Qualcomm Incorporated
Combined average of Moat (AI Resilience), Growth, and Valuation scores.
Moat Score
Qualcomm's moat is anchored by an essential 5G/6G standard-essential patent (SEP) royalty franchise (QTL) that compounds with every smartphone shipped globally, plus chipset IP leadership in mobile, automotive, and (newly) data center AI inferencing. The mobile handset business is structurally cyclical and increasingly threatened by Apple's in-house modem transition, but the licensing royalty stream is extraordinarily durable.
Qualcomm's competitive position rests on regulatory lock-in (essential 5G/6G patents), proprietary chipset IP (Snapdragon SoC + Hexagon NPU), and high-stickiness automotive design wins that span 7-10 year vehicle development cycles:
- QTL: The 5G/6G Royalty Tax: Qualcomm Technology Licensing (QTL) collects royalties on essentially every 3G/4G/5G handset sold worldwide, including from competitors who do not use Qualcomm chips. The portfolio of 140,000+ patents covering cellular standards is the single most defensible IP asset in mobile communications. Cross-licensing agreements with Apple, Samsung, Huawei, and Xiaomi have all been renewed through the late 2020s. As the industry moves toward 6G standardization (2028-2030), Qualcomm's continued contribution to 3GPP standards extends the royalty runway by another decade.
- Snapdragon: The Premium Android Default: Snapdragon 8 Gen 4 and the recently launched Oryon CPU architecture have re-established Qualcomm as the performance leader in premium Android SoCs. Samsung's Galaxy S26 launched globally on Snapdragon (vs. prior Exynos split), validating the platform advantage. Hexagon NPU integration enables on-device generative AI features that competing chipsets struggle to match at comparable power envelopes.
- Automotive: From Infotainment to Full ADAS: Qualcomm's Snapdragon Digital Chassis has accumulated a design-win backlog exceeding $45B across major OEMs (BMW, Mercedes, GM, Hyundai, Stellantis). Each automotive design win is a 5-7 year revenue stream with high margins and high switching costs once embedded in the vehicle E/E architecture. Q2 FY2026 set a record for auto chip revenue, and the segment is on track to be a $10B+ business by FY2029.
- AI200/AI250: The Inference-First Data Center Bet: Qualcomm announced AI200 (2026) and AI250 (2027) rack-scale data center accelerators based on the Hexagon NPU architecture, optimized for inferencing workloads with 768GB LPDDR per card and direct liquid cooling. Saudi Arabia's HUMAIN (PIF-backed) committed to 200MW of AI200 racks starting in 2026 — the first marquee customer. This is a credibility play in the inference-tier AI accelerator market currently dominated by NVIDIA and AMD; success is not guaranteed but optionality is meaningful.
Moat Verdict
Qualcomm's moat structure is led by regulatoryLockIn (essential 5G/6G patents), businessLogic (modem + Hexagon NPU IP), and talentScarcity (RF/standards engineers). These moats are highly AI-resilient — generative AI does not threaten cellular standards, and on-device AI actually plays into Hexagon's strengths. The structural risks are not AI-related: they are (1) Apple's continued modem in-housing, which compresses QCT handset revenue, and (2) execution risk on data center AI inference (AI200/AI250) where NVIDIA and AMD are entrenched. The QTL royalty stream alone justifies a meaningful floor valuation; the question is whether automotive + AI compound on top of a stable handset business or merely backfill against Apple-driven mobile decline.
49.9 resilient · 63.2 vulnerable · 80/20 = 52.6 · = 53
Open a moat to read its note.
Snapdragon developer SDKs and the Hexagon NPU toolchain create some learned-interface stickiness for OEM SoC integration teams, but the moat is modest — competing SoCs (MediaTek, Apple Silicon, Samsung Exynos) have comparable developer tooling.
Qualcomm's modem and RF front-end designs embody decades of cellular signal-processing know-how, and the Hexagon NPU carries similar low-power inference IP. Apple now ships its own C1/C2 modem in part of the iPhone lineup, which shows a well-funded entrant can rebuild that know-how given years, so it rates intact rather than strong.
Qualcomm is a chip and IP licensing company, not a data platform; controlling access to a public data source is not part of the business model.
RF/modem engineering and standards-committee participation talent is genuinely scarce globally — the 3GPP standardization process has only a few hundred engineers worldwide qualified to contribute meaningfully. Qualcomm employs a large fraction of them. Custom NPU and Oryon CPU teams (acquired via Nuvia) compound this advantage.
Snapdragon SoC + Snapdragon X modem + RF front-end + WiFi/BT combo chips are sold as a bundled platform to handset OEMs, raising switching costs and giving Qualcomm pricing leverage. Snapdragon Digital Chassis bundles compute + connectivity + ADAS for autos.
