Mobileye
Operational Inflection Meets the Intel Cap

Snapshot: Executive Summary

Mobileye Global Inc. is past the inflection of a semiconductor cyclical recovery that has continued into 2026, layered against a four-front structural competitive squeeze and a persistent Intel ownership overhang. The operating trajectory has cooled: FY25 revenue landed at ~$1.89B (+15%) after three guidance raises, and Q1 2026 extended it (revenue $558M, +27%), but Q2 2026 was roughly flat, revenue $508M (+0.4%), adjusted operating income $155M, with management raising the FY26 revenue midpoint to ~$2.0B, though the raise and the Q2 adjusted-EPS beat ($0.19) were flattered by a ~$93M one-time Israeli R&D tax credit¹. The catch is a $3.788B non-cash goodwill impairment in Q1 (a write-down of the 2017 Intel-acquisition value) that drove a GAAP loss and helped halve the stock, and now founder-CEO Amnon Shashua has announced his departure with no successor named, adding key-person and governance risk. The product setup remains the strongest since the October 2022 IPO: three Surround ADAS Western OEM wins (VW Group mainstream brands, an unnamed second Western OEM, and Mahindra across 6+ models from 2027), Polestar 4 SuperVision in production, MOIA's L4 ID.Buzz robotaxi in Hanover pre-series production, and a 2028 L3 Imaging Radar nomination by a "leading global automaker."

The operational inflection collides with a competitive reality that has fundamentally narrowed Mobileye's addressable market. The premium L2+/L3 race in Western markets has effectively bypassed SuperVision: NVIDIA Drive Thor owns Mercedes-Benz, BYD, XPeng, Li Auto, Volvo, JLR, Lucid, NIO, Rivian, and Zoox⁴; Qualcomm Snapdragon Ride Pilot debuts in BMW iX3 (2026 MY) and has been opened to all OEMs and Tier-1s; Tesla's AI5 vertical integration permanently closes the highest-value socket. In China, Horizon Robotics' 47.7% domestic basic ADAS share (+58% FY25 revenue growth), Black Sesame's 1H25 +40% YoY growth, BYD's in-house Xuanji A3 chip on the God's Eye platform (>2.3M installed), and Geely's Haohan 2.0 ejection of SuperVision from Zeekr 001 collectively confirm the in-house-silicon trend has reached the largest volume OEMs. SuperVision unit guidance of only ~150K in 2026: essentially flat YoY with Porsche/Audi pushed from late 2026 to Q1 2027, confirms the aspirational 1.2M-unit narrative lacks a near-term customer-by-customer bridge.

The Intel ownership structure compounds the multiple ceiling. Per the 2026 proxy and Q1 2026 filings, Intel still controls Mobileye with ~77% economic / ~96.9% voting power (50M Class A + all 597.77M Class B at 10 votes per share, convertible 1:1)⁸. Under Intel CEO Lip-Bu Tan's "use the cash" divestiture mandate (Altera 51% sold to Silver Lake for $4.46B cash as the template), Intel executed a July 2025 secondary of 50M shares at $16.50, roughly one-quarter the 2023 secondary price, with a concurrent Class B-to-A conversion shrinking the high-vote pool. Three sitting Intel executives plus the chair direct Mobileye's board. Every quarter carries secondary-announcement risk (historically -5-10% on print), and the goodwill impairment is the accounting confirmation of the de-rating that overhang has driven.

We rate MBLY Hold with a 12-month price target of $10. At ~$8.73 that is ~15% upside, coherent with a Hold and deliberately below the ~$13 Street average given the CEO's announced departure, the soft Q3 guide, and Intel's ~96.9% voting control. The distribution is genuinely two-sided: the cyclical-recovery thesis is only partly in motion (FY25 +15% and Q1 +27%, though Q2 flattened to +0.4%; guidance raised) yet caps at a discounted multiple the Intel overhang and the impairment signal prevent from re-rating toward the AMBA 7-8x peer level; the competitive-erosion thesis is real but offset by a cash-rich balance sheet and a durable base ADAS franchise. Probability-weighted blended FV of $9.40 (Bull $13 / Base $10 / Bear $6 at 20/50/30 weights). The stock is cheap but range-bound; we would upgrade on a Chauffeur production-OEM win or a credible Intel monetization framework that lifts the overhang.

