Microsoft Sits Flat at $493.51 as Meta Rips 5.75% and Alphabet Sheds 3.04%

Microsoft (MSFT) trades at $493.51, down $0.44 or 0.09%, against Tuesday's close of $493.95. Forty-four cents of movement on a $3.67 trillion company, on a day when Meta ripped 5.75% to $648.77 and Alphabet shed 3.04% to $328.07.

That non-reaction is the most informative thing about the stock today.

Meta launched Muse, a paid consumer AI agent, Tuesday evening. The market took roughly $135 billion out of Alphabet and put roughly $107 billion into Meta. Microsoft moved less than half a percent. The market has decided that whatever Muse is, it does not touch Microsoft's business — and it is right, because Microsoft sells AI to IT departments on seat licenses and consumption contracts, not to consumers through app stores.

Underneath that indifference sits a stock that has already had its repricing. Microsoft closed at $390.54 the day before its fiscal fourth-quarter report on July 29 and jumped 8.13% to $422.30 in after-hours trading, then extended to roughly 17% the following session. It has since worked to $493.51 — a 26.4% advance in six weeks — and sits 10.9% below its 52-week high of $553.72 against a low of $349.20.

The report that produced it delivered numbers that ended a specific argument. Azure grew 43% and crossed $100 billion in annual revenue. Commercial remaining performance obligation reached $678 billion, up 84%. Microsoft 365 Copilot passed 30 million paid seats. Annual revenue exceeded $331 billion, up 18%.

The report also delivered the number that will define the next four quarters: fiscal 2027 capital expenditure guidance of $255 billion to $260 billion, against $190 billion committed for 2026 — a 35% increase in a spending program that was already the largest in corporate history.

The thesis running through this piece: Microsoft is the only hyperscaler that has shown the market a demand backlog large enough to justify its capital program, and the $678 billion RPO figure is why the stock trades at 27.5 times earnings while peers with worse cash conversion trade at similar or higher multiples. The question the October 27 report answers is whether the backlog converts fast enough to absorb $260 billion of spending.

The Tape: $499.70 To $493.95 To $493.51

The recent price sequence shows a stock consolidating after a violent move rather than reversing it.

Microsoft closed at $499.70 on Friday, September 4. Markets were closed Monday for Labor Day. Tuesday brought a $5.75 decline of 1.15% to $493.95, with the session ranging from $490.15 to $495.19 and the stock opening $6.00 below the prior close. Wednesday has produced 44 cents of movement to $493.51.

That is a stock that gave back a point on Tuesday and has done nothing since. The $490 level held on the only real test of the week.

The broader tape is worse than the stock. The S&P 500 trades at 7,651.41, down 0.29%. The Nasdaq 100 sits at 29,372, off 0.46%. The Dow has shed 312 points or 0.59%. Apple is down 0.92% to $313.31, Amazon off 1.88% to $252.13, Alphabet down 3.04%, Nvidia lower by 0.47% at $224.67, Oracle off 0.44% at $161.81.

Microsoft's 0.09% decline is better than every megacap peer except Meta and Tesla. On a day with Brent at $101.071, the 10-year yielding 4.8120% and a 60% probability of a Federal Reserve hike on September 16, holding flat is a genuine relative result for a high-multiple growth stock.

The pre-market print at $492.82 and the near-flat regular session together describe an absence of sellers rather than an arrival of buyers. Volume has not been the story this week.

The context that makes the current level meaningful is what it took to get here. Microsoft entered its July 29 earnings report down roughly 19% year to date, having beaten expectations in each of the three prior quarters and been punished each time — shares fell 3.01%, then 9.80%, then 5.00% following those beats. Investors had stopped believing that capital expenditure would convert into revenue.

The fourth-quarter report broke that pattern in a single session. The stock is now up on the year and trading above its 200-day exponential moving average for the first time since the May-June selloff.

Consolidating 26% higher after six weeks is what a stock does when the repricing is complete and the next catalyst is a quarter away.

Technicals: $500 Is The Line, $553.72 Is The Ceiling

The chart is constructive with one obvious hurdle directly overhead.

