RIVN (Trans): R2 to drive vols; GPM seen turning positive by YE

I. Earnings Recap: Key Takeaways

1. Liquidity and External Capital

a. Cash, cash equivalents and short-term investments were approx. $5.5bn at end-Q3. The $1.0bn, 10-year, 6% fixed-rate term loan with Volkswagen Group funded on Oct 7. b. Total expected JV capital commitments are $5.8bn, with $5.3bn already funded and the remaining $460mn expected by end-2027. The second milestone under the Uber collaboration is expected this year, triggering an additional $250mn equity investment. c. Including conditional facilities from Uber, Volkswagen and the DOE loan, available liquidity plus targeted future capital totals approx. $13.0bn.

2. Outlook

a. Deliveries: Full-year guide raised to 67k–70k units from 65k–70k. R2 will account for the majority of Q4 deliveries. b. Profit: Full-year Adj. EBITDA loss guided at $1.75bn–$1.85bn, with the midpoint improving by $100mn, driven by regulatory credits, IEEPA tariff refund receivables, and higher deliveries. c. Vehicle GP: Expect to exit the year with positive vehicle GP (not positive for the full Q4). Further expansion in 2027 as R2 ramps. d. Opex: Q4 R&D to grow low double digits QoQ. In 2027, R&D growth will outpace 2026 materially, focused on ADAS/AV hardware and software and core vehicle tech.

3. CapEx

a. 2026 CapEx midpoint lowered by $300mn to $1.4bn–$1.5bn, driven purely by timing shifts in large equipment and contractual payments at the Georgia plant. b. CapEx will rise significantly YoY in 2027 as Georgia construction accelerates.

4. Key Financials This Quarter

a. Total: Revenue of $2.155bn (+38% YoY). GP of $176mn with GPM of 8% (includes $166mn D&A and $28mn SBC). Adj. EBITDA loss of $435mn as the company scaled sales, marketing and service for R2 and stepped up autonomy investment, lifting opex. b. Automotive revenue: $1.55bn (+36% YoY) on deliveries up 46% YoY, with $81mn regulatory credits. Mix shift toward R2 and commercial vans weighed on blended ASP, while R1 ASP rose on mix. c. Vehicle GP: Loss of $80mn, narrowing by $50mn vs. a $130mn loss a year ago, driven by higher deliveries and per-unit GP improvement, increased regulatory credits, and IEEPA tariff refund receivables. Partly offset by R2 ramp costs. d. Software & Services: Revenue of $605mn (+45% YoY), led by E/E architecture and software development for the JV, maintenance and repair services, and Autonomy+, partly offset by lower used car remarketing sales. The Volkswagen JV contributed $354mn (~60%). GP was $256mn with GPM of 42% on a richer-than-expected high-margin mix. e. Production and one-offs: Built 19,751 units and delivered 19,248. R2 ramp added approx. $120mn of incremental operating costs vs. normalized run-rate.

5. Finance Leadership: CFO Claire McDonough stepped down. VP of Finance Derek Mulvey became interim CFO effective the day after the call.

II. Earnings Call Details

2.1 Management Commentary: Key Messages

1. R2 Demand and Sales

a. Launched in Jun, R2 surpassed R1 and EDV in Q3 to become the top-selling model, driving record deliveries. b. Q3 test drives for R1 and R2 exceeded 103k, topping the Q2 record. Launch Edition reservation conversion continues to beat expectations, with more trims rolling out over coming months. c. R2 won Top Gear's 2027 SUV of the Year.

2. Manufacturing and Supply Chain

a. Record output at the Normal plant. R2 second shift has started, with utilization to rise over the next few quarters. b. Construction of the new Georgia plant is on track. c. Macro and geopolitical factors add complexity, cost and uncertainty. The company is addressing supply chain risks and higher commodity and memory costs.

3. Commercial (EDV)

a. Amazon continues to expand its North America electric delivery fleet. b. Deliveries of a new rural-route variant began in Oct, adding AWD and a larger pack with ~30% more range.