Tactical: MBLY is trading at ~$8.73, ~13% below our $10 12-month target. The stock is cheap (~3x forward EV/Sales) and cash-rich against a guidance-raising business, but the recovery stalled in Q2 (flat revenue, a credit-flattered EPS beat, soft Q3 outlook) and the gap stays open until a binary catalyst clears the structural overhang (a Chauffeur production-OEM announcement, a credible Intel divestiture framework, CEO-succession clarity, or China stabilization). Rating Hold; tactical positioning requires an event-driven thesis rather than a directional view.

Investment Thesis

Rating: Hold. Probability weights: 20% Bull / 50% Base / 30% Bear → blended fair value ~$9.40 against the ~$8.73 reference price. The Hold reflects a stock that is cheap and recovering but structurally capped: the three lenses (cyclical-recovery, competitive-erosion, Intel-overhang) converge on a fair value modestly above spot with two-sided risk. The 50% base weight reflects high conviction in the operational recovery but moderate conviction the multiple re-rates while Intel controls ~96.9% of the vote and a $3.8B impairment sits on the books; 30% bear weight respects the four concurrent competitive threats plus Intel sell-down risk; 20% bull weight respects the Chauffeur and Intel-monetization optionality. We would move to Buy on a catalyst that clears the overhang.

Business Overview

Mobileye is a vertically-integrated ADAS and autonomous-driving supplier built around the EyeQ system-on-chip family, proprietary computer-vision silicon bundled with software stacks, REM (Road Experience Management) crowdsourced HD mapping data, and turnkey driving systems. EyeQ SoC sales accounted for 86% of FY2024 revenue, with SuperVision contributing most of the remainder¹. Mobileye has shipped >230M EyeQ chips cumulatively across 1,200+ vehicle models and 50+ OEMs since founding⁷, an installed base that anchors a powerful data flywheel.

Revenue history: the destock-and-recovery V

EyeQ unit shipments: the leading volume indicator

SuperVision unit trajectory: the premium-tier story

Product stack: four layers of autonomy

• Base ADAS: front-camera and safety-mandate compliance running on EyeQ4 (legacy) and EyeQ6 Lite (7nm, 5 DL TOPS int8, ~46M lifetime program units). The high-volume legacy business.
• Surround ADAS: newer mid-tier on EyeQ6 High at $100-150 ASPs with ~70% gross margins. Two confirmed Western OEM wins at the snapshot date: VW Group mainstream brands (Q1 2025), unnamed second Western OEM (Q3 2025). (Subsequent event: a third Mahindra agreement for 6+ models from 2027 was announced February 2026.)
• SuperVision: hands-off / eyes-on Level 2+ with front+rear cameras and radar on 2x EyeQ6 High, ASP ~$1,300, currently in Polestar 4 (China SOP Aug 2023, global rollout 2024). VW Group Audi/Bentley/Lamborghini/Porsche programs pushed from late 2026 to Q1 2027.
• Chauffeur: eyes-off Level 3 conditional autonomy using 3x EyeQ6 High plus Mobileye's proprietary 4D Imaging Radar, ASP $2,500-3,000. A "leading global automaker" selected May 2025 for a 2028 SOP, the named OEM has not been publicly disclosed.
• Drive: Level 4 fully driverless using 4x EyeQ6 High. Powers MOIA's ID.Buzz robotaxi (Hanover pre-series production Q1 2025, ~100 test vehicles in 6 cities, driver-out testing 1H 2026, Lyft running Dallas fleet, Uber using ID.Buzz in LA).

Chip roadmap: multi-generation cadence

EyeQ5 (~24 TOPS, premium tier, ~10% of mix in Q3 2025 peaking ~15% in 2026); EyeQ6 Lite and EyeQ6 High both 7nm and ramped in 2024-2025; EyeQ7 sampling in 2025 on 5nm/3nm node with performance comparable to NVIDIA Thor, targeting "eyes-off"; and EyeQ8 announced October 2025 in design with 3-4x EyeQ7 performance, targeted for "mind-off" systems in 2029-2030.

