Built by Claude FableValuation, Concerns, Catalysts and Risks (5/10)
Bottom line: QCOM trades at roughly half the peer multiple on FY+1 earnings, and the next 90 days hold an unusually dense set of hard catalysts: the Arm trial verdict (in progress now), the FQ4'26 report in early November, first data center ASIC revenue in the December quarter, and the Apple and Huawei license renewals signed in the last two weeks. Against that, China ship-to revenue is about 46% of the total, the Apple modem roll-off sped up in FQ3'26, Samsung is pushing Exynos back into the Galaxy S27, and three separate regulatory and legal overhangs remain open (Arm, China SAMR, US export controls). The rubric puts the cheap valuation and catalysts against the China and regulatory exposure. That lands at a 5.
QCOM FY+1 is FY27 (Sep-27). Peers use the fiscal year closest to calendar 2027: AVGO FY Nov-27, MRVL FY Jan-28, ARM FY Mar-28, and CY27 for the others. Prices, EV and consensus come from FMP /stable (quote, key-metrics-ttm EV, analyst-estimates averages) as of 2026-10-08 intraday. Bloomberg and Visible Alpha were not available for this run.
| Metric | FY+1 Estimate | Multiple | Peer Avg |
|---|---|---|---|
| P/E (FY27 EPS) | $10.13 (21 analysts) | 17.3x | 36.8x (median 32.7x; ex-ARM 28.0x) |
| EV/EBITDA (FY27 EBITDA) | $15.29B | 12.7x | 28.1x (median 25.0x; ex-ARM 21.1x) |
| P/E, mature subset (NXPI, TXN, MediaTek) | 17.3x | 24.4x | |
| EV/EBITDA, mature subset | 12.7x | 18.1x | |
| P/E FY28 (EPS $13.07, 9 analysts) | 13.4x | n/a |
| Peer | FY+1 P/E | FY+1 EV/EBITDA |
|---|---|---|
| AVGO | 18.6x | 15.0x |
| MRVL | 37.7x | 25.0x |
| ARM | 89.5x | 70.1x |
| NXPI | 12.6x | 9.5x |
| TXN | 27.9x | 19.2x |
| AMD | 38.5x | 32.4x |
| MediaTek (2454.TW) | 32.7x | 25.6x |
- The peer set comes from Daloopa industry_id 62. FMP consensus for MediaTek has CY27 EPS more than doubling, so treat its multiple as noisy.
- The discount is partly earned. Street FY27 EPS ($10.13) sits below FY26 ($10.54). The drivers are the Apple step-down, memory-driven handset weakness, input-cost inflation and data center OpEx running ahead of revenue. Every peer in the table is growing EPS into FY+1. QCOM only screens cheap on a growth-adjusted basis if FY28 ($13.07, a thin 9 analysts) is real. That is the core debate.
- Only NXPI is cheaper. QCOM sits at about 0.5x the peer median on both P/E and EV/EBITDA.
- The FY24 drop in China share is mostly a reporting change, not real de-risking. US revenue jumped from 3.5% to 24.9% of the total in one year, and South Korea more than doubled. That points to a change in how Qualcomm attributes revenue (customer HQ and billing versus ship-to), not a shift in the business. Even on the new basis, China is 46% of revenue, more than 4x the rubric's 10% line. On the old basis it ran 62% to 67%.
- Customer concentration (10-K, unnamed in Daloopa): Customer X was 21% of FY25 revenue, Customer Y 20% and Customer Z 13%. The three together are 54% of revenue. Our reading is that X and Y are Apple and Samsung and Z is a Chinese Android OEM (likely Xiaomi). The filing does not name them in the Daloopa series, so this is inference.
- China OEM handsets are the swing factor. Management said China handset revenue bottomed in FQ3'26 and guided "double-digit" sequential growth in FQ4 with "very high confidence" (FQ3'26 call). Total handset revenue was $5,086M in FQ3'26, versus $6,328M a year earlier, down 19.6% y/y. That was the worst print in eight quarters. The cause was Chinese OEMs cutting builds and drawing down channel inventory as AI data center demand starved them of memory.