The 52-week range runs $349.20 to $553.72. At $493.51 the stock sits 41.3% above the low and 10.9% below the high — well into the upper half of its annual range and above its 200-day simple moving average, which it reclaimed following the July earnings report.

The immediate resistance is $499.70, Friday's close and the round $500 handle sitting immediately above it. Microsoft has been rejected there this week, with Tuesday's session opening $6.00 below that close and never recovering it. Reclaiming $500 restores the short-term uptrend.

Above that, the path runs toward the mid-$530s and then to the $553.72 high, which is 12.2% above current price. That high was set before the year-to-date drawdown and represents the pre-capex-panic valuation.

Beneath price, $490.15 is Tuesday's low and the first genuine support — 0.7% below spot. It held cleanly on the only test this week. Below that, the stock has an open gap zone back toward the mid-$460s from the post-earnings run, and then the $422.30 area where the July repricing began. That level is 14.4% below current price and represents the floor of the current structure.

The 200-day EMA sits well beneath, having been reclaimed in late July. A stock trading above aligned 20-day and 50-day moving averages, with the 200-day recovered, is in an uptrend by every standard definition.

Momentum has cooled without breaking. The stock has fallen 1.18% over the past week and roughly 1.52% over the past month — a pause after a 26% six-week advance rather than distribution. That is healthy consolidation, and it is what a base looks like before the next leg.

The technical signal from moving averages and momentum indicators currently reads as a buy on the daily timeframe.

The level that matters most for the next quarter is $553.72. Clearing a 52-week high on a $3.67 trillion company requires a catalyst, and the only one on the calendar is the October 27 earnings report. Between now and then, the range from $490 to $500 is where the stock lives.

The Q4 Print That Changed The Stock's Year

The fiscal fourth quarter ended June 30 and the numbers were a clean beat on every headline metric.

Revenue came in at $90.01 billion, up 18% year over year against consensus estimates of roughly $87.67 billion. Non-GAAP earnings per share printed $4.74 against a $4.24 consensus — a 11.8% beat. GAAP earnings per share reached $4.81 with net income of $35.77 billion, up from $27.23 billion and $3.65 per share in the year-ago quarter. Net income grew 31.4% on 18% revenue growth, which is operating leverage working in the right direction.

The Productivity and Business Processes segment, covering Office, Dynamics and LinkedIn, produced $37.85 billion in revenue, up 14.3% and ahead of the $37.19 billion consensus.

For the full fiscal year, annual revenue surpassed $331 billion, up 18%. Microsoft Cloud revenue exceeded $214 billion, up 27%. Azure crossed $100 billion in annual revenue for the first time, growing 41% over the year.

Set that against the third quarter for the trajectory. Q3 fiscal 2026 delivered revenue of $82.9 billion up 18%, operating income of $38.4 billion up 20%, and EPS of $4.27 against a $4.13 consensus. Azure grew 40% in that quarter, its fastest pace in over a year, with AI services accounting for roughly 16 percentage points of that growth. Microsoft Cloud revenue reached $54.5 billion, up 29%. The AI annual revenue run rate stood at $37 billion, up 123%.

Sequential acceleration from 40% to 43% Azure growth is the detail that mattered most, because it directly contradicted the decelerating-cloud narrative that had taken the stock down 19% year to date.

Guidance extended the story. The chief financial officer projected 45% Azure growth at constant currency for fiscal first quarter, above the 41.4% consensus, and management provided fiscal 2027 guidance for double-digit revenue and operating income growth.

The market response was the largest single-session repricing in the stock's recent history. Shares moved from $390.54 to $422.30 in after-hours trading, an 8.13% gain, and extended to roughly 17% the following day. That reaction followed three consecutive quarters where beats produced declines of 3.01%, 9.80% and 5.00%.

Sentiment does not shift on one print unless the print answers the specific question the market was asking. This one did.

Azure Past $100 Billion And 43% Growth

Azure and other cloud services grew 43% year over year in the fourth quarter, ahead of the company's own guided range of 39% to 40%, and pushed the business past $100 billion in annual revenue for the first time.

Both halves of that sentence carry weight. A 43% growth rate on a business at $100 billion of annual scale is extraordinary — the law of large numbers says growth should decelerate as the base expands, and Azure accelerated from 40% in the third quarter to 43% in the fourth.