Qualcomm has telemetry from billions of Snapdragon-powered devices but does not aggregate or monetize this as a data product. Modest secondary advantage at best — the company's value is in chip design IP and patents, not in data network effects.
Qualcomm's 140,000+ patents covering 3G/4G/5G/6G cellular standards are essential to industry compliance — every handset OEM globally must license them. FCC/CE-equivalent regulatory certifications for cellular modems require extensive testing that Qualcomm has already cleared. This is the single most durable moat in the company.
Modest indirect network effects via Snapdragon developer ecosystem (Qualcomm AI Engine Direct SDK, game developer relations), but these are not a primary moat. The cellular standards process itself creates a kind of network effect for the SEP holder, but that's already captured in regulatory lock-in.
Snapdragon chips are physically embedded in 70%+ of premium Android handsets and an expanding share of vehicles, but they are not embedded in the financial transaction layer. Some embedding via secure enclave / SE for mobile payments, but this is not a primary moat.
Qualcomm does not maintain a system-of-record platform; it sells chips and licenses IP.
Fabless or equipment franchise whose edge is IP and process know-how, rated elsewhere; manufacturing volume does not give it a unit-cost lead rivals cannot buy.
Buyers are enterprises choosing on switching cost, integration and performance, which the other pillars rate. The name carries reputation, not a price premium it could hold on brand alone.
Combined average of Moat (AI Resilience), Growth, and Valuation scores.
Moat Score
Qualcomm's moat is anchored by an essential 5G/6G standard-essential patent (SEP) royalty franchise (QTL) that compounds with every smartphone shipped globally, plus chipset IP leadership in mobile, automotive, and (newly) data center AI inferencing. The mobile handset business is structurally cyclical and increasingly threatened by Apple's in-house modem transition, but the licensing royalty stream is extraordinarily durable.
Growth Score
Q3 FY2026 (June quarter) revenue $9.9B (-4% YoY), at the high end of guidance; net income $2.0B (-25% YoY). QCT handsets fell 20% YoY to $5.1B on memory-constrained Android builds and softer China OEM demand, while automotive set another record at $1.6B (+61% YoY) and QCT automotive + IoT together grew 28%. The September quarter is guided to $9.7–10.5B, with handsets ~$5.2B as lower Apple product revenue offsets sequential Android growth, and automotive guided to another record at ~+60% YoY. Data center revenue begins ramping in the December quarter, and Qualcomm raised its FY2029 non-handset ambition to >$24B across automotive and IoT plus >$15B in data center. The Apple modem roll-off is now in the numbers rather than ahead of them; the open questions are how fast the non-handset lines can outgrow a shrinking handset base, and when elevated memory input costs stop pressuring QCT margins.
Valuation Score
At ~$194 (September 24, 2026), Qualcomm trades at ~18× FY2026 consensus non-GAAP EPS (~$10.5–10.7) and ~18–19× FY2027 (~$10.1–10.8) — no longer the ~14× bargain the May review described, because the earnings base itself has been cut. Q3 FY2026 non-GAAP EPS fell 20% YoY to $2.21 on revenue of $9.9B (-4%), handset revenue fell 20% on OEM inventory drawdowns and memory costs, Q4 is guided to $2.05–2.25, and management now expects its share of the new iPhone to be materially below its prior 20% estimate. The price sits just above the base target ($190), ~49% above the bear ($130) and ~27% below the bull ($265): automotive (record $1.6B, +61% YoY) and data-centre optionality are what the multiple pays for, against a flat 2027 EPS consensus.
The Standard-Essential Patent Royalty Engine
Qualcomm's competitive position rests on regulatory lock-in (essential 5G/6G patents), proprietary chipset IP (Snapdragon SoC + Hexagon NPU), and high-stickiness automotive design wins that span 7-10 year vehicle development cycles:
- QTL: The 5G/6G Royalty Tax: Qualcomm Technology Licensing (QTL) collects royalties on essentially every 3G/4G/5G handset sold worldwide, including from competitors who do not use Qualcomm chips. The portfolio of 140,000+ patents covering cellular standards is the single most defensible IP asset in mobile communications. Cross-licensing agreements with Apple, Samsung, Huawei, and Xiaomi have all been renewed through the late 2020s. As the industry moves toward 6G standardization (2028-2030), Qualcomm's continued contribution to 3GPP standards extends the royalty runway by another decade.
- Snapdragon: The Premium Android Default: Snapdragon 8 Gen 4 and the recently launched Oryon CPU architecture have re-established Qualcomm as the performance leader in premium Android SoCs. Samsung's Galaxy S26 launched globally on Snapdragon (vs. prior Exynos split), validating the platform advantage. Hexagon NPU integration enables on-device generative AI features that competing chipsets struggle to match at comparable power envelopes.