- Export control and Huawei. Commerce revoked QCOM's Huawei product export license in May 2024, so product revenue from Huawei is already zero. The live risk is broader: tighter rules on US silicon sold into Chinese OEMs, or Chinese retaliation. China's SAMR has had an open antitrust probe since 2025-10-10 over the unfiled Autotalks deal. The statutory fine cap is 10% of China sales, which would be up to roughly $2B on FY25 China revenue. The bigger risk is that the probe becomes leverage in a trade dispute. Status as of today is unclear from public sources.
| Concern | What changed in FQ3'26 (2026-07-29) | Size / evidence | Direction |
|---|---|---|---|
| Apple modem roll-off | The step-down now starts in FQ4'26, not FY27. Share of the new iPhone is "materially lower" than the prior 20% assumption. FY27 Apple product revenue falls below the earlier ~$2B+ | Management says FY27 non-handset growth of >60% "will replace the entire Apple product revenue" of FY26. Apple's December quarter is gone from the model | Worse, pulled forward |
| Samsung Galaxy S share | Snapdragon is in ~70% of flagship devices at Unpacked (FQ3'26 call), down from the ~75% expected for the S26 family (FQ1'26 call) | Reports say Exynos 2700 is in mass production for S27/S27+/S27 Pro in most regions, with Snapdragon only in the S27 Ultra plus US/Canada/Mexico | Worse for FY27 |
| Memory squeeze on Android | OEMs are buying prior-generation chips to offset memory BOM inflation. Handsets fell 19.6% y/y | A double-digit chip price increase is "small" next to the jump in memory cost (Amon) | Bottoming, per mgmt |
| Input-cost inflation | Wafers, packaging and memory costs rising. Price increases are coming, but the margin benefit lags | FQ4 QCT EBT guide is 23% to 25%, versus 26% in FQ3 | Worse near term |
| QTL licensing | QTL revenue was $1,278M, the lowest in 8 quarters, versus $1,318M a year earlier. EBT margin was 69%, versus 77% in FQ1'26 | Since the call, the Apple license was renewed (effective 2027-04-01, terms undisclosed) and a Huawei cross-license with 5G signed on 2026-10-05 (terms and direction of payment undisclosed) | Better: overhang largely removed |
| Arm litigation | Qualcomm already won the Nuvia ALA case (jury Dec-2024, full dismissal Sep-2025; Arm appealing). Qualcomm's own suit against Arm went to a 5-day Delaware jury trial on 2026-10-05 | Qualcomm is seeking to halt royalties to Arm for up to five years. The judge is weighing a cap on recovery | Asymmetric: upside if QCOM wins; the downside of the appeal is the larger tail |
| Data center execution | HBC Gen 1 taped out. Two hyperscaler ASIC programs start revenue in the December quarter. First merchant HBC product mid-2027. OpEx ~$2.7B in FQ4 includes Modular and Alphawave | Targets are $5B in FY27 and $15B in FY29 from a base of ~zero. Management concedes "investors want to see more proof points" | Unproven; it is a new-entrant execution risk |
| Custom-silicon competition | Broadcom and OpenAI's custom inference chip ramps in 2027 (local news cache) | Competes directly for the inference workloads behind QCOM's $15B FY29 target | Rising |
| Date | Catalyst | What to watch | Skew |
|---|---|---|---|
| Oct 2026 (verdict imminent) | QCOM v. Arm jury trial, Delaware | Whether QCOM wins royalty relief; any cap on damages | Positive optionality |
| Early Nov 2026 (date TBC) | FQ4'26 report and FQ1'27 guide | China handset rebound (guided double-digit q/q), Auto ~+60% y/y, Apple step-down size, first FY27 framing. Guide is $9.7B to $10.5B rev, $2.05 to $2.25 EPS; consensus $10.15B / $2.17 | Neutral to positive (China bottom is high-confidence) |
| Dec quarter 2026 | First data center ASIC revenue from two hyperscalers | The first proof point for the $5B FY27 target | Positive if on time |
| 2026 to 2027 | AI200 rack deployments (Humain 200MW starting 2026), AI250 in 2027, HBC mid-2027 | Named customers beyond Humain; disclosed revenue | Binary |
