Bid Cap
Company library Semis & Hardware

Company report

QUALCOMM Incorporated QCOM

Three-pass checked

The bet you're really making is that Qualcomm keeps putting the Snapdragon application processor and cellular modem inside the world's Android phones, and keeps collecting a fee on nearly every smartphone sold anywhere for using its patents. You're betting that as Apple finishes its own modem and stops buying Qualcomm's, the newer lines, chips for cars, laptops and factory gear, grow fast enough to offset the revenue decline. Right now it is going the wrong way: sales fell 4% in the June quarter to the lowest in over a year, and the margin slipped to 53% from 56%. You pay about 17 times earnings, the low end of where the stock has sat for a decade and a fraction of what its chip rivals fetch.

Key data

Price$168.74
52-week range$121.99 – $259.92
P/E (TTM / FY28E)16.8x / 12.8x
EV/EBITDA13.9x

QCOM · price with moving averages

Daily · 6MWeekly · 3Y
$95$137$179$221$263 Oct '23May '24Dec '24Jul '25Feb '26Oct '26 Bid Cap
EMAs82140

Source: market data.

The business

Qualcomm earns money two ways. The larger half, QCT, designs and sells Snapdragon chips: the processor and the cellular modem inside most premium Android phones, plus a growing line for cars, PCs and connected industrial gear. The smaller but far richer half, QTL, licenses Qualcomm's foundational cellular patents and collects a royalty on essentially every 3G, 4G and 5G handset made, whether or not it carries a Qualcomm chip. Licensing is the moat: the patent pool is deep enough that rivals building phones still pay to use it. It is a fabless model, so TSMC and Samsung actually etch the silicon, and the piece a buyer would recognize is the Snapdragon badge on a flagship phone. Concentration is the whole story: a handful of Android makers and one large U.S. customer now building its own modem.

The numbers

Read the operating line, not the bottom line, because tax items have whipped reported profit around. The June quarter is the honest tell.

QuarterRevenueOp. incomeDiluted EPS
Q2 FY25 (Mar '25)$11.0B$3.1B$2.52
Q3 FY25 (Jun '25)$10.4B$2.8B$2.43
Q1 FY26 (Dec '25)$12.3B$3.4B$2.78
Q2 FY26 (Mar '26)$10.6B$2.3B$6.88
Q3 FY26 (Jun '26)$9.9B$1.6B$1.87

Operating income has fallen three quarters running, from $3.4B to $1.6B, and June's $1.6B is down 41% from the same quarter a year ago. That is the real deterioration. The $6.88 in March is not earnings power; it is a roughly $5B non-cash tax benefit from recognizing deferred tax assets, the mirror image of a large charge the year before. Strip both out and the picture is a business whose top line is shrinking and whose margins are giving ground, with gross margin down to 53% from 56% as the chip segment leads the slide.

Fiscal yearRevenueOp. incomeDiluted EPS
FY2021$33.6B$9.8B$7.87
FY2022$44.2B$15.9B$11.37
FY2023$35.8B$7.8B$6.42
FY2024$39.0B$10.1B$8.97
FY2025$44.3B$12.4B$5.01
9M FY26$32.8B$7.3B$11.53

Fiscal 2025 set a revenue record and lifted operating income 23%, then fiscal 2026 rolled over: nine-month revenue of $32.8B is a hair below the same span last year, and the $11.53 EPS is flattered by that one tax quarter. The diversification the last look leaned on has not yet reached the consolidated line, which fell again in June. At $169 you pay about 17 times trailing earnings, the low end of a decade that ranged from 10 to 33 times, and roughly a third of the 54 times its fabless-chip peers command. The market is not pricing a compounder; it is pricing a handset-and-licensing asset it expects to shrink. What this memo believes the market underweights: the cash engine is intact even as profit falls, and the single print that settles it is the direction of chip-segment revenue and gross margin over the next two quarters.