4. Autonomy

a. Autonomy+ is available across all R2 and R1 MY25+, with take rate above expectations. Equipped vehicles can drive hands-free across millions of roadway miles. b. Point to Point: After inputting an address, the vehicle can complete the full trip under driver supervision, handling city streets, traffic lights and highway interchanges. Initial limited release via invite-only Autonomy Early Access Program this year, with broader availability to all Autonomy+ subscribers in 2027. Management views this as a key purchase driver and share catalyst. c. Uber: Collaboration is on plan, with the next milestone expected by year-end to unlock a $250mn additional equity investment. Robotaxi is targeted for 2028. d. Announced the acquisition of Third Dimension AI last week to support AV development and closed-loop simulation. e. R&D will accelerate over the next few years, prioritizing the Large Driving Model and higher-level autonomy for consumer models and Robotaxi. An Autonomy and AI Day will be held on Dec 9.

5. Software and Brand

a. Oct OTA added Gmail integration to Rivian Assistant. b. Partnered with Sony Pictures and Bill Murray on a Ghostbusters-themed Halloween mode, showcasing rapid feature releases enabled by deep hardware-software integration.

2.2 Q&A

Q: How much of the autonomy platform can be reused on commercial vehicles?

A: The autonomy platform is built around a Large Driving Model, a large-parameter base model in training that adapts across vehicle types and use cases, covering R1, R2 and commercial vans. Applications span consumer (with Point to Point as the next major milestone), Robotaxi with Uber (launching 2028), and multiple commercial scenarios in the Vans program.

Q: Does 'positive vehicle GP at year-end' mean Q4 as a whole is positive, or only the Dec run-rate?

A: It means exiting the quarter with positive vehicle GP, not positive for the full Q4. This sets a solid base for continued R2 ramp and full-year GP improvement in 2027.

Q: What exactly is the $120mn ramp cost? Excluding ramp inefficiencies and under-absorbed fixed costs at current pricing and mix, what is R2's marginal contribution?

A: The $120mn is the delta vs. the two-shift full-run-rate conversion cost and depreciation, essentially reflecting under-absorption before full two-shift utilization. It also includes expedite freight and other scale-inefficient items that fade with volume. Per delivered unit COGS fell by approx. $12k vs. Q2, mainly from R2 introduction, which both carries a lower cost base and brings scale to the Normal plant. Ex-regulatory credits and on a cash basis, vehicle GP turned positive this quarter and should continue improving as R2 ramps.

Q: Since Mar pricing, memory, PCB, electronics, steel and copper costs rose (aluminum roughly flat). How will engineering and manufacturing efficiency offset this to preserve target margins on new trims?

A: While ramping and unlocking Normal capacity, engineering is also refining designs. Some changes will apply to models built at the Georgia Stanton Springs plant, and some will flow into current models over coming months and quarters. Commodities are notably above plan, with memory particularly acute, and we are pursuing offsets to hold delivered pricing across trims at $45k entry and roughly $57k top-end. Cost actions span factory operations detail, logistics and packaging from suppliers to plant, and part-level cost optimizations with suppliers that do not impact performance, often via redesign to consolidate or eliminate parts.

Q: For R2 delivery growth into Q4 and 2027, what is the key swing factor: converting record test drives and buzz into orders, or supply chain and manufacturing ramp?

A: Supplier ramp is as critical as the plant itself. R2 added many new suppliers and we intentionally streamlined Tier-1 count; the overall pace is gated by the slowest supplier. There is no single severe bottleneck, but synchronized ramp across suppliers is required, with supply chain and logistics teams closely coordinating. The COO added that scaling hundreds of suppliers and their tier-2/3s is a major program, and we deploy dedicated teams to help expand capacity where needed. Expanding the supply base is a top focus.

Q: How much did Autonomy+ contribute to the 45% YoY growth in Software & Services? Has it become a key purchase driver?

A: (Not quantified.) Autonomy+ take rate is above expectations. Today, Universal Hands-Free supports hands-free driving across billions of miles of marked roads in the U.S. The bigger step is Point to Point: hands-free plus full-route navigation to destination, traversing four-way stops, complex interchanges and on/off ramps. We believe this will unlock incremental demand, with Tesla FSD being the only similar offering today. As more customers experience it, this capability will become a core purchase criterion over the next few quarters and years.

Q: EDV revenue with Amazon hit another quarterly high. Beyond Amazon (e.g., vans seen with Walmart livery), what is the adoption potential with other customers?

