QCOM · QUALCOMM Incorporated · Stock research

Current Thesis

The data-center re-rating was attempted and rejected. Qualcomm's ~2026-06-24 Investor Day put a hard number on the pivot $15B of data-center revenue by 2029 and the stock spiked into a 52-week high of $259.92. Seventeen sessions later it closed 2026-07-17 at $171.78, roughly 34% below that high, with a $181B market cap and a 17.4x forward multiple. Nothing about the 2029 guide changed in those three weeks. What changed is that the Street re-weighted the near-term math: Apple's in-house modem transition takes Qualcomm's iPhone 18 modem share to about 20% in fiscal 2027 and Apple chip revenue to roughly $2B, down from more than $6B two years earlier, against a full-year FY26 revenue base of about $42.4B. Revenue leaves on a known schedule; the replacement is purchase orders and two unnamed hyperscaler conversations. A stock that gives back an entire catalyst move in under a month is telling you the catalyst was the exit liquidity for that leg. Consensus now sits at $221.90 against a $171.78 tape the same targets that clustered below price in late June now sit 29% above it, because the analysts did not move and the price did. That inversion is the setup's whole problem: it makes the name look cheap without any structural evidence that the decline is finished, and it arrives seven trading days before a binary print.

Bullish and bearish views on QUALCOMM Incorporated

The model's bull view on QUALCOMM Incorporated (QCOM), in brief: 2026-06-24/25 Investor Day: management guided to $15B of data-center revenue by 2029 the first quantified, multi-year number attached to the pivot rather than a product preview. The bear view: The Apple hole is arithmetic, not sentiment. Both cases follow in full.

Bull Case

• 2026-06-24/25 Investor Day: management guided to $15B of data-center revenue by 2029 the first quantified, multi-year number attached to the pivot rather than a product preview.
• 2026-07-06: Benchmark's Cody Acree reaffirmed a street-high $300 target after a fireside chat with senior management, citing fiscal-2027 earnings supported by purchase orders, custom chips already in production, and active discussions with two hyperscale customers. Stock rose 5.80% that session to $186.48.
• 2026-07-13: TD Cowen maintained Buy and raised its target to $225 a rating-plus-target move, not the Neutral-maintain pattern that dominated the June cluster.
• Valuation reset is real: 17.4x forward earnings, a $181B cap, and a $3.68 annual dividend (2.14% yield) on a business that did $44.28B of FY25 revenue (+13.66%).
• 2026-07-16: Micron announced strategic customer agreements with Tier-1 global automotive suppliers including Qualcomm and Hyundai Mobis evidence the automotive silicon line, the second diversification leg, is signing real supply commitments.
• 2026-05-28 Snapdragon C launch widened the Windows-on-ARM funnel into entry-level laptops, extending the PC attach story into 2026 shelves.

Bear Case

• The Apple hole is arithmetic, not sentiment. CEO Cristiano Amon has said Qualcomm supplies modems for roughly 20% of iPhones in 2026 and effectively none by 2027; Apple historically represented about a fifth of total revenue. A ~$4B annualized gap opens before the data-center line ships at scale.
• 2026-07-01/02: a cluster of moves to Hold, with cautious updates from Bernstein and Mizuho, flagged execution and margin risk from redirecting resources out of the smartphone franchise into an unproven AI-silicon business.
• 2026-07-02: Elon Musk called reports of a SpaceX AI phone using Snapdragon silicon "utterly false," and the stock fell 5.10%. A name that drops 5% on the removal of an unconfirmed rumor was carrying rumor premium.
• 2026-07-14: the ARM/Nuvia licensing dispute resurfaced as a live overhang on the Snapdragon X Elite line, with royalty-escalation or injunction risk attached to the PC franchise. Shares fell 3.2% to $178.10.
• 2026-07-16: QCOM fell 4.14% to $170.61 alongside MU -5.80%, SNDK -9.98% and NVDA -2.46% it traded as a high-beta casualty of a memory/AI-chip drawdown without owning the memory upside on the way up.
• Consensus Q3 FY26 revenue of $9.667B is a sequential step down from Q2 FY26's $10.6B, and FY25 earnings already fell 45.37% to $5.54B. The multiple is low because the earnings line is contested.
• Analyst dispersion runs $100 to $300 with a Hold consensus across 36 analysts. A 3x spread between the low and high target is not a mispricing to arbitrage; it is a market saying nobody can underwrite fiscal 2027.

