Qualcomm (QCOM) at 20 Times Earnings, Minus Apple

Something like $7.5 billion a year is walking out of Qualcomm’s biggest customer relationship, and the company finally put a number on how fast. Apple’s share of the modems going into new iPhones is now landing well under the roughly 20% Qualcomm once expected to hold, and product revenue from Apple is set to fall about 50% sequentially from the September quarter to the December one. That is the scary part of the June-quarter report. The less scary part, the part the stock has not fully priced in, is that Qualcomm has spent two years building a replacement business, and the last quarter’s numbers are the first real evidence it is working.

The stock trades around $177.72 as I write this, 31.0% off its 52-week high of $258, at 20.3 times trailing earnings against a five-year average of 18.6. That is not a growth multiple. It is closer to what the market pays for a company it expects to shrink before it grows again, and on the handset side that expectation is correct.

My thesis: Qualcomm’s automotive and IoT segments are growing fast enough in dollar terms to absorb most of the Apple exit over the next two fiscal years, even though the December quarter will look ugly in isolation, because the replacement revenue is diversified rather than concentrated in one customer the way the chip business used to be.

Three numbers anchor that view: automotive up 61%, IoT up 9%, handset down 20%. The mix is shifting faster than the total revenue line lets on.

The quarter that beat while a customer left

Revenue for the June quarter was $9.9 billion, down -4% from a year earlier but ahead of what analysts had modeled going in. Non-GAAP earnings per share came in a shade under estimates. Underneath the headline, the mix had already shifted. Handset revenue inside the chip division fell 20% to $5.09 billion, a decline Qualcomm and outside analysts tied to industry-wide memory chip shortages compressing phone builds as much as to Apple, a supply squeeze I wrote about from the other side when Micron’s pricing power came back. Automotive revenue hit a record $1.59 billion, up 61% year over year and the 23rd straight quarter of double-digit automotive growth. IoT revenue rose 9% to $1.83 billion. Shares still fell on the print, which is the normal reaction when the biggest customer is leaving and the core phone business is soft in the same quarter, even if the total beat.

Why the Apple exit accelerated

Qualcomm and Apple settled a bruising patent fight in 2019 with a modem supply deal that always had an expiration date attached, since Apple has been developing its own modem silicon since at least 2018. What changed this summer was pace, not direction. Apple’s in-house modem reached volume faster than Qualcomm had been planning for, and management now expects fiscal 2027 Apple product revenue to come in under roughly $2 billion, down from a prior planning assumption near double that. Qualcomm’s own release framed the near-term math bluntly: the entire non-handset business, growing more than 60% year over year, is what management is counting on to replace the departing Apple revenue, not a single new customer win. That is a real change in tone from a year ago, when Apple commentary on earnings calls was still framed around “if,” not “when the share falls below 20%.”

Automotive is the piece with a track record

Automotive is the strongest evidence the replacement plan is more than a slide in an investor deck. Qualcomm’s Snapdragon Digital Chassis platform is now designed into vehicles across most major automakers, and management raised its automotive annualized revenue run-rate outlook to about $7 billion following the June quarter, up from prior guidance. Twenty-three consecutive quarters of double-digit growth is not a fluke tied to one design win; it reflects a multi-year backlog of auto contracts that convert to revenue as vehicles actually ship, which is a slower but stickier kind of growth than a phone refresh cycle. The company’s longer-range target is $40 billion in combined non-handset revenue by fiscal 2029, and automotive is the segment furthest along toward proving that number is more than aspiration.

Data center is still mostly a promise

Data center is the newer, less proven leg. Qualcomm has sized data center revenue at roughly $5 billion for fiscal 2027, scaling toward $15 billion by fiscal 2029, built around AI inference chips rather than the training silicon Nvidia and AMD dominate. That is a real market, but it is also a market Qualcomm has never shipped meaningful revenue into before, competing against incumbents with years of hyperscaler relationships already built. I would treat every dollar of that guidance as provisional until a specific hyperscaler customer and contract size shows up in a filing. This is the piece of the story I am least willing to take on faith, and it is also the biggest single swing factor in whether the $40 billion 2029 target is realistic or optimistic.