A: EDV has been optimized for structural cost advantages. Amazon has deployed at scale from early in the program through this year, gathering extensive operational data that validates true structural cost benefits, hence continued orders. Demand from Amazon stepped up meaningfully this year, and other commercial operators evaluating efficiency gains are now assessing EDV, which we view positively.

Q: Progress on the E/E architecture JV with Volkswagen? Are other OEMs interested in adopting the tech?

A: The first mass-production outcome will be Volkswagen's ID. EVERY1 launching next year in Europe, an EV priced around €20k, showcasing the tech outside the Rivian brand and the U.S. It will compete directly with European and Chinese OEMs, and we are confident in cost-efficiency at that feature set and price, treating it as a 'billboard' for the technology. We see potential beyond the JV over time, but our focus remains Rivian programs and Volkswagen Group workstreams.

Q: Are consumers already buying in anticipation of future Point to Point? Will Point to Point and Gen 3 hardware drive a step-change in demand?

A: To date, autonomy has not been the primary purchase driver for most U.S. consumers, with most vehicles sold featuring limited autonomy. But the industry is at the start of a major inflection, with more consumers seeking higher-level capability. Point to Point is the unlock: enter a destination and the cognitive load of driving and navigation falls substantially. We are prioritizing resources here, from sensor suite and compute architecture to neural base-model development, as a first step toward L4 targeted for 2028.

Q: Does acquiring Third Dimension help achieve L4 and expand ODD?

A: In model-based AI approaches versus rules-based 'Autonomy 1.0', evaluation is critical. Open-loop, via real-world road testing, is widely used by us and others and will continue. We also aim for closed-loop, testing models under real and simulated conditions with heavy iteration and injected faults and edge cases. Third Dimension strengthens our existing simulation, enabling faster closed-loop cycles and quicker model deployment and iteration.

Q: What will the Dec Autonomy Early Access Program validate ahead of the 2027 broad release? Is it to gather data and verify safety outcomes?

A: Early access is a standard path for new tech, used to collect broad user feedback on functions, interactions and UI across diverse customers, driving environments and geographies. It prepares us for the 2027 full push to all Autonomy+ subscribers.

Q: What is the initial Point to Point scope? How will long-tail scenarios like pick-up/drop-off be handled, and how will it evolve?

A: Pick-up/drop-off is particularly important for Robotaxi/L4, especially in dense urban areas with no curb space and lane-only stopping, which remains challenging in current deployments. Consumer Point to Point is supervised, so complex pick-up/drop-off can be handled by the driver. The focus is UI: what the vehicle does on arrival and how it prompts the driver, and we are working to make that interaction seamless and intuitive.

Q: R&D (ex-SBC) rose 7% YoY. How are pricing and availability for rented AI training compute?

A: Compute rental prices have increased. We expect Q4 R&D to grow low double digits QoQ, and 2027 R&D growth to be well above 2026, driven entirely by strategic investment in autonomy software, hardware and select vehicle tech. We are efficient with compute, but will expand usage next year as the Large Driving Model progresses.

Q: With higher steel, logistics/freight and DRAM costs, what are the mitigation levers?

A: Costs break down into materials, factory conversion and logistics, and each has dedicated teams working savings from quick wins to multi-year redesigns or supplier changes. This is mandatory even without headwinds, and rising raw materials and memory costs make it more urgent. The most impactful near-term lever is ramp: lifting output from existing lines and teams to improve fixed-cost absorption, pushing supplier ramp, and efficiently converting parts into finished vehicles.

Q: Why lower 2026 CapEx to $1.4bn–$1.5bn?

A: Timing. Georgia plant spend includes large equipment and big contracts with lumpy cadence, and invoice/payment shifts by weeks can move between quarters. The revision is purely timing-related.

Q: Media showed Rivian vehicles with Walmart branding. Is this testing or a corporate partnership?

A: The commercial van program continues discussions with customers beyond Amazon. The Amazon partnership is progressing very well with a sharp volume increase this year, demonstrating performance and cost advantages of the platform. We will continue to pursue additional partners. (No direct comment on Walmart.)

Q: Is Third Dimension essentially converting regular video into a 3D world replica? Are other capabilities needed to complete the AV stack?

