QCOM is one of the best value stocks in the market right now

QCOM at $193: the market is handing you a 19x fwd toll booth with three free call options stapled to it

Been holding off posting this until the price got stupid. It got stupid. Down from $250 to $193 in two weeks and basically none of it was company specific. A hot payroll print killed rate cut hopes, Broadcom’s guide spooked the whole tape, Korea got smoked, Marvell dropped 10 percent in sympathy. Qualcomm just got dragged with the tide. Thats the setup value guys wait for, price dislocating from business.

The numbers:

Roughly $42B revenue this year, $12.63 in EPS, about $10B of free cash flow. Three legs:

Licensing: ~$5.3B a year at 72 percent margins. Patents on essentially every 5G phone on earth. This is the toll booth, it pays the 1.9 percent dividend by itself and it doesnt care what the chip cycle does. 6G standards land 2028-2030 which extends this annuity into the 2040s.

Chips: handsets are soft right now (memory shortage, China inventory burn, management already guided the bottom for this quarter with recovery after) but automotive just printed a record $1.33B quarter up 38 percent, guided +50 percent next quarter, exiting the year above a $6B run rate. Over a million cars on Snapdragon already, VW signed. This line alone backfills most of the Apple modem rolloff everyone obsesses over.

Capital return: $20B buyback authorization, $2.8B repurchased in one quarter, share count already down from 1.08B to 1.05B. At $193 every buyback dollar retires more shares. The drop literally works for you.

Why the earnings are about to inflect:

Apple may have just done Qualcomm the biggest favor in its history. They stood on stage and told a billion people that a phone without a serious AI chip and 12-16GB of RAM cant run the new software. They made the installed base feel obsolete. And Apple cant sell a single chip to the other 80 percent of the phone market. So when Samsung, Xiaomi, Honor and the rest respond over the next 12 months, theres one counter at the flagship window: Snapdragon 8 Elite, the highest ASP highest margin part Qualcomm makes, $60-80 a unit versus $25-40 midrange. The whole android ecosystem gets pushed upmarket whether it wants to or not. We ran this exact movie with 5G.

The data center kicker aka why june 24 matters:

Qualcomm already has real, inked business here: the ByteDance ASIC deal, Saudi Arabia’s Humain deploying 200MW of their AI200/AI250 racks, and a confirmed custom silicon engagement with a “leading hyperscaler” shipping this December that management is saving the name on for the june 24 investor day. The street’s consensus whisper, and multiple analysts have now said it out loud going into the event, is that the hyperscaler is Meta, consistent with SemiAnalysis reporting Meta as lead customer on Qualcomm’s Arm server CPU. Im treating it as high probability, not certainty. If Amon says that name on stage, this stops trading like a phone company the same afternoon.

JPMorgan isnt waiting. They took their target from $160 to $265 and put the stock on positive catalyst watch into the event, modeling ~$3B of data center revenue by FY27 scaling toward $35B by 2031. Thats a second company that wasnt in anyone’s model 18 months ago.

And the device layer is already won: Ray-Ban Metas, the new Meta display glasses, Snap Specs, Samsung and Google’s glasses, all Snapdragon. Category grew 211 percent last year. Meta is tuning Llama for Snapdragon directly. Theres also street chatter the OpenAI phone silicon work is tilting Qualcomm’s way after partner stumbles. Optionality, priced at zero.

Valuation (Rerate to peers upside is crazy):

~19x forward earnings. Broadcom 35x. Marvell 65x. ARM near triple digits. Semi industry median 34x. Same AI narrative, half the multiple or less. Yes the PEG looks ugly (~8) because consensus models ~3 percent growth, but thats precisely the bet: youre paying nothing for growth, so any growth surprise is pure multiple expansion. Cheapness is the margin of safety, the re-rate is the upside.

Bear case, addressed:

Apple modem exit: real, $7-8B over a couple years. Offset by auto’s ramp plus the buyback, which is why FY27 consensus is flat, not collapsing. Already in the price at 19x.

Nvidia’s PC chip: premium PC was maybe $1B of revenue. Half of it lost is ~2 percent of EPS. Stock fell 10 percent on the headline. Thats the inefficiency, not the risk.

“Broadcom owns data center”: owns accelerators, yes. Qualcomm is taking the inference, sovereign, and server CPU lanes in a market growing fast enough that the number three player still prints billions. $35B by 2031 is JPMs number not mine.

“Value trap”: correct call for five years. Auto at +38, the AI phone cycle, and june 24 are the specific things that break the assumption. Youre not paying for them anyway.

TLDR:

FY28 EPS ~$15 as auto, edge AI premiumization, and data center layer onto $12.63. At 24x, still a turn discount to Broadcom, thats ~$360, about 85 percent from here. Half-works case: 19x on $13 = $250, +29 percent. Downside support $175-180 where buyback, dividend, and licensing cash flow take over, ~8 percent below. Risking 8 to make 29 base, 85 bull, dated catalyst in two weeks.

Other catalyst thoughts (the gravy) (some repeats);

1. It’s the cheapest AI chip stock on the board. About 19 times forward earnings versus Broadcom at 35, Marvell at 65, and a semiconductor industry median around 34. Qualcomm trades roughly 44 percent below that median, cheaper than every AI name except a dead heat with Nvidia, while owning real edge AI exposure. 2. Qualcomm owns the AI device wave. Almost every pair of AI smart glasses and every decent smartwatch runs on Snapdragon. Meta Ray-Bans, Snap’s new Specs, Samsung, Google, Motorola, all of them. Smart glasses sales jumped 211 percent last year and Qualcomm is the default chip inside. Jensen Huang himself told investors to buy the stock. 3. The AI phone upgrade cycle is just starting. Phones more than two years old can’t run on-device AI well. A premium Snapdragon runs 60 to 80 dollars versus 25 to 40 for mid range, so as the NPU becomes the reason people upgrade, Qualcomm sells more chips and pricier ones. Same playbook as the 5G wave, which they won. 4. Automotive is quietly on fire. Revenue hit a record 1.33 billion last quarter, up 38 percent, with next quarter guided up around 50. They’re exiting the year above a 6 billion run rate, Volkswagen signed on, and over a million cars already run Snapdragon. This is the engine that backfills the Apple modem business as it rolls off. 5. The data center pivot is basically free in the stock. Humain in Saudi Arabia is deploying 200 megawatts, an unnamed hyperscaler ships this December, the ByteDance deal is millions of ASICs, and SemiAnalysis says Meta is the lead customer for their Arm server CPU. Futurum pegs it north of 10 billion in revenue over a few years, and after the selloff the market is paying close to nothing for it. 6. They’re buying back stock aggressively. A 20 billion authorization, 2.8 billion in a single quarter, share count already down from 1.08 to 1.05 billion. At these lower prices every dollar retires more shares, so the drop actually helps you. 7. The licensing business is a cash machine. Around 5.3 billion in royalties at 72 percent margins, patents on basically every 5G phone on earth, and 6G stretches that runway into the 2040s. It funds the roughly 10 billion in annual free cash flow, pays the 1.9 percent dividend, and covers everything else no matter what the chip cycle does. 8. You get a stack of free call options. The rumored OpenAI phone chip targets 300 to 400 million units a year by 2028. There’s the budget AI PC market with Snapdragon C, plus the whole robotics and physical AI push. None of that costs you anything at this price. 9. Catalysts are right in front of you. June 24 investor day is in two weeks, where they name the hyperscaler and lay out the data center roadmap. July 29 earnings is where China handsets bottom and start to turn. JPMorgan already put it on positive catalyst watch and moved their target to 265 from 160.