Customer concentration: structural through Tier 1s

Geographic mix

The REM data flywheel

Every EyeQ-equipped vehicle anonymously uploads ~10KB/km of road data into the Mobileye Roadbook. >7.5B km cumulative had been mapped by early 2021 disclosure; the harvested fleet is multiples larger today given the 200M+ EyeQ shipments since. GM is the largest REM customer (uses for OnStar mapping and Super Cruise expansion), VW Group is integrated, and Q1 2025 brought a new Korean OEM contract for "REM harvesting + cloud-enhanced ADAS", a new monetization vector beyond AV maps. REM is widely cited as Mobileye's deepest moat because the data flywheel only grows with each new EyeQ shipped, and no current competitor (Horizon, Black Sesame, NVIDIA, Qualcomm) has a comparable crowd-sourced HD-map asset.

Autonomy Stack & Competitive Position

Mobileye occupies a defensible-but-narrowing position in a four-front competitive landscape. The premium L2+/L3 race in Western markets, the China mid-tier, the vertically-integrated premium OEM tier, and the Tier-1 turnkey-stack tier are each being attacked by a different competitor, and Mobileye is winning fewer of them than the IPO-era thesis assumed.

Front 1: Premium L2+/L3 in Western markets (NVIDIA Drive Thor)

NVIDIA Drive Thor (announced 2024, production from 2025) is the platform of choice for Mercedes-Benz (CLA on MB.OS in 2026 using NVIDIA DRIVE AV stack), BYD (next-gen EVs on Thor), GAC Hyper (L4 starting 2025), XPeng, Li Auto, Volvo, JLR, Lucid, NIO, Rivian, and Zoox⁴. NVIDIA's automotive segment grew to $2.3B in FY26 (+39% YoY) with a >$5B run-rate target, multiples larger than Mobileye SuperVision/Chauffeur. NVIDIA is the clearly preferred chip for the new tier of full-stack-developing OEMs.

Front 2: Turnkey Tier-1 stack (Qualcomm Snapdragon Ride Pilot)

Qualcomm's Snapdragon Ride Pilot, co-developed with BMW over three years, debuting in the iX3 (Neue Klasse, 2026 MY) and validated in 60 countries (target 100+ by 2026), has been deliberately opened to ALL global OEMs and Tier-1s⁵. This is a direct attack on Mobileye's mid-tier and premium addressable market with the same "turnkey stack" value proposition Mobileye historically owned. Qualcomm Q2 FY26 auto revenue $1.33B (+38% YoY) confirms execution velocity. The risk isn't BMW specifically, it's that Mobileye's structural mid-tier moat (offering OEMs a complete drive-out-the-door solution) now has a credible Tier 1 alternative.

Front 3: China mid-tier (Horizon Robotics + Black Sesame + BYD in-house)

Horizon Robotics holds 47.7% of China domestic-brand basic ADAS share and 44.2% of NOA under RMB 200K (FY25 revenue +57.7% YoY to RMB 3.758B; Journey series shipments 4.01M units +38.8%)⁶. Journey 6 launched 2025 at up to 560 TOPS. The CARIZON JV (CARIAD + Horizon) is materially negative for Mobileye's China SuperVision exposure: first ADAS production 2025, VW vehicles with CARIZON ADAS in market 2026, with Phase 2 involving VW developing its own SoC with Horizon, signaling VW's China ADAS roadmap is being built around Horizon, not Mobileye.

BYD's vertical integration ("God's Eye" platform with >2.3M vehicles equipped; Xuanji A3 4nm in-house chip in development with combined >2,100 TOPS targeted) and Black Sesame Technologies (1H25 revenue +40%; Huashan A2000 in US export-clearance pipeline) confirm the in-house-silicon and domestic-chip trend has reached the largest volume OEMs. Geely's Haohan Intelligent Driving 2.0 already ejected SuperVision from Zeekr 001 (the original SuperVision launch customer), the canonical case study of how this plays out.

Front 4: Vertically-integrated premium (Tesla AI5)

Tesla is structurally NOT a Mobileye customer. Tesla has telegraphed AI5 tape-out for early 2026 on TSMC 2nm/3nm (also planning Samsung Taylor, TX as a second source). AI5 is expected to deliver ~5x the useful compute of dual-SoC HW4 with 144GB unified memory (9x HW4's 16GB), with mass production targeted for late 2026 and volume in mid-2027. Tesla is now collapsing inference (AI6) and training (Dojo 3) onto a converged chip roadmap, a deliberate strategy that closes off Tesla as an addressable customer permanently. This represents zero TAM impact to Mobileye, but is a strategic threat in that Tesla FSD performance increasingly defines the premium ADAS bar.