The composition is what makes it credible. In the third quarter, AI services accounted for roughly 16 percentage points of Azure's 40% growth. That means the non-AI base business was growing at roughly 24%, and the AI layer was additive rather than cannibalizing. The company's AI annual revenue run rate reached $37 billion in the third quarter, up 123%.

Guidance for fiscal first quarter sits at 45% constant-currency Azure growth against a 41.4% consensus, with management expecting continued acceleration through the first half of fiscal 2027. A cloud business guiding to acceleration at $100 billion scale is a rare thing.

The competitive comparison sharpens it. Amazon Web Services grew 37% to $42.23 billion in its second quarter, with the AI business at a $25 billion annualized run rate. Google Cloud grew 82% to $24.77 billion in Alphabet's second quarter. Azure is growing faster than AWS from a larger base and is roughly four times the size of Google Cloud.

Google Cloud's 82% growth rate is faster, but it is compounding on a $24.77 billion quarterly base against Azure's $100 billion annual scale. The absolute dollar addition favors Microsoft substantially.

What Azure growth actually determines is the capital conversion rate. The company is spending $190 billion in calendar 2026 and guiding to $255 billion to $260 billion in fiscal 2027 specifically to build Azure capacity. Azure growth is the only measure of how much of that spending is generating revenue in the same fiscal year.

At 43% growth with 45% guided, the conversion is working. At 35% or below, the capex-to-revenue relationship breaks and the entire investment case comes under review.

The October 27 report is the next reading.

The $678 Billion Backlog Is The Real Number

Commercial remaining performance obligation surged 84% year over year to $678 billion. That figure is the single most important disclosure in the fourth-quarter report and it is the reason Microsoft's stock behaved differently from Meta's and Alphabet's this earnings season.

RPO measures contracted future revenue — money customers have committed to spend that has not yet been recognized. It is not a pipeline, not a forecast, and not a management estimate. It is signed business.

At $678 billion against annual revenue of $331 billion, Microsoft is carrying slightly more than two years of total company revenue in contracted commitments. Growing that number 84% in twelve months means the commitments are being signed far faster than they are being consumed.

This is what separates Microsoft from every other company in the AI capital cycle. Meta is spending $130 billion to $145 billion in 2026 with free cash flow at $784 million and no contracted backlog at all — its revenue is advertising, sold in real time. Alphabet spent $44.92 billion in a single quarter with free cash flow at negative $5.9 billion and suspended its buyback. Both are building capacity against expected demand.

Microsoft is building capacity against $678 billion of signed demand.

That distinction is why the market rewarded a similar capex number in Microsoft's case and punished it elsewhere. Infrastructure spending backed by concrete contractual commitments is investment. Infrastructure spending backed by conviction is a bet.

The caveat is duration and cancellability. RPO includes multi-year contracts stretching over horizons that give customers room to renegotiate, and a portion of it relates to arrangements with AI customers whose own funding depends on capital markets remaining open. An 84% increase concentrated in a small number of very large agreements carries counterparty risk that a broad enterprise base does not.

Microsoft has not disclosed the concentration. That is the disclosure worth pressing for on October 27.

For now, $678 billion is a genuine moat and it is what the current 27.5 times multiple is paying for.

Copilot At 30 Million Seats And The Billing Shift

Microsoft 365 Copilot surpassed 30 million paid seats, up from more than 20 million as of April, with net seat additions more than doubling sequentially and user satisfaction scores doubling over the last three quarters.

Ten million net paid seats added in a single quarter, with the pace of additions doubling, is the strongest consumer or enterprise AI adoption datapoint any company has disclosed. At the pricing levels these seats typically carry, 30 million paid seats implies an annual run rate in the neighborhood of $10 billion.

GitHub Copilot now has 50 million users. Hundreds of enterprise customers have purchased millions of seats for high-end E7 productivity bundles.

The more consequential development is structural rather than numerical. Microsoft is introducing usage-based billing models for Copilot, Cowork and GitHub Copilot alongside per-seat licensing. That changes the revenue model in a way that expands the addressable market meaningfully.

