsoft’s scale. The current generation Maia 200, built on TSMC’s 3-nanometer process and delivering approximately 10 petaflops of FP4 compute performance, is already in production at two U.S. data centers (Iowa and Arizona). The MAI software models running on that hardware have demonstrated specific cost reductions against Nvidia alternatives: MAI-Image-2.5-Pro reduces GPU costs by up to 84% compared with OpenAI’s GPT-Image-2 for image generation in PowerPoint; MAI-Voice-2-Flash delivers GPU cost reductions of up to 89% versus the OpenAI model it replaced in Dynamics 365 Contact Center. If Maia 300 scales those economics to 300,000 or more units, the inference cost math changes<\/a> for Azure’s entire AI workload changes structurally \u2014 and that changes the return on the capital being committed.<\/p>\n The competitive gap is real. Google produces roughly 3 million Tensor Processing Units annually and began recognizing revenue from direct TPU sales to external customers in the quarter ended June 2026. Amazon’s Trainium and Inferentia chips are deployed at AWS scale. Microsoft is the last major hyperscaler to reach commercial-scale custom silicon. The Maia 300 program, if it delivers on its timeline, is Microsoft’s bid to close that gap \u2014 but J.P. Morgan analysts have flagged that projects concentrated on TSMC’s N3 process and CoWoS advanced chip-packaging technology face supply tightness through 2027<\/a>, a constraint directly relevant to the 300,000-unit target.<\/p>\n One additional complication from the same report: Anthropic, which Microsoft has identified as a potential Maia 300 customer, confirmed earlier this month that it is forming its own custom silicon team<\/a> to design chips for its Claude models. That makes Anthropic simultaneously a potential buyer of Maia 300 and an emerging long-term competitor in custom AI silicon \u2014 the kind of dynamic that is hard to price from outside.<\/p>\n Analyst Targets After a Rally That Already Happened<\/p>\n The relevant analyst estimates to understand are not the buy ratings \u2014 virtually every major firm rates MSFT a buy. The relevant figure is the gap between where the stock is now and where analysts see fair value.<\/p>\n Morningstar analyst Dan Romanoff maintained a $600 fair value estimate and a 5-star rating in the period leading up to earnings, when the stock was significantly below that level. Following the post-earnings rally to roughly $500, Morningstar noted the stock has moved into a range it considers fairly valued. Morningstar assigns Microsoft a wide economic moat, anchored primarily by switching costs \u2014 the friction that keeps enterprise customers from migrating off Azure, Microsoft 365, and GitHub \u2014 with network effects and cost advantages as secondary contributors.<\/p>\n Tigress Financial analyst Ivan Feinseth raised his price target to $690<\/a>, citing Azure’s 43% quarterly growth, the $59.3 billion in Microsoft Cloud revenue for the quarter, the 30 million Copilot paid seats, and Microsoft’s proprietary Maia and Cobalt chips as tools for improving efficiency. At the current price near $500, that target implies approximately 38% upside.<\/p>\n The 24\/7 Wall St. team, writing after the rally when MSFT was around $487, set a $573.79 price target<\/a> with a buy rating at 90% confidence, noting the commercial backlog as the key differentiator against peers. GuruFocus’s GF Value model placed fair value at $575 as of August 10, suggesting the stock is approximately 13% undervalued at $500.<\/p>\n On valuation multiples: Microsoft currently trades at approximately 28x trailing earnings and roughly 26x forward earnings. That compares with Alphabet at approximately 17.4x forward earnings and Amazon at roughly 35x. The premium Microsoft commands over Alphabet reflects either a judgment that Azure’s growth profile justifies it or \u2014 as bears would argue \u2014 AI optimism that has run ahead of demonstrated margin expansion.<\/p>\n Short interest remains minimal, indicating institutional skepticism is limited. The more significant institutional signal is the insider activity: over the past three months, insiders have sold approximately $17.8 million worth of shares with no offsetting insider purchases. Separately, hedge fund TCI, led by Sir Christopher Hohn, sold nearly all of its approximately $8 billion MSFT stake by end of March 2026, citing concerns that rapid AI advancement could undermine Microsoft’s dominance in productivity software and create risk for Office as AI-driven workflows and competing platforms emerge. TCI’s exit is the most specific institutional bear case on record for this period \u2014 worth weighing against the analyst consensus, even as that consensus remains overwhelmingly bullish.