Where Mobileye is still winning

• High-volume base ADAS. ~230M EyeQ SoCs deployed lifetime, 1,200+ vehicle models, 50+ OEMs. The legacy franchise is structural and durable; EyeQ6 Lite at ~46M lifetime program units extends it through the 2027 horizon.
• Surround ADAS mid-tier ($100-150 ASPs). Two confirmed Western OEM wins at snapshot (VW Group mainstream, unnamed second Western OEM); a Mahindra agreement was added in February 2026. ~70% gross margin, scalable volume.
• VW Group anchor relationship. March 2024 expanded partnership across Audi, Bentley, Lamborghini, Porsche for SuperVision/Chauffeur; Surround ADAS at mainstream brands; MOIA ID.Buzz L4 robotaxi partnership.
• Polestar 4 SuperVision in production. China SOP Aug 2023, global rollout 2024, proves SuperVision works in commercial deployment.
• Imaging Radar for L3 (2028). Unnamed "leading global automaker" selected May 2025, the highest-ASP product tier ($2,500-3,000) reaches commercial production with a major OEM. Mobileye's terminated FMCW lidar program freed R&D for Imaging Radar focus.
• REM data moat. >7.5B km mapped (2021 disclosure; multiples larger now). GM, VW, new Korean OEM in 2025. No competitor has a comparable crowd-sourced HD-map asset.

Competitive matrix: at a glance

The synthesis: Mobileye still owns the high-volume base ADAS franchise and is winning at the Surround-ADAS mid-tier, but has lost the premium L2+/L3 race in the West (to NVIDIA and Qualcomm) and is structurally losing share in China (to Horizon, Black Sesame, BYD in-house). SuperVision unit guidance of only ~150K in 2026, essentially flat YoY with Porsche/Audi pushed to Q1 2027, confirms the aspirational 1.2M-unit SuperVision figure lacks a near-term customer-by-customer bridge. The Chauffeur (L3 eyes-off) and Drive (L4 robotaxi w/ VW MOIA) optionality is real but unproven.

Intel Overhang & Share Structure

The Intel ownership structure is the single most important non-operational variable in the MBLY thesis. Mobileye operates with a dual-class share structure: Class A carries 1 vote, Class B carries 10 votes per share, with all Class B shares held exclusively by Intel (via Intel Overseas Funding Corporation, a wholly owned subsidiary) and convertible 1:1 into Class A at any time⁸. This structure ensures Intel maintains near-total voting control even after meaningful economic dilution. Note: this section reflects the December 11, 2025 snapshot using the Q3 2025 10-Q and the July 2025 secondary offering as primary disclosure sources. Where subsequent filings (FY2025 10-K, 2026 DEF 14A proxy) further refine the figures, the differences are non-material and called out parenthetically.

Current ownership snapshot

Figures below reflect the post-July-2025-secondary state per the Q3 2025 10-Q and offering disclosures (the most recent disclosures available at the Dec 11, 2025 snapshot date). The subsequently filed FY2025 10-K and 2026 proxy confirmed the directional trajectory.

Post-IPO trajectory of Intel's stake

• October 2022 IPO: Intel held ~94% economic at the $21 debut price. Initial implied market cap ~$16.7B.
• June 2023 secondary: Intel sold 38.5M Class A shares at $42.00 (~$1.5B gross proceeds), reducing economic ownership to ~88.3%. This was the high-water-mark of the post-IPO valuation.
• July 2025 secondary (the watershed transaction under Lip-Bu Tan): Intel sold 50M Class A shares at $16.50: roughly one-quarter the 2023 price, reflecting MBLY's stock decline. Concurrent with the offering, Intel converted 100M Class B shares to Class A, and Mobileye repurchased 6,231,985 Class A from Intel at $16.05 (~$100M).
• End-FY25: Economic ~79.5%, voting ~97.3%. The Class B-to-A conversion mechanic permanently shrinks the high-vote pool, accelerating the eventual path to loss of controlled-company status if Intel continues selling.

The Lip-Bu Tan catalyst: "use the cash"

The catalyst for accelerated sell-down is unambiguous. Pat Gelsinger was ousted as Intel CEO on December 1, 2024 (severance ~$7M+); Lip-Bu Tan became CEO in March 2025 with an explicit mandate to divest non-core assets. Tan completed the Altera sale (51% to Silver Lake for $4.46B cash, implying ~$8.75B Altera enterprise value) as the first move; CFO David Zinsner publicly stated Intel could "use the cash" from a Mobileye sell-down. The July 2025 secondary was the direct expression of this strategy.