A: The acquisition enhances internal simulation to build efficient closed-loop validation and model improvement. This work was already underway, and the acquisition accelerates it. Simulation-based validation is increasingly important for rapid iteration of model architectures and designs. (No comment on additional capabilities.)

Q: What are the top priorities to raise brand awareness over the next 2–3 years? How will you measure conversion to customer acquisition and revenue?

A: R1, as the flagship, is the brand door-opener, but price limits consideration for many consumers. R2, starting at $45k and topping just over $50k, overlaps with the U.S. new-car Avg. price and sits in the popular midsize crossover/SUV segment, with rising road presence lifting awareness on its own. We also expand brand awareness via creative marketing, such as this year's Ghostbusters collaboration with Sony and Bill Murray, crossing cultures and audiences.

Q: The R2 LiDAR variant was planned for consumers this year but is now employee-only testing/purchase. Was it parts delay or stronger-than-expected launch performance?

A: The Large Driving Model will deploy across all R2 (including Launch Edition) and R1 MY25+, making Point to Point available on those vehicles. The next-gen hardware features the in-house RAP1 processor at 800 TOPS per chip, 1,600 TOPS per vehicle with two chips, and an enhanced perception stack using cameras, mmWave radar and LiDAR. That version enters employee programs and validation builds by year-end, with market launch in 2027.

Q: Test drives surged after R2 launch. How is test-drive-to-order conversion versus R1?

A: R2 launched with the top-priced high-performance trim, and while we had assumptions, actual conversion is above expectations. We are similarly optimistic about upcoming trims. Test drives are a key KPI, and R2 materially lifted volumes. (No specific conversion rate or R1 comparison disclosed.)

Q: How is demand for next year's lower-trim, smaller-pack R2? What is the margin impact?

A: Trims fall along two axes: interior grade (premium/standard, including features like the rear drop glass) and powertrain (AWD Performance, AWD Base, RWD with small/large packs). The portfolio covers the demand curve, offering choices across performance, range, price and features from $45k to $57k. On unit economics, early single-shift production and supplier ramp mean limited absorption of fixed costs like depreciation and plant labor; as volumes grow, those fixed costs are spread over more units, materially lowering vehicle cost. We intentionally lead with higher-priced trims and add others as scale builds, ultimately supporting healthy positive GP on R2, which given its scale will drive overall financials.

Q: Hands-free take rate is above expectations. Is it near Tesla FSD's ~75%?

A: Today, hands-free works on nearly all lane-marked roads, covering millions of miles, but it cannot yet exit ramps into city streets or traverse traffic lights and four-way stops. Point to Point does, and we see it as the autonomy experience inflection; the CEO uses it daily for commuting. We believe it will be a major demand driver, but the pace at which the broader U.S. market treats it as a core purchase criterion is uncertain. That is why we invested heavily for years, from launching the in-house RAP1 chip to vertically integrating the stack, with the perception platform feeding and running the Large Driving Model. We expect Point to Point availability to Autonomy+ users to lift take rates further. No specific rate disclosed: 'maybe later, not now'.

Q: Will pricing rise after Point to Point launches (Tesla FSD is about $99/month)?

A: Over the long run, pricing has room to move up, but specifics will depend on market dynamics. Today it is a standalone subscription, but it may be embedded in vehicle pricing in the future, and models could include monthly, one-time buyout or bundled. Autonomy will reshape industry structure and likely drive major share shifts among companies; pricing will be market-driven and dynamic, not static for a decade.

Q: With R2 sales nearly 2x R1, what is R1's long-term role? Will you leverage R2 scale and supply chain to lower R1 pricing toward other 3-row SUVs to boost volume?

A: We expect R1 ASP to tick up modestly by design. We discontinued the R1 standard trim to move the product upmarket and separate its price band from R2. As the flagship, R1 includes a quad-motor variant over 1,000 hp, a tri-motor over 800 hp, plus a performance dual-motor, with strong on-road and off-road capability; per-vehicle revenue this quarter reflects that shift. R1 is not a volume play and will not compete with lower-priced two- or three-row models; it is positioned as a premium offering.

<End of text>

Risk Disclosure and Disclaimer:Dolphin Research Disclaimer and General Disclosure