| Sep 2026 (done) | Snapdragon Summit: 8 Elite Gen 6 and 8 Elite Extreme Gen 6 on 2nm, agentic AI positioning, Xiaomi 18 Pro / Pro Max launch customers | Premium mix and ASP lift in an "agentic smartphone cycle" | Modest positive |
| Ongoing | Auto: BMW lead-compute win for next-gen ADAS plus cockpit (FQ3'26). Auto revenue $1,588M, versus $984M a year earlier, up 61% y/y | ADAS pipeline conversion | Positive, already visible |
| 2026 to 2027 | AI PC: X2 platforms, ~150 designs in 2026, Gemini Chromebook design-win share | Windows on Arm share past the ~9% of >$600 US retail laptops (CQ2'25) | Small |
| Pending | Huawei license closing and regulatory approval; Apple license economics from 2027-04 | QTL revenue run-rate in FY27 | Positive |
Yes, in both directions. The mix inflection is real. Non-handset QCT grew 28% y/y. Auto rose from $984M to $1,588M. IoT rose from $1,681M to $1,830M. And the FY29 non-handset target nearly doubled to $40B, including $15B from data center. But total revenue still fell 4.0% y/y, to $9,947M from $10,365M. AI demand is also hurting the core business through memory scarcity, and the Apple hole got bigger and arrived sooner. Growth reaccelerates only if the China bottom holds in FQ4 and data center revenue shows up on schedule in the December quarter. FY27 consensus EPS below FY26 says the street is not yet underwriting that.
- The China Android bottom in FQ3 is confirmed by a double-digit q/q rebound. Price increases restore QCT margins in 1H FY27.
- Non-handset growth of >60% in FY27 (auto ADAS ramp, industrial pipeline above $7B, $5B data center) fully offsets Apple. FY28 EPS of ~$13 then makes the stock 13.4x forward on a reaccelerating top line.
- Apple and Huawei license renewals secure QTL's ~70% margin stream. An Arm win cuts royalty cost.
- Management repeats the data center message the street does not believe. That is the "management is contrarian" setup in our principles, and the lukewarm consensus (34 Hold, 30 Buy/Strong Buy) leaves room for sentiment to invert.
- Handset revenue (still about half of QCT) is shrinking. Apple is going to zero and Samsung is moving S27 volume to Exynos. Chinese OEMs face both memory rationing and a domestic-champion push (Huawei Kirin, plus SAMR leverage).
- QCOM is a new entrant against NVIDIA, AVGO and MRVL in data center. $5B in FY27 from a near-zero base is a very large ask, and the data center business dilutes gross margin.
- With 46% of revenue in China and an open SAMR probe, any escalation in US-China trade hits QCOM before most peers.
- The low multiple is a value trap if FY28 estimates (9 analysts) prove too high. FY27 is already a down EPS year.
QCOM fails the China test outright: China is 46% of revenue on the current basis and 62% to 67% on the old ship-to basis, both far above the rubric's 10% line. It also carries real regulatory and legal overhangs: an open China SAMR probe, US export-control exposure, and a live Arm trial plus appeal. Those push the score toward 1. Two things pull it back. The valuation is clearly below peers (17.3x P/E and 12.7x EV/EBITDA on FY27, versus peer medians of 32.7x and 25.0x). And the near-term catalyst set is dense and mostly skewed positive: the Arm verdict, the FQ4 China rebound, first hyperscaler ASIC revenue in the December quarter, and the Apple and Huawei license renewals that remove QTL risk. FQ3'26 changed the story in both directions. The AI mix shift and the $40B FY29 non-handset target got bigger, but the Apple hole and the memory squeeze arrived sooner. With a cheap stock, real catalysts and heavy China and regulatory risk, this lands at the rubric midpoint, 5/10. The FQ4 report and December-quarter data center revenue are the evidence that would move it.
Sources: Daloopa fundamentals (QCOM company_id 157, linked inline); FMP /stable quote, key-metrics-ttm, analyst-estimates (2026-10-08); QCOM earnings call transcripts FQ4'24 to FQ3'26 (FMP); Qualcomm-Apple license renewal; Huawei-Qualcomm license; QCOM v. Arm trial; Arm ruling Sep-2025; SAMR Autotalks probe; Galaxy S27 Exynos 2700; Snapdragon Summit 2026; Huawei export license revocation; AI200 / Humain. Data sourced from Daloopa.