Intel's residual contractual rights

Even at reduced economic ownership, Intel retains durable contractual control rights:

• Most-Favored-Nation pricing. As long as Intel holds 20%+, Mobileye must offer Intel pricing equivalent to its best single customer. This is a structural ASP drag on the relationship.
• Section 355 spin-off cooperation rights. Mobileye must cooperate with any tax-free spin Intel pursues. Preserves the spin option for Intel.
• Controlled-company governance. As long as Intel holds majority voting power, Mobileye remains a controlled company under Nasdaq rules, exempt from independent-majority board requirements.
• Anti-dilution option terminated. Intel's previous right to maintain ≥80.1% ownership through future Mobileye issuances terminated after the July 2025 offering because Intel fell below 80%. Meaningful milestone, doesn't change voting control.

Board control: three Intel executives + Intel-affiliated chair

Intel can unilaterally elect the entire board. (Subsequent event after our Dec 11, 2025 snapshot: the February 2026 Mentee Robotics acquisition, $900M; ~$612M cash + 26.2M Class A shares, became a vivid example: a related-party transaction (CEO Shashua and CTO Shalev-Shwartz are co-founders/shareholders of Mentee), approved by Intel as sole Class B holder, illustrating how controlled-company status enables major capital decisions without minority shareholder approval.)

Lockup mechanics: the current overhang

The July 2025 offering carried a 90-day lockup with Goldman Sachs and BofA Securities as gatekeepers (no sales without their written consent through ~early October 2025). That lockup has now lapsed, meaning Intel is once again unconstrained to sell. Every quarter through 2026-2027 carries secondary-announcement risk; historically, secondary announcements have produced -5-10% reactions on print.

What unwinds the overhang: the path to multiple expansion

For MBLY's multiple to expand to AMBA-7x or Horizon-high-teens parity, one of the following needs to happen:

• Accelerated run-off bringing Intel voting power below 50%. At ~97% voting today with 597.8M Class B outstanding, this requires multiple secondaries and substantial Class B-to-A conversions. Multiple years and many transactions away absent a strategic event.
• Section 355 spin-off announcement. Intel could distribute Mobileye to Intel shareholders tax-free, removes the overhang in one stroke. Cooperation right preserves this option; no formal announcement.
• Strategic sale of Intel's stake. A strategic buyer (Apple, Qualcomm, Samsung, an OEM consortium) acquiring Intel's full stake would remove the overhang but introduce a different governance question. No current rumors.
• Multi-year monetization framework disclosure. If Intel announced a pre-committed multi-quarter sell-down with defined tranches and pricing methodology, the market could price the overhang once rather than discounting it every quarter. Zinsner's "use the cash" framing suggests run-off without a formal framework, currently.

Market implication

Every quarter carries secondary-announcement risk (historically −5-10% on print), MFN pricing rights compress ASPs while Intel stays >20%, and three sitting Intel executives plus the chair direct capital allocation. This overhang structurally prevents any operational re-rate from translating into multiple expansion to the AMBA 7-8x or Horizon Robotics high-teens levels; the stock can earn into a higher absolute price but cannot close the discount to peer ADAS franchises until Intel ownership clarity emerges. The Hold rating is the analytical expression of this fact.

Financial Health & Trends

Q2 2026 print highlights (reported Jul 23, 2026)

The Q2 2026 print showed the recovery stalling: revenue was roughly flat at $508M (+0.4%) on 10.0M EyeQ units (+3%), and management raised the FY26 revenue midpoint to ~$2.0B, but the raise and the $0.19 adjusted-EPS beat were flattered by a ~$93M one-time Israeli R&D tax credit; underlying earnings were weaker and the Q3 outlook is soft (revenue −5% to −6% YoY). No new impairment was taken this quarter; the Q1 $3.788B non-cash goodwill impairment stands but does not touch cash or the operating trajectory. Founder-CEO Amnon Shashua also announced his departure with no successor named. Negatives shown in (parentheses) per house convention.