Per-seat licensing caps revenue at the number of employees an organization has. Usage-based billing does not. An enterprise with 10,000 employees can only buy 10,000 seats, but it can consume an unbounded amount of agent compute if the agents are doing work that previously required more people. Converting from seats to consumption is how a software company escapes the ceiling of its customer's headcount.

The architectural approach behind it is a mixed-model design that lets enterprises use frontier models for their strengths, low-cost models for efficiency, and train their own — the same pattern Microsoft uses internally in Copilot and GitHub Copilot, now being made available to customers.

The comparison to Meta's Muse is instructive precisely because it is not competitive. Muse is a consumer subscription sold to individuals for personal tasks. Copilot is enterprise software sold to IT departments with security review, procurement cycles and multi-year commitments. Different buyer, different sales motion, different revenue durability.

Enterprise revenue is stickier and slower. Consumer revenue is faster and churns. Microsoft holds the former, and that is why 30 million paid seats is a more durable number than any consumer agent adoption figure will be.

The risk is that agent capability commoditizes and per-seat pricing collapses toward the cost of inference. The usage-based pivot is Microsoft's hedge against exactly that.

Capex $255–260 Billion And The Accounting Change

Fiscal 2027 capital expenditure guidance sits at $255 billion to $260 billion, against $190 billion committed for calendar 2026 — an increase of roughly 35%. Fourth-quarter capital expenditure alone reached approximately $41 billion.

Those figures are without precedent. $260 billion in a single fiscal year exceeds the annual revenue of all but a handful of companies on the planet, and it is being deployed by one company into data centers, GPUs, CPUs, power and memory.

The accounting overlay matters and it has been widely misread. Effective fiscal 2027, Microsoft is extending the estimated useful life of data centers and office buildings from 15 to 25 years. That change affects only the timing of future depreciation, with minimal benefit to fiscal 2027 operating income.

The larger effect is on the capex line itself. More future data center leases shift from finance leases, which are included in capital expenditure, to operating leases, which are not. That adjustment brings the calendar 2026 capex expectation to approximately $175 billion while underlying investment plans remain unchanged.

Read that carefully: the reported capex number and the actual spending are diverging because of a classification change. Anyone comparing Microsoft's capex line to a peer's without adjusting for lease treatment is comparing different things.

The defensive argument management has made against oversupply risk is that capital expenditure is heavily weighted toward short-lived assets — CPUs and GPUs — which carry shorter lead times. Short lead times mean the spending can be throttled quickly if demand disappoints. That flexibility is real and it is the reason a $260 billion program is survivable rather than terminal.

The counterargument is that extending data center useful life from 15 to 25 years while simultaneously buying assets described as short-lived is an uncomfortable pairing. Depreciating a building over 25 years is defensible. The GPUs inside it will not last 25 years.

The memory shortage is compounding the cost. Soaring memory chip prices stemming from the AI infrastructure buildout are inflating the per-unit cost of every server rack, which means $260 billion buys less capacity than $260 billion would have bought a year ago.

Capex is the single biggest swing factor for margins and the stock over the next several quarters.

Valuation: 27.5x, $3.67 Trillion, 16% To Target

At $493.51 against trailing EPS of $17.95, Microsoft trades at 27.5 times earnings. Market capitalization stands at $3.67 trillion on roughly 7.43 billion shares.

That multiple sits below the stock's own history. Microsoft's 10-year average price-to-earnings ratio is 31.52. The 15-year average is 29.46. The 20-year average is 27.43. At 27.5 times, the stock is trading essentially at its two-decade average and 13% below its ten-year average — despite growing revenue 18% and Azure 43%.

The comparison to peers frames it further. Meta trades at 24.7 times trailing earnings with a 12-point operating margin collapse and free cash flow down 90.8%. Microsoft trades at 27.5 times with margins intact and a $678 billion contracted backlog. The three-point premium for a materially better business is modest.

The analyst distribution is unusually consolidated for a stock this size. The average twelve-month target sits at $572.92, implying 16.1% upside from $493.51. The high estimate is $870. The low estimate is $400, implying 19% downside. Fifty-two analysts recommend buying with zero sell ratings, producing a Strong Buy consensus across 55 covering firms.