The honest risk here is sequencing. Automotive took years of design wins before the revenue showed up in a quarterly print, and data center is starting from a smaller base with less of that groundwork already laid. If the same multi-year lag applies, fiscal 2027’s $5 billion figure could prove optimistic even if the eventual $15 billion target is directionally right. I would rather see one named hyperscaler contract than another slide with a big total addressable market number on it.

The licensing business nobody argues about

Underneath the chip segment sits Qualcomm Technology Licensing, the patent royalty business that collects a cut of essentially every 4G and 5G device sold regardless of whose chip is inside it, Apple’s own modems included. That business does not care whether Apple buys Qualcomm silicon, only whether Apple sells phones with Qualcomm-patented radio technology inside them, which every iPhone on the market still does. It is a smaller, higher-margin business than the chip segment, and it is the one part of Qualcomm’s Apple relationship that survives the modem transition intact. Investors modeling a clean break with Apple’s revenue sometimes forget this piece keeps collecting regardless.

What the multiple already assumes

At 20.3 times trailing earnings and 32.3 times forward estimates, the market is pricing in the earnings hit from Apple, then some. Forward EPS estimates imply earnings actually falling versus the trailing figure, -37%, which tells you the “much lower” Apple guidance for fiscal 2027 is already baked into consensus rather than sitting as a surprise waiting to happen. The average analyst price target sits at $203, about 14% above the current price, with only 33% of the 24 analysts covering the stock rated buy, a split that tells you Wall Street is not fully convinced the replacement math closes cleanly. Compare that to Apple itself, trading at 38.5 times earnings with 60% of analysts at buy: the market is willing to pay a much richer multiple for the company losing the modem business than for the one gaining share in it, because Apple’s growth is diversified across a much larger base and Qualcomm’s is not yet.

Metric	QCOM	AAPL
Price	$177.72	$336.13
P/E (TTM)	20.3	38.5
5yr avg P/E	18.6	31.1
Revenue growth (FY2025)	14%	6%
Dividend yield	2.02%	0.31%
Analyst buy rating	33%	60%
Avg price target upside	14%	3%

Where the position sizing question comes in

Qualcomm’s beta has run high through this transition, and a stock swinging on segment mix the way this one does deserves the position-sizing discipline I laid out in how I think about volatile names, not a full-weight bet on either the bull or bear case. Short interest sits at 3.2% of the float, modest but not nothing, which suggests the skeptics have not fully capitulated even after the automotive numbers. The stock’s average earnings-day move has been 7.5%, and the actual move after the July 29 report was -2.6%, smaller than the historical average despite the Apple news, which is itself informative: the market had mostly priced the exit before the number arrived.

The counter-case and the number I would wait for

My case would be wrong if automotive growth slows from the low-60s percentage range toward something closer to 20% over the next two quarters, because that would mean the design-win backlog is thinner than 23 straight quarters of growth suggests, and the fiscal 2029 non-handset target would need to be pushed out. The single number I want to see next is the automotive revenue print for the December quarter: another sequential step up toward the $7 billion run-rate would confirm the backlog is converting on schedule; a flat or declining print would be the first real crack in the replacement thesis, not just a rough Apple-transition quarter. I would also watch whether Qualcomm names an actual data center customer by name in the next two earnings calls, the way it names automotive OEM wins today; silence on that front after two more quarters would tell me the $5 billion fiscal 2027 data center figure is softer than management’s tone suggests. For readers weighing Apple itself against the supplier it is walking away from, I looked at Apple’s own earnings reaction after a similar guidance surprise, which is a useful contrast in how differently the market treats the two companies’ misses.

Analysis and opinion only, not investment advice. Figures come from Qualcomm’s third-quarter fiscal 2026 results on SEC EDGAR and its investor site; Apple figures come from its filings on SEC EDGAR; valuation multiples are approximate and were checked on September 22, 2026.