FY25 guide evolution: three raises

Margin trajectory: GAAP vs non-GAAP divergence

Three-year P&L narrative

• FY22 $1,869M: Pre-IPO peak operational year; Intel-era reporting baseline.
• FY23 $2,079M (+11%): Cyclical peak with ~37.4M EyeQ unit shipments. Stock peaked $47 area on this trajectory.
• FY24 $1,654M (-20%): Severe Tier-1 inventory correction. ZF, Valeo, Aptiv sitting on ~6-7M units of excess EyeQ inventory built during 2023 chip shortage. By Q2'24 the excess was "almost fully behind," but H2'24 China softening (Zeekr in-sourcing + weaker EV demand) removed ~3.5M units. GAAP net loss $(3.09B) driven by the $2,695M goodwill impairment.
• FY25 $1,894M (+15%): Recovery year. EyeQ +22% units, three guidance raises, OCF +150%.
• FY26E ~$2.0B (+5%): Guidance raised to ~$1.995B midpoint at Q2 2026; after Q1 revenue grew +27% YoY, Q2 was roughly flat (+0.4%) at $508M, and a ~$93M one-time Israeli R&D tax credit flattered adjusted EPS while the Q3 outlook softened (−5% to −6% YoY).

Forward consensus dispersion is meaningful

FY2026 32-analyst consensus revenue ~$1.68B (range $1.52-$1.85B), Jefferies modeling ~$1.98B (+2.8%), Mobileye's own internal targeting reportedly trending toward ~$1.995B midpoint. FY2027 27-analyst consensus revenue ~$2.07B (range $1.67-$2.62B), Jefferies at $2.24B (+15.2%). The wide dispersion reflects uncertainty about SuperVision ramp timing and China share trajectory, the bull/bear gap is real, not a calibration issue.

Capital Allocation & Runway

Mobileye carries minimal funded debt and ~$1.5B net cash at Q2 2026 (after the Mentee Robotics acquisition, which was largely stock-funded, and the start of a new $250M buyback). Balance sheet liabilities are primarily accounts payable, accrued expenses, and employee benefits, providing multi-year runway despite GAAP losses³.

Cash-generation divergence from GAAP optics

The gap between GAAP losses (driven by ~$1B+ annualized SBC + intangible amortization from the 2017 Intel acquisition) and FCF reflects the non-cash nature of the largest expense lines. 9M 2025 operating cash flow of $489M (+~150% YoY) and capex of $52M imply ~$437M free cash flow YTD. The structural setup, zero debt, $1.7B cash, $437M YTD FCF, ~55% R&D intensity, frames Mobileye as a self-funded R&D platform whose investment thesis hinges on volume recovery plus SuperVision/Chauffeur ASP scale-up rather than near-term GAAP earnings. Solvency risk is not material.

R&D intensity at ~55%: extraordinarily high

R&D was $861M for the nine months ended September 27, 2025, roughly $287M/quarter, running at ~55-60% of revenue on a GAAP basis. This is extraordinarily high for a semiconductor company but appropriate for an IP/software-led ADAS platform investing through the cycle in SuperVision, Chauffeur, Drive, EyeQ7 (5nm/3nm sampling 2025), and EyeQ8 (3-4x EyeQ7 performance, design for 2029-30 SOP).

M&A optionality and subsequent events

At the Dec 11, 2025 snapshot, Mobileye had not announced a material organic acquisition. (Subsequent event: in February 2026, Mobileye announced the acquisition of Mentee Robotics for ~$900M; ~$612M cash + 26.2M Class A shares, a related-party transaction with founders Shashua and Shalev-Shwartz, approved by Intel as sole Class B holder. The deal is structurally interesting, humanoid-robotics IP could extend Mobileye's autonomy stack beyond automotive, and is a vivid illustration of how controlled-company status enables major capital decisions without minority shareholder approval. The capital deployment reduces post-deal net cash by ~$612M, materially changing the runway math.)

Capital allocation priorities

• Organic R&D investment: dominant priority. ~55% of revenue. The single most important strategic spend; underinvesting risks losing future program wins.
• Selective M&A. The $100M July 2025 buyback was structurally a transaction with Intel during the secondary, not an open-market program. Material organic M&A had not occurred at the Dec 11, 2025 snapshot date.
• No dividend. No formal dividend program; not appropriate given growth-stage GAAP losses.
• No open-market buyback program at scale. Repurchases have been transaction-related (Intel secondary windows), not capital-return signaling.