<\/p>\n There is also an active securities class action with a lead-plaintiff deadline of August 11, 2026, covering investors who purchased MSFT between May 1, 2025, and January 28, 2026. This does not directly affect the investment thesis but represents a legal overhang that informed investors should monitor.<\/p>\n Three Conditions the $500 Price Requires<\/p>\n Buying Microsoft at $500 today is not the same trade as buying it at $389 before earnings. The record single-day market cap addition of approximately $450 billion \u2014 the largest in stock market history<\/a>, surpassing Nvidia’s prior record of $441 billion set in April 2025 \u2014 has already happened. Three specific conditions now need to hold for the current price to prove justified rather than borrowed.<\/p>\n The first is Azure growth durability. Hood guided 45% growth for fiscal Q1 2027. As the annual revenue base grows past $100 billion, each percentage point of growth requires more absolute revenue than it did at a smaller scale. Microsoft’s commercial backlog gives the growth a contractual foundation, but investors will watch the Q1 FY2027 print \u2014 expected around October 2026 \u2014 for the first evidence of whether the guided 45% is achievable on a $100 billion-plus base.<\/p>\n The second is Copilot revenue conversion. The 30 million paid seats are a compelling adoption signal. But the question that seats alone cannot answer is what revenue per seat looks like at scale. Microsoft has offered steep discounts in enterprise competitive displacement scenarios. Until the company discloses average revenue per seat and weekly active use rates alongside its seat count, the headline figure captures the top of the commercial funnel without confirming what is flowing through it.<\/p>\n The third \u2014 and today’s new development \u2014 is Maia 300 execution. The capital expenditure commitment for fiscal year 2027 is defensible if custom silicon economics materialize at scale: the MAI architecture’s demonstrated 84% to 89% cost reductions on Nvidia inference multiply into billions of dollars of margin impact if Maia 300 reaches 300,000 or more units in production by 2027. That is an execution bet. The Maia 200 faced production delays and remains deployed in only two U.S. data centers. Supply constraints at TSMC’s advanced process nodes are a documented risk. A Q1 FY2027 print that shows Azure growth below 40% or capex efficiency worsening \u2014 without a credible Maia 300 production timeline \u2014 would change the calculus substantially.<\/p>\n How the Alphabet Rivalry Looks Today<\/p>\n Google Cloud posted 82% revenue growth in its most recent quarter and began recognizing revenue from commercial TPU sales \u2014 a development Microsoft has not yet matched<\/a>. At approximately 17.4x forward earnings, Alphabet trades at a meaningful discount to Microsoft’s approximately 26x multiple. Both companies are investing at unprecedented scale: the four major hyperscalers are collectively projected to spend approximately $725 billion on AI infrastructure in 2026<\/a>, up 77% from roughly $410 billion in 2025.<\/p>\n The critical difference in the Alphabet comparison is Microsoft’s enterprise lock-in. When an organization already runs Exchange, SharePoint, Teams, and GitHub through Microsoft, adding Azure AI workloads carries lower switching friction than migrating to Google Cloud or AWS. The $678 billion backlog is the most quantitative expression of that stickiness.<\/p>\n What neither side of the Alphabet comparison resolves is the custom silicon question. Google enters the second half of 2026 already commercially selling TPUs. Microsoft enters it planning a September chip reveal. If Maia 300 delivers on its stated timeline, the gap closes materially. If it slips \u2014 as the Maia program has slipped before \u2014 the cost structure comparison with Google becomes more challenging.<\/p>\n This article is for informational purposes only and does not constitute investment advice. Consult a qualified financial professional before making investment decisions.<\/p>\n Frequently Asked QuestionsIs Microsoft stock a buy at $500 after the post-earnings rally?