Fiscal yearOperating cash flowBuybacks
FY2022$9.1B$3.1B
FY2023$11.3B$3.0B
FY2024$12.2B$4.1B
FY2025$14.0B$8.8B

Cash generation kept climbing to $14B while buybacks more than doubled to $8.8B, so the company is retiring shares into the weakness.

Management

CEO Cristiano Amon sold about $3.65M of stock over two days in early May 2026, part of 64 insider sales totaling $7.2M over the year and set against zero insider buys; the vendor feed does not carry the 10b5-1 footnotes, so plan status is not disclosed. The capital record is the more useful signal: management doubled buybacks in fiscal 2025 and repaid about $2B of long-term debt in the June quarter, leaving net debt near 0.8 times EBITDA. Roughly 22% of revenue still goes to R&D, which is where the automotive, PC and data-center bets are being funded. Pay is conventional and the guidance record is unremarkable, small beats on non-GAAP prints against a falling GAAP base.

How it fails or surprises you

Apple builds Qualcomm out. Apple's in-house modem is replacing Qualcomm's across the iPhone line, and management has guided its share of that socket toward near zero by 2027. Losing one of its largest chip customers pulls revenue and high-margin dollars at once. The tell is the sequential chip-segment revenue line in the next two 10-Qs.

The offset hasn't arrived. Operating income fell 41% year over year in June while the newer lines were supposedly maturing; if cars, PCs and industrial were filling Apple's hole, the consolidated top line would not still be falling. This is the fact the bull read explains least well. Watch chip revenue and gross margin turn up together, or admit the hole is winning.

Edge AI re-rates it (right tail). Qualcomm has stood up a data-center segment and is pushing on-device AI, which lifts the silicon content in every AI-capable phone and PC. At 17 times earnings the market pays nothing for this. The first real data-center revenue disclosure, or an AI-PC attach rate that shows up in chip revenue, would force the multiple back toward peers.

Closing thoughts

The market prices this as a melting asset, near a decade-low multiple and a third of peer valuations, and the evidence half-agrees: revenue and operating profit are declining now, not in a forecast. But the left tail is a slow erosion, not a solvency event. Operating cash flow is $14B, the current ratio is 2 to 1, net debt is under one turn of EBITDA, and the company is buying back stock at roughly 5% of market cap annually. That combination pays you to wait. The fatter tail is probably still down, a value trap where handset earnings erode faster than buybacks shrink the count, but the right tail, a diversification and edge-AI re-rating off a cheap base, is real and unpaid-for. That is judgment, not a number.

The bet is still that Qualcomm keeps the Snapdragon application processor and cellular modem inside the world's Android phones and keeps collecting a fee on nearly every smartphone sold anywhere for using its patents, while the newer lines, chips for cars, laptops and factory gear, grow fast enough to offset the revenue decline as Apple finishes its own modem. It breaks if chip revenue keeps falling and gross margin keeps slipping at the same time; those two numbers, together, tell you first. The read is wrong if June is the last quarter chip revenue declines and it isn't; watch the next two prints.

Methodology

Fundamentals read this run from filed XBRL in the 10-Q dated 2026-07-29, period ended 2026-06-28; operating income, revenue, cash flow and buybacks are company-disclosed. Reported net income and diluted EPS for the March 2026 quarter include a large non-cash deferred-tax benefit; operating income is used as the cleaner earnings measure. Price, 52-week range, multiples and analyst consensus are vendor market data as of September 6, 2026; forward P/E uses FY2028 consensus EPS of $13.20. Insider figures cover the trailing 12 months; 10b5-1 plan status is not disclosed in the vendor feed. Documentation prepared with AI assistance. Not investment advice.

Fact check: Buyback rate corrected to ≈5% of market cap (FY2025 buybacks $8.8B vs Sep 2026 market cap $177B, not "more than 6%"); bet-block metaphors ("brains," "fill the hole") made literal per framework; all numerical claims verified against filed 10-Q and vendor data. Final analysis verified as of Sep 6, 2026.

Bid Cap

Daily ideas, a 390-name database, and a model long/short book from an investor who mostly covers financials. $70 a month or $700 a year.

Subscribe on Substack