Valuation Overview

Why EV/Sales over P/E for MBLY

P/E is not the right lens because non-GAAP EPS is modest (~$0.50 FY26 underlying estimate, ~17.5x forward P/E) and GAAP EPS is structurally negative through at least FY27E due to SBC + intangible amortization (and the FY26 goodwill impairment). EV/Sales is the right lens because (a) revenue and unit volumes are clean, measurable metrics; (b) net cash position is meaningful relative to enterprise value; (c) the peer set (ADAS/autonomy silicon) is uniformly valued on EV/Sales by sell-side and buy-side; (d) the FY26-FY27 mix-shift story (Surround ADAS, SuperVision, Chauffeur ASPs) is best modeled through revenue × multiple.

Where the multiple "deserves" to sit

A franchise growing mid-teens with 68-72% target gross margins, ~15% adj. op-margins scaling toward 25-30% on Surround ADAS / SuperVision / Chauffeur mix, plus credible 2027+ Chauffeur and Drive optionality, would justify 5-7x NTM EV/Sales. But the Intel overhang structurally caps this, the comparable in-house-stack franchises (AMBA, Horizon Robotics) don't carry a controlling shareholder with explicit monetization mandate. A ~3.6x multiple is the right anchor for the base case, only just above today's ~3.0x spot, with ~4.6x reserved for the bull where the overhang is partially neutralized by a Chauffeur production-OEM catalyst; both sit well below AMBA's ~7.5x, and the size of that discount is the Intel overhang compounded by the founder-CEO's announced departure and the soft Q3 guide.

The MBLY vs AMBA multiple gap: anatomy of the discount

AMBA trades at 7.5x; MBLY at 3.0x. The ~4-point gap decomposes roughly as: (a) Intel overhang and float-release risk: ~2 points; (b) China share-loss trajectory vs AMBA's more diversified end-markets: ~1 point; (c) higher gross margins and larger installed base at MBLY (-0.5 points offset to MBLY's favor); (d) AMBA's earlier 2024-2025 recovery cycle vs MBLY mid-recovery: ~1 point timing premium. Each component is independently observable; a Chauffeur announcement could compress the (c) overhang component by ~1 point; an Intel monetization framework could compress it by 1.5-2 points.

SuperVision ramp scenarios: the swing factor

The aspirational 1.2M SuperVision unit figure (cited by management) implies FY28E SuperVision revenue of ~$1.5-1.6B at $1,300 ASPs, alone roughly equal to 70% of FY24 total revenue. This is the upside case; current 2026 guide is just ~150K units. A credible customer-by-customer bridge requires Porsche, Audi, Bentley, Lambo all delivering on the Q1 2027 SOP (currently delayed once already from late 2026); China stabilization preventing further defections; and at least one additional Western OEM beyond VW Group. The bear case ($1.85B FY27 revenue) effectively assumes the SuperVision bridge breaks.

EV/Sales Scenario Model

FY27E revenue × EV/Sales multiple = enterprise value, plus net cash, divided by diluted shares = forward fair value, discounted one year at 10% (ADAS-cycle visibility + Intel-overhang risk premium) to today's PT.

Sensitivity grid: FY27E revenue × EV/Sales multiple (PT in $)

Quick PT calculator

Risk / Reward calculator

Upcoming Catalysts

Risk Factors (Including Upside Cases)

For a Hold rating, the risk taxonomy must be symmetric. We list both bear-case risks (the thesis is wrong, downside materializes) and upside-case risks (the thesis is wrong, upside materializes). For MBLY both directions are credible, that's what defines a Hold.

Bear-case risks (downside drivers)