A zero-sell-rating distribution on a $3.67 trillion company is worth noting as a sentiment reading. It means the entire covering universe agrees on direction and disagrees only on magnitude — the $470 spread between the $400 low and $870 high reflects differing views on how much capex risk is already priced in.

The dividend is a rounding error at $0.91 per share most recently declared and a trailing yield of 0.73%, with a payout ratio near 20%. Nobody owns Microsoft for income, but a growing dividend at a 20% payout ratio signals a management team confident enough in cash generation to commit to a rising distribution while spending $260 billion on infrastructure.

EBITDA stands at $198.69 billion with an EBITDA margin of 58.39%. A company converting 58 cents of every revenue dollar into operating cash before interest, tax and depreciation has the capacity to fund the capital program internally.

The valuation debate reduces to one input: whether fiscal 2027 depreciation on $260 billion of spending compresses the earnings that the 27.5 multiple is applied to.

Margins, Cash, And The 58% EBITDA Machine

The financial engine underneath the AI story is what makes the capital program possible, and it has held up better than any peer's.

Microsoft Cloud revenue exceeded $214 billion for fiscal 2026, up 27%. Total revenue passed $331 billion, up 18%. Fourth-quarter net income reached $35.77 billion on $90.01 billion of revenue — a 39.7% net margin. Third-quarter operating income was $38.4 billion on $82.9 billion of revenue, growing 20% against 18% revenue growth.

Operating income growing faster than revenue while capital expenditure runs at $41 billion a quarter is the metric that separates Microsoft from Meta, whose operating margin compressed to 31% from 43%, and from Alphabet, whose free cash flow turned negative at -$5.9 billion.

EBITDA at $198.69 billion with a 58.39% margin describes a business with pricing power in enterprise software that no amount of AI competition has yet eroded. Office, Windows, Dynamics and LinkedIn generate the cash that funds Azure's buildout, and Productivity and Business Processes grew 14.3% to $37.85 billion in the fourth quarter — respectable growth from a mature base.

The concern flagged by credit analysis earlier this year was cash generation relative to spending. Microsoft has ranked toward the bottom of large software companies on cash generation metrics precisely because the capital program consumes so much of the operating cash flow. That is arithmetic rather than deterioration — the business generates more cash than it ever has, and it spends more than it generates on infrastructure.

The company employs 223,000 people. Headcount discipline alongside a $260 billion capital program is the same trade Meta is running: spending on machines while holding the line on people.

The balance sheet retains capacity. Microsoft carries one of the small number of remaining top-tier corporate credit ratings, and that access to debt markets at low spreads is a genuine competitive advantage in a capital cycle this large. A competitor without that access has to fund from cash flow or equity.

The item to watch in fiscal 2027 is the depreciation line. Extending asset lives from 15 to 25 years pushes charges into later periods, which flatters near-term operating income at the cost of later years.

Competition: Alphabet, Amazon, And The Enterprise Split

The hyperscaler comparison in the second quarter of calendar 2026 lays out the competitive position precisely.

Azure grew 43% in its fourth fiscal quarter and crossed $100 billion in annual revenue. Amazon Web Services grew 37% to $42.23 billion in Amazon's second quarter, with an AI business at a $25 billion annualized run rate. Google Cloud grew 82% to $24.77 billion.

By absolute scale, Azure is roughly 2.4 times Amazon's quarterly cloud revenue on an annualized basis and roughly four times Google Cloud. By growth rate, Google Cloud leads, Azure is second and AWS is third. By capital efficiency, Microsoft is the only one of the three that showed a contracted backlog large enough to justify the spending.

Alphabet trades at $328.07, down 3.04% today. Its second-quarter revenue reached $119.8 billion, up 24.2%, with the Gemini App at 950 million monthly active users. It spent $44.92 billion on capex in the quarter and suspended its buyback. Google Cloud's 82% growth is genuinely faster than Azure's, and Gemini's consumer reach is larger than anything Microsoft has in consumer AI.

Amazon trades at $252.13, down 1.88%. AWS remains the largest infrastructure provider by installed base, and Amazon has the advantage of selling both the compute agents run on and the commerce inventory agents transact against.

