Microsoft (MSFT) at 27.5 Times Earnings: What the Price Asks

Microsoft trades around $493.78 as I write this, at 27.5 times trailing earnings. Its own five-year average is 32.5. So the stock that people bought without thinking at 32 times is available at 27, and the honest question is whether that is a discount or a warning.

I will say where I land now. At this price I would call Microsoft fairly priced for a business of this quality, and a purchase I could defend, though not a bargain. The multiple has already come down from the level that worried me earlier, but the AI capital bill is running through the results, and the gross margin already shows it. My view rests on that tension, and the rest of this post tests it with numbers.

What the numbers say today

Start with the business. Fiscal 2026 revenue was $331.8 billion, up 18% from $281.7 billion the year before, and $211.9 billion three years earlier. Operating income was $155.2 billion, a 47% margin. The latest quarter brought in $90.0 billion, also up 18%. For a company this size, growing revenue at 18% is unusual, and it is the fact that keeps holders patient.

The stock ranges between $349 and $549 over 52 weeks and sits 10.1% below the high, about 42% above the low. Market capitalization is around $3,666.6 billion. On the last earnings date, 2026-07-29, the shares moved +15.5%, against an average earnings-day move of 8.1%. Volatility around reports has been high for a stock people call boring.

Metric	Value	Context
Price	$493.78	52-week range $349 to $549
P/E (trailing)	27.5	five-year average 32.5
P/E (forward)	25.1	forward EPS $19.71
Revenue growth	18%	$331.8 billion in fiscal 2026
Operating margin	47%	gross margin 67.9% vs 68.8% prior year
Analyst target	$573	range $510 to $700, 32 analysts

Where the growth comes from

Microsoft reports three segments, and only one carries the valuation. Productivity and Business Processes holds Office, Teams, LinkedIn and Dynamics, a sticky and slow-growing annuity. More Personal Computing covers Windows, devices, search and Xbox and tracks the PC market, which means it grows slowly. Intelligent Cloud, and inside it Azure, is the piece the market pays for. Management has said a growing share of Azure growth now comes from AI services rather than classic migration of company servers to the cloud.

That shift matters for two reasons. AI workloads pull in more spending per customer than a storage or hosting contract does, so revenue can grow faster than customer counts. They also demand far more hardware per dollar of revenue, which is where the margin question begins. I would read the company’s segment tables in the latest 10-Q on SEC EDGAR before trusting any single growth figure I or anyone else quotes, because segment definitions have been reshuffled before.

The capital bill shows up in gross margin

Gross margin slipped from 68.8% to 67.9%, a fall of about nine-tenths of a point. That is small. It is also the first visible trace of a very large investment in data centers and chips, because depreciation on new hardware lands in cost of revenue long before the AI revenue it supports arrives at full scale.

Operating margin at 47% says the company is absorbing this comfortably today. Net income rose to $133.7 billion from $101.8 billion, a jump of about 31% on 18% revenue growth, and I would not extrapolate that gap. Earnings that outgrow sales by that much often include items that do not repeat, and I have not separated them here. If the margin gap between earnings growth and revenue growth closes next year, forward EPS will look less generous than it does now.

I laid out the method in my note on reverse-DCF math for expensive stocks: figure out what growth a price requires, then ask whether the company has ever delivered it. Microsoft has delivered the required growth so far, which is why the argument is about how long, not whether.

What a 27.5 multiple asks for

The earnings yield at a 27.5 multiple is about 3.6%. Trailing earnings per share are $17.95, and analysts expect $19.71, roughly 10% higher. On that forward figure the stock trades at 25.1 times.

Try the arithmetic in reverse. If I want a forward multiple of 25 on $19.71, I pay about $493, which is where the stock is. At 20 times I would pay about $394, close to the 52-week low of $349. At 30 times, the price would be about $591, above the high of $549. The current price sits near the bottom of the range that Microsoft’s history has treated as normal. A holder who buys here and sees the multiple settle at the five-year average of 32.5 on 10% earnings growth gets a total return in the low teens for the year, counting the 0.72% dividend. Take the multiple to 22, a fair bear case, and the return turns negative even with the earnings growth intact.

Analysts are more generous. The average target is $573, 16% above the price, and 97% of the 32 analysts rate it a buy. I treat that as a mood reading, not a forecast, since even the lowest target of $510 is above today’s price. A group that agrees that much has stopped providing information about risk.

Comparing it with its neighbors

Microsoft looks different when set beside other large technology names. In my piece on Alphabet’s discount to Microsoft I argued that the market has been paying more for Microsoft’s predictability. That premium has narrowed. I also wrote about Amazon’s split between retail and cloud, where the cloud unit competes with Azure directly, and the two companies’ capital plans are in effect a bet on the same customers. If one of them cuts data center spending, the other’s pricing power improves, which is worth remembering when you read guidance.

The quantitative scoring I follow moved from a B to a D on this name, and I take that seriously. It reflects the shift in the price and estimate picture, not a collapse in the business. A rating of D on a 47% operating margin company says the stock is no longer cheap on the models’ inputs, and that momentum has been mixed.

Dividends, buybacks and what holders actually collect

The dividend is small: $3.56 a share over the trailing twelve months, a 0.72% yield. Nobody owns Microsoft for income, and I would not compare it with a utility. What matters is that the payout is a sliver of earnings, about 20% of $17.95, so it has room to grow even if profit growth slows to single digits. A holder who reinvests that dividend adds very little to the return in a given year, and I would not lean on it in the arithmetic above.

Short interest sits near 1.0% of the float, which is low. Nobody is leaning hard against the stock, so a sharp squeeze is not part of the upside case, and a wave of forced covering is not part of the downside either. The price moves when earnings move, and the 8.1% average reaction on report day tells you how much is at stake each quarter.

How I would size a position

If a portfolio already holds Microsoft through an index fund, it owns a lot of it: the company is one of the largest weights in every major US index. A second purchase on top adds concentration, not diversification. I would think of any extra buying as a decision about how much of the AI spending cycle to carry, and I would cap it at a level where a 29% drop, which is what the 52-week low implies from here, would not change how I sleep or spend.

Spreading purchases over three or four quarters also suits a stock that has moved +15.5% in a day. Buying in tranches around earnings dates means no single report decides the whole outcome.

Where I would add, and what proves me wrong

I would add on weakness toward the 22 to 24 times forward range, which translates to about $434 to $473 on current estimates, if gross margin holds within a point of 67.9%. I would not add at the highs, and I would not sell a long-term holding just because the multiple compressed, since revenue growth of 18% pays for a lot of patience.

The counter-case is specific. If quarterly revenue growth, now 18%, falls below 12% while capital spending keeps rising, then the multiple should compress toward 22 and the stock has a lot further to fall than the current price suggests. The other risk is on the upside: the shares jumped +15.5% on the last report, and a stock that can gain that much in a day can also lose it, so position size matters more than the entry multiple.

There is also a limit to what I know. I have not verified the cost of the company’s AI build-out from filings for this post, and I cannot tell how much of the current earnings power is one-time. Both facts matter and I have left them as open questions rather than pretending to a number.

Two thresholds guide me from here. Gross margin below 67.9% minus one point would tell me the AI spending is costing more than expected. A quarterly revenue growth rate under 12% would tell me the growth story is ending sooner than the price assumes. Until either happens, 27.5 times earnings for this quality is a fair price rather than a gift.

Analysis and opinion only, not investment advice. Figures come from Microsoft’s filings on SEC EDGAR and its investor site; valuation multiples are approximate and were checked on September 22, 2026.