<\/p>\n It depends on what you believe about three conditions: whether Azure can sustain 45% growth on a $100 billion-plus annual base through fiscal year 2027; whether Copilot’s 30 million paid seats convert into disclosed revenue per seat that justifies the enterprise pricing; and whether the Maia 300 chip program executes on its timeline to reduce AI inference costs at scale. Multiple analysts including Morningstar ($600 fair value)<\/a> and Tigress Financial ($690 price target)<\/a> see upside from current levels, and GuruFocus’s model suggests approximately 13% undervaluation at $500. The bear case, articulated by TCI’s exit and the capex-to-free-cash-flow inversion, is that the fiscal year 2027 capital expenditure program may not generate returns before balance sheet strain becomes visible. This article is for informational purposes only and does not constitute investment advice.<\/p>\n What is the $678 billion Microsoft backlog, and why does it matter?<\/p>\n Commercial remaining performance obligations represent contracted but not yet recognized revenue \u2014 money enterprise customers have committed to pay Microsoft for future cloud and AI services. The $678 billion figure, up 84% year-over-year<\/a>, exceeds Microsoft’s entire fiscal year 2026 revenue of $331.8 billion. It represents the contracted foundation for Azure’s future growth, but it carries two caveats: OpenAI historically accounted for approximately 45% of the backlog per prior reporting<\/a>, and the backlog converts to revenue only as Microsoft builds the data center capacity to deliver those contracted services \u2014 at a cost that analysts estimate at over $255 billion in fiscal year 2027 capital expenditure.<\/p>\n What is the Maia 300, and how does it change Microsoft’s investment case?<\/p>\n The Maia 300 is Microsoft’s next-generation custom AI accelerator, reported by The Information on August 10, 2026, to be targeting a September reveal with TSMC manufacturing talks for 300,000-plus units<\/a> by 2027. Its predecessor, the Maia 200, uses TSMC’s 3-nanometer process to deliver approximately 10 petaflops of FP4 compute \u2014 and Microsoft’s MAI software models deployed on that hardware have demonstrated 84% to 89% GPU cost reductions<\/a> versus Nvidia-based inference. If Maia 300 scales to 300,000-plus units, it turns what is currently a massive annual capital expenditure line into a structural cost-reduction program rather than pure infrastructure buildout \u2014 which changes the free cash flow recovery timeline materially.<\/p>\n Why did Microsoft’s capital spending exceed free cash flow for the first time in Q4?<\/p>\n In fiscal Q4 2026, Microsoft spent $41 billion on capital expenditure and finance leases while generating approximately $19.6 billion in free cash flow after capital expenditure. The company is building AI data center capacity ahead of the demand it has under contract \u2014 the $678 billion backlog<\/a> represents committed future demand that does not yet have the physical infrastructure to serve it. CFO Amy Hood says the company expects to remain free cash flow positive for the full fiscal year 2027. The question is whether Maia 300 and the MAI model efficiency gains can accelerate the cash-generation recovery fast enough to justify the investment before investors demand evidence of margin recovery.<\/p>\n","protected":false},"excerpt":{"rendered":"Microsoft Chairman and Chief Executive Officer Satya Nadella speaks during the Microsoft Build conference opening keynote in Seattle,…\n","protected":false},"author":2,"featured_media":135139,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[11],"tags":[3074,420,7853,67063,425,416,320,7852,15330,67064,67062,305],"class_list":["post-135138","post","type-post","status-publish","format-standard","has-post-thumbnail","category-microsoft","tag-ai-chip","tag-azure","tag-azure-copilot","tag-azure-growth-forecast","tag-cloud-computing","tag-copilot","tag-microsoft","tag-microsoft-copilot","tag-microsoft-earnings","tag-microsoft-maia-300-chip","tag-microsoft-stock-msft","tag-tech-stocks"],"_links":{"self":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/135138","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/comments?post=135138"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/135138\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media\/135139"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media?parent=135138"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/categories?post=135138"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/tags?post=135138"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}