- Automotive: From Infotainment to Full ADAS: Qualcomm's Snapdragon Digital Chassis has accumulated a design-win backlog exceeding $45B across major OEMs (BMW, Mercedes, GM, Hyundai, Stellantis). Each automotive design win is a 5-7 year revenue stream with high margins and high switching costs once embedded in the vehicle E/E architecture. Q2 FY2026 set a record for auto chip revenue, and the segment is on track to be a $10B+ business by FY2029.
- AI200/AI250: The Inference-First Data Center Bet: Qualcomm announced AI200 (2026) and AI250 (2027) rack-scale data center accelerators based on the Hexagon NPU architecture, optimized for inferencing workloads with 768GB LPDDR per card and direct liquid cooling. Saudi Arabia's HUMAIN (PIF-backed) committed to 200MW of AI200 racks starting in 2026 — the first marquee customer. This is a credibility play in the inference-tier AI accelerator market currently dominated by NVIDIA and AMD; success is not guaranteed but optionality is meaningful.
Moat Verdict
Qualcomm's moat structure is led by regulatoryLockIn (essential 5G/6G patents), businessLogic (modem + Hexagon NPU IP), and talentScarcity (RF/standards engineers). These moats are highly AI-resilient — generative AI does not threaten cellular standards, and on-device AI actually plays into Hexagon's strengths. The structural risks are not AI-related: they are (1) Apple's continued modem in-housing, which compresses QCT handset revenue, and (2) execution risk on data center AI inference (AI200/AI250) where NVIDIA and AMD are entrenched. The QTL royalty stream alone justifies a meaningful floor valuation; the question is whether automotive + AI compound on top of a stable handset business or merely backfill against Apple-driven mobile decline.
49.9 resilient · 63.2 vulnerable · 80/20 = 52.6 · = 53
Open a moat to read its note.
Snapdragon developer SDKs and the Hexagon NPU toolchain create some learned-interface stickiness for OEM SoC integration teams, but the moat is modest — competing SoCs (MediaTek, Apple Silicon, Samsung Exynos) have comparable developer tooling.
Qualcomm's modem and RF front-end designs embody decades of cellular signal-processing know-how, and the Hexagon NPU carries similar low-power inference IP. Apple now ships its own C1/C2 modem in part of the iPhone lineup, which shows a well-funded entrant can rebuild that know-how given years, so it rates intact rather than strong.
Qualcomm is a chip and IP licensing company, not a data platform; controlling access to a public data source is not part of the business model.
RF/modem engineering and standards-committee participation talent is genuinely scarce globally — the 3GPP standardization process has only a few hundred engineers worldwide qualified to contribute meaningfully. Qualcomm employs a large fraction of them. Custom NPU and Oryon CPU teams (acquired via Nuvia) compound this advantage.
Snapdragon SoC + Snapdragon X modem + RF front-end + WiFi/BT combo chips are sold as a bundled platform to handset OEMs, raising switching costs and giving Qualcomm pricing leverage. Snapdragon Digital Chassis bundles compute + connectivity + ADAS for autos.
Qualcomm has telemetry from billions of Snapdragon-powered devices but does not aggregate or monetize this as a data product. Modest secondary advantage at best — the company's value is in chip design IP and patents, not in data network effects.
Qualcomm's 140,000+ patents covering 3G/4G/5G/6G cellular standards are essential to industry compliance — every handset OEM globally must license them. FCC/CE-equivalent regulatory certifications for cellular modems require extensive testing that Qualcomm has already cleared. This is the single most durable moat in the company.
Modest indirect network effects via Snapdragon developer ecosystem (Qualcomm AI Engine Direct SDK, game developer relations), but these are not a primary moat. The cellular standards process itself creates a kind of network effect for the SEP holder, but that's already captured in regulatory lock-in.
Snapdragon chips are physically embedded in 70%+ of premium Android handsets and an expanding share of vehicles, but they are not embedded in the financial transaction layer. Some embedding via secure enclave / SE for mobile payments, but this is not a primary moat.
Qualcomm does not maintain a system-of-record platform; it sells chips and licenses IP.
Fabless or equipment franchise whose edge is IP and process know-how, rated elsewhere; manufacturing volume does not give it a unit-cost lead rivals cannot buy.
Buyers are enterprises choosing on switching cost, integration and performance, which the other pillars rate. The name carries reputation, not a price premium it could hold on brand alone.