Microsoft's differentiated position is the enterprise software layer sitting above the cloud. Neither Amazon nor Alphabet owns Office, Windows, Teams, Dynamics, LinkedIn or GitHub. Copilot at 30 million paid seats and GitHub Copilot at 50 million users exist because Microsoft already had the distribution inside the enterprise before AI arrived.

That is the same distribution argument Meta is making with WhatsApp, applied to a different market. Microsoft owns the workplace surface. Meta owns the messaging surface. Alphabet owns search and Android. Amazon owns commerce and infrastructure.

The nearest competitive threat is that agent capability lets a customer replace the Office layer entirely — an AI that drafts, calculates and presents without needing Word, Excel or PowerPoint underneath it. That is a multi-year risk rather than a quarterly one, and the usage-based billing pivot is Microsoft positioning to monetize the agent even if the application layer erodes.

Why Muse Did Not Move Microsoft

Meta's Muse launch produced a 5.75% move in META, a 3.04% decline in Alphabet, and 44 cents of movement in Microsoft. The reason deserves stating explicitly because it explains what Microsoft actually sells.

Muse is a consumer AI agent distributed through WhatsApp and a standalone app, sold on tiered consumer subscriptions to U.S. users aged 18 and over. It books travel, sends emails, fills out forms and makes purchases on an individual's behalf.

Copilot is enterprise software. It is sold to IT departments through procurement processes involving security review, compliance certification, data residency requirements and multi-year commitments. Its 30 million paid seats sit inside organizations that have signed contracts contributing to a $678 billion backlog.

Those are not competing products. They compete for the same underlying model capability and the same GPU supply, but not for the same buyer, the same budget or the same sales cycle.

The market understood that immediately, which is the most encouraging signal in Microsoft's tape today. A market that repriced Alphabet by $135 billion on a consumer agent launch and left Microsoft untouched has correctly identified that Microsoft's revenue is insulated from consumer AI competition.

The intersection that does exist is the partnership layer. Meta has collaborations with Microsoft, Nvidia, AMD, Broadcom and OpenAI. Meta's Muse Spark models and its $130 billion to $145 billion capital program create demand for the same infrastructure Microsoft is building. Every frontier lab that does not build its own data centers is a potential Azure customer.

That is the underappreciated angle. Microsoft's exposure to the consumer agent war is not competitive — it is as a supplier. More consumer agents mean more inference demand, and inference demand runs on cloud infrastructure.

Microsoft's own AI products keep shipping alongside the enterprise stack, including the Project Perception cybersecurity system and continued investment across the portfolio.

The stock's indifference today is the market saying Microsoft is upstream of the fight.

Legal: The OpenAI Lawsuits And The Publisher Problem

The most persistent overhang on Microsoft is not competitive or financial. It is legal, and it is growing.

The Seattle Times Company and Newsday have sued OpenAI and Microsoft in federal court, alleging the companies copied the newspapers' journalism without permission to train AI systems. That follows the closely watched New York Times lawsuit, where recent filings covered a range of copyright law and cultural arguments.

Microsoft being sued by the parent company of its hometown daily newspaper is a notable escalation in a litigation trend that now spans multiple publishers across multiple jurisdictions.

The financial exposure is difficult to size and that is precisely the problem. Copyright damages in the United States can be assessed per work, and training corpora contain enormous numbers of individual works. A ruling that establishes per-work liability for training data would produce a number that no reserve accounts for.

The more consequential risk is structural rather than monetary. An adverse ruling on whether training on copyrighted material constitutes fair use would require licensing arrangements across the industry, which raises the cost of every model Microsoft trains or serves and compresses the margin on AI revenue that the entire investment case depends on.

Microsoft's position is complicated by the OpenAI relationship. The company is named alongside OpenAI in these actions, which means it carries litigation exposure for a partner's training decisions as well as its own.

Set against that, the practical market treatment has been dismissive. The stock dipped on the latest lawsuit filings and recovered. A $3.67 trillion company with $198.69 billion of EBITDA can absorb settlements measured in billions without a material change to the investment case, and the market is pricing settlements rather than injunctions.