Setup & Price Structure

Price structure is a failed breakout that has become a downtrend. The $248–250 zone that capped the tape through late May was reclaimed on the Investor Day spike to $259.92, then lost entirely the classic shape of a catalyst that marks a high rather than starts a leg. The 2026-07-06 bounce to $186.48 on the $300 target reaffirmation printed a lower high and failed within four sessions. From there: $178.10 on 2026-07-14, $170.61 on 2026-07-16, $171.78 on 2026-07-17, with an intraday range that has probed down to $164.77. RSI sits near 40 with MACD at -2.481 not oversold enough to bounce mechanically, not weak enough to be a capitulation low. The 52-week floor is $121.99, set earlier in 2026, and there is little visible shelf between the mid-$160s and that zone.

On the beginner-trap matrix this scores badly in the two ways that matter. First, it is squarely in averaging-down territory: a familiar large-cap 34% off its high, at 17.4x forward with a 2.14% dividend and a consensus target 29% overhead, generates exactly the "it's cheap now" impulse that the playbook exists to refuse. A 2.14% yield does not offset a 34% drawdown. Second, earnings land 2026-07-29 inside the window where any position taken here is a bet on a print, not on a trend. There is no cluster confirmation to lean on either: the semiconductor complex sold off with it on 2026-07-16, so the peer group is confirming weakness rather than strength. A base-building case only earns attention after price holds above the mid-$160s, reclaims $186.48 on expanding volume, and prints a higher low none of which exists today.

Catalyst Calendar (next 30 days)

• 2026-07-29 (confirmed): FY26 Q3 results after the close, conference call 1:45 p.m. Pacific. Consensus EPS $1.40, revenue $9.667B versus $10.6B in Q2 FY26. This is the binary.
• 2026-07-29 (confirmed, same release): FY26 Q4 guidance. The guide, plus any quantified data-center booking or backlog figure and updated FY27 Apple modem-share commentary, is the actual mover the reported quarter is history.
• ~2026-07-30 (est.): Apple fiscal Q3 results. Any commentary on in-house modem timing is a direct read-through to the Qualcomm royalty and chip line.
• ~2026-08-10 (est.): TSMC monthly revenue disclosure a coincident read on high-end mobile and AI silicon order flow.
• ~2026-08 (est.): $0.92 quarterly dividend ex-date. Income event, not a price catalyst; do not confuse it with a reason to own the drawdown.

Elapsed catalysts

• Undated: ARM/Nuvia litigation docket movement. No scheduled date, so it cannot be positioned around it is a tail risk that shows up unannounced, as it did on 2026-07-14. _(passed 5d ago)_

What Would Change Our Mind

The constructive case requires structure, not a better story. Concretely: price holds the mid-$160s on a weekly closing basis, then reclaims and closes above the 2026-07-06 lower high of $186.48 on volume above the 20-day average, establishing a higher low in between. On the fundamental side, the 2026-07-29 call would need to convert the 2029 target into near-term evidence a named hyperscaler, a disclosed data-center backlog or booking number, or an FY27 revenue framing that shows the pivot filling the Apple gap rather than following it. A Q4 guide above the $9.667B-ish run rate with data-center revenue broken out as a line item would be the strongest single tell. Absent those, a bounce toward $186 is a lower-high retest in a downtrend and should be treated as such.

The bear case deepens if the print reiterates 2029 targets without near-term numbers, if FY27 Apple modem share is revised below the ~20% already communicated, or if the ARM/Nuvia matter produces an adverse procedural ruling on the Snapdragon X line. Any of those, with price already below all the June structure, argues for the $122 area as the operative reference rather than the $221.90 consensus target.

Correlation Notes

• Memory/AI-chip complex (MU, SNDK, NVDA): on 2026-07-16 QCOM -4.14% traded alongside MU -5.80%, SNDK -9.98%, NVDA -2.46%. QCOM now carries downside beta to the memory tape without the memory revenue an asymmetry that hurts on drawdown days.
• MU (commercial, not just correlated): the 2026-07-16 Tier-1 automotive supply agreements name both Qualcomm and Hyundai Mobis, tying the two through the auto silicon channel.
• AAPL: direct inverse exposure. Apple's modem insourcing schedule is the single largest determinant of Qualcomm's FY27 revenue base; Apple product and earnings commentary moves QCOM independently of the AI narrative.
• ARM: the Nuvia-derived IP dispute makes ARM a legal counterparty, not just a peer. Litigation headlines transmit directly into the Snapdragon X PC franchise.
• NVDA: competitive rather than correlated on the AI-PC axis the RTX Spark launch (2026-06-01) and Jensen Huang's stated CPU ambitions contest Snapdragon-X territory, and Qualcomm was absent from the named beneficiary list around that launch (Dell, TSMC, MediaTek, Microsoft).
• AVGO / MRVL: the custom-silicon comparables Qualcomm is asking to be valued against. Relative performance versus these two is the cleanest live scoreboard on whether the data-center pivot is being believed.