Growth Analysis
Growth Drivers
Key Risk
With the Apple modem roll-off already in the handset line, the residual downside is on the two things the diversification case still assumes: if Samsung moves a larger share of the Galaxy S27 lineup to Exynos, handsets fall further than the current reset, and if the data center ramp that starts in the December quarter has not produced a visible revenue line by the end of FY2027, the >$15B FY2029 data center target — and the non-handset growth that is meant to offset the handset decline — loses credibility.
Score Derivation
62.5 base + 0.8 trajectory − 4 margin − 5 risk = 54
Base 62.5 (3–7% CAGR, midpoint 5%, baseFromCagr) + 0.8 trajectory (automotive and data center accelerating, handsets decelerating, IoT and QTL stable: (2 − 1) / 5 × 4) − 4 compressing margin (QCT EBT margin 26% in Q3, guided 23–25% for Q4; gross margin guided slightly below the historical range on elevated memory input costs; net income −25% YoY) − 5 moderate keyRisk. Severity moves high → moderate because the Apple half of the old risk has materialised: handsets −20% YoY and a Q4 handset guide carrying lower Apple revenue are now charged in the cagrEstimate and the handset driver's decelerating trend, and leaving severity at high would charge them twice. What remains unmaterialised is a Samsung Exynos share shift and the data center ramp missing its targets. The old author string (base 55, +5 automotive backlog, +5 data center optionality) did not follow the formula, which computed 60 on the old inputs = 54
Price Scenarios (12–24 Months)
Valuation Multiples
| Forward P/E (FY2026) | ~18× |
| Forward P/E (FY2027) | ~18–19× |
| Q4 FY2026 EPS Guide | $2.05–2.25 |
| Market Cap | ~$205B |
| Consensus Price Target | ~$193–204 |
QCOM no longer screens cheap: FY2026 EPS estimates have fallen to ~$10.5–10.7 and FY2027 is expected to be roughly flat as the Apple modem step-down accelerates and Android OEMs work down inventory. At ~18× that trough-ish EPS, the stock is priced near fair value; upside needs automotive and data-centre revenue to lift FY2027–28 EPS above a flat consensus, and downside is a further leg of handset estimate cuts.
Approximate figures as of September 2026.
Where We Are vs Targets
Loading live price…
Apple completes full modem in-housing, Samsung returns to Exynos, AI200 fails to win additional marquee customers beyond HUMAIN, and the QCT handset business shrinks to ~50% of total revenue.
- Apple ships its own modem in 100% of iPhone lineup by FY2027, removing $4-5B/year of QCT handset revenue and ~$1B of operating profit
- Samsung Galaxy S27 returns to majority Exynos, eroding Snapdragon's premium Android share
- AI200 ramp disappoints — HUMAIN remains the only meaningful customer through 2027 — and Qualcomm writes down a portion of the data center investment
- FY2027 EPS slips to ~$10 (the low end of consensus) and the multiple compresses to ~13× as Qualcomm is re-rated as a structurally declining mobile business with optionality that did not materialize — implying ~$130
Handset revenue stabilizes after the Q3 FY2026 inventory trough, the accelerated Apple step-down is absorbed, automotive keeps compounding, and FY2027 EPS lands near the ~$10.5 consensus midpoint at a multiple slightly above Qualcomm's historical median.
- China OEM handset revenue troughs in Q3 FY2026 and returns to double-digit sequential growth in Q4 as channel inventory drawdowns ease, as management guided
- Automotive revenue keeps compounding off a record $1.6B quarter (+61% YoY) as the design-win backlog converts to production
- QTL royalty stream remains stable as 5G adoption deepens in emerging markets, partly offsetting the Apple modem step-down
- ~18× FY2027 consensus EPS of ~$10.5 ≈ $189 — about one turn above the ~17× five-year median, reflecting trough Apple-exit earnings and a richer auto/data-centre mix — fair value ~$190
AI200/AI250 emerges as a credible NVIDIA inference-tier alternative, automotive scales faster than $10B/yr, on-device AI drives premium Android ASP expansion, and Qualcomm re-rates to peer-average semi multiples on durable diversification.
- AI200/AI250 capture 5-10% of the rack-scale AI inference market by 2028 (a $20B+ TAM by then), driving $3-5B of incremental high-margin data center revenue
- Automotive accelerates to $10B+ run-rate by FY2028 (one year ahead of plan) as ADAS Level 3+ ships across the BMW, Mercedes, and Stellantis fleets
- On-device generative AI becomes a premium Android differentiator, lifting Snapdragon 8-series ASPs by 15-20% and partially offsetting Apple modem revenue loss
- Multiple expands to ~22× on FY2027 EPS of ~$12 (≈11–19% above consensus as handsets recover and data-centre revenue arrives) — implying ~$265