• CEO transition / key-person risk. Founder-CEO Amnon Shashua announced his departure in Q2 2026 with no successor named. Shashua is the technical and strategic architect of Mobileye's autonomy roadmap (EyeQ, SuperVision, Chauffeur, REM); an unmanaged transition is a governance and key-person risk layered directly on top of the Intel control overhang.
• Earnings quality: one-time-credit-flattered EPS. Q2 2026 adjusted EPS of $0.19 beat, but was boosted by a ~$93M one-time Israeli R&D tax credit, and the FY26 guidance raise mostly reflects the same credit rather than an operational step-up. Underlying earnings power is weaker than the headline, and the Q3 outlook is soft (revenue −5% to −6% YoY), so the "beat and raise" should not be read as a re-rating catalyst.
• Intel secondary offering announcement. Post-July-2025 lockup has lapsed; Lip-Bu Tan's "use the cash" mandate creates persistent overhang. Historical announcement reaction −5-10%. Intel holds 597.8M Class B shares + 50M Class A as of April 2026 proxy.
• China ADAS share continued loss. Horizon Robotics 47.7% domestic basic ADAS share, 44.2% sub-RMB200K NOA, FY25 revenue trajectory +57% YoY (per Horizon's interim disclosures). CARIZON JV building VW China roadmap around Horizon. BYD Xuanji A3 in development. Black Sesame Huashan A2000 in US export-clearance pipeline, first potential export risk to Western markets.
• Porsche/Audi SuperVision slip past Q1 2027. Already pushed once from late 2026 to Q1 2027; further delay would crater the 2027 SuperVision unit bridge (the aspirational 1.2M figure lacks customer-by-customer credibility).
• Tier 1 customer concentration with ZF / Valeo / Aptiv (~71% of revenue FY2023). Any one of these losing share to in-house OEM modules or Qualcomm-based reference designs would directly hit Mobileye volumes.
• Qualcomm Snapdragon Ride Pilot mid-tier penetration. Opening Ride Pilot to all OEMs and Tier-1s with 60-country (target 100+) validation directly attacks Mobileye's Surround ADAS positioning at OEMs evaluating turnkey alternatives.

Upside-case risks (Hold could be wrong because…)

• Chauffeur L3 production-OEM announcement. Mobileye selected by "leading global automaker" for 2028 SOP with Imaging Radar (announced May 2025); formal naming unlocks $2,500-3,000 ASPs and would catalyze multiple re-rate toward AMBA-7x. Materially positive.
• Mobileye Drive / MOIA L4 commercial milestones. Pre-series production at Hanover in Q1 2025, ~100 ID.Buzz test vehicles in 6 cities, driver-out testing 1H 2026, Lyft (Dallas) and Uber (LA) as fleet operators, first commercial revenue would validate Drive optionality not in current Street numbers.
• SuperVision China stabilization. Polestar 4 (China SOP Aug 2023, global rollout) and VW Group China (Audi/Bentley/Lambo/Porsche by end-2026/2027) collectively could offset Zeekr defection if execution holds.
• Intel monetization framework or Section 355 spin announcement. If Intel announces a multi-quarter pre-committed sell-down framework or a tax-free spin, the overhang resolves once rather than quarterly. Either would catalyze a one-step multiple re-rate.
• EyeQ7 sampling translates to design wins. EyeQ7 (5nm/3nm, comparable to NVIDIA Thor in compute) sampling in 2025. Any Tier 1 or OEM design win on EyeQ7 in 2026 would refresh the eyes-off Chauffeur tier and could re-establish the premium-tier narrative.
• Emerging-OEM Surround ADAS scale. Mahindra (subsequently announced Feb 2026, 6+ models from 2027) + REM monetization at Korean OEMs + base-ADAS regulatory volume (Euro NCAP, UN regulations) could expand the volume base independent of premium-tier outcomes.

Bull vs Bear Debate

Technical Analysis

Trader's view

• Established a $6.47-$15.81 range over the trailing year; now basing in the high-single-digits (~$8.73) after the impairment-driven sell-off and the soft Q2 print.
• Key support: $8 (recent basing zone). A close below opens the path toward the $6.47 52-week low; overhead resistance sits at the $10 PT zone.
• Key resistance: $17 (analyst median PT cluster + structural distribution shelf where Intel sold July 2025 at $16.50 + Mobileye repurchased at $16.05). A weekly close above $17 with volume would signal that the next secondary risk is being absorbed; unlocks $20 zone where Q2 2025 highs sat.
• R/R from ~$8.73 to the $13 bull target with a $6 stop = ~1.6:1, only a modest edge for an event-driven long (Chauffeur announcement, China stabilization print). The $10 base offers ~+15% with two-sided risk; the CEO transition, the soft Q3 guide, and Intel secondary risk argue for patience and tighter stops.
• Volume profile shows a heavy distribution shelf at $16-18, this is now structural resistance because the next Intel tranche would likely price in that zone if executed.
• Options market implied moves around earnings have averaged 8-12%, consistent with the binary nature of guidance prints and secondary risk.