That judgment is probably correct and it is not certain. The distinction between a damages award and an injunction on training practices is the difference between a one-time charge and a business model change.

For anyone modeling Microsoft over multi-year horizons, the publisher litigation is the tail risk that does not appear in any capex or Azure growth calculation.

MSFT Price Forecast: Levels, Scenarios, Probabilities

The executable map.

Upside, in order: $499.70 as Friday's close and the $500 handle, 1.3% above spot. Then the mid-$530s, and $553.72 as the 52-week high, 12.2% above current price. Beyond that, the average twelve-month target at $572.92 implies 16.1% upside, with the high estimate at $870.

Downside, in order: $490.15 as Tuesday's low, 0.7% below spot. Then the mid-$460s where the post-earnings gap sits. Then $422.30, the level from which the July repricing began, 14.4% beneath current price. The low analyst estimate sits at $400, implying 19% downside. The 52-week low is $349.20.

Base case at 48% probability: Microsoft consolidates between $470 and $520 into the October 27 fiscal first-quarter report. The stock holds its post-earnings gains without extending them, because the next catalyst is a quarter away and the $500 handle needs a reason to break. Target range $480 to $510.

Bull case at 33% probability: fiscal Q1 Azure growth comes in at or above the guided 45% constant currency, commercial RPO extends beyond $678 billion, Copilot seat additions maintain their doubling pace, and management holds capex guidance without raising it. The stock clears $500, works through the mid-$530s, and tests $553.72. Upside 12.2%, with $572.92 the twelve-month target.

Bear case at 19% probability: Azure growth decelerates below 40%, fiscal 2027 capex guidance moves above $260 billion, or the depreciation line from the useful-life extension proves less flattering than modeled. Add a Federal Reserve hike on September 16 lifting the 10-year through 4.8120%, and a 27.5 multiple compresses toward 24. That takes the stock to roughly $430. Downside 13%.

The distribution skews bullish because the $678 billion backlog is contractual rather than aspirational, and because the valuation at 27.5 times sits below the ten-year average of 31.52 and roughly at the twenty-year average of 27.43.

The single most important date is October 27. The single most important number is Azure constant-currency growth against the 45% guide.

Verdict: The Only Hyperscaler With The Receipts

Microsoft at $493.51, down 0.09%, is the least interesting stock on the tape today and the most defensible one.

The constructive case is documented and specific. Annual revenue surpassed $331 billion, up 18%. Microsoft Cloud exceeded $214 billion, up 27%. Azure crossed $100 billion in annual revenue growing 43% in the fourth quarter against a guided 39% to 40%, with 45% constant-currency growth guided for the current quarter against a 41.4% consensus. Commercial RPO reached $678 billion, up 84% — more than two years of total company revenue in signed commitments. Copilot passed 30 million paid seats with net additions more than doubling sequentially and GitHub Copilot at 50 million users. Fourth-quarter net income of $35.77 billion on $90.01 billion of revenue is a 39.7% net margin, and EBITDA of $198.69 billion carries a 58.39% margin. The stock trades at 27.5 times trailing earnings against a ten-year average of 31.52, with 52 buy ratings and zero sells.

The cautious case is equally real. Fiscal 2027 capital expenditure guidance of $255 billion to $260 billion is a 35% increase on a program that was already the largest in corporate history, and the memory shortage means those dollars buy less capacity than they did a year ago. The useful-life extension from 15 to 25 years and the finance-to-operating lease shift make the reported capex line less comparable to peers and to Microsoft's own history. The stock sits 10.9% below its 52-week high having already run 26.4% in six weeks, which means the good news is in. Publisher litigation alongside OpenAI is expanding rather than resolving. And a 60% probability of a Federal Reserve hike on September 16, with the 10-year at 4.8120%, is the wrong environment for a 27.5 multiple.

The verdict is constructive with the catalyst deferred. Microsoft is the only company in the AI capital cycle that has shown the market a contracted demand backlog large enough to justify what it is spending, and $678 billion is the number that separates it from every peer building on conviction alone.

Hold $490.15 and the post-earnings structure survives. Clear $500 and the path to $553.72 opens. Between those numbers, ten dollars wide, is a stock waiting for October 27.

That's TradingNEWS