Rivian Automotive (RIVN)

Summary

1. Execution is visibly improving. Q3 deliveries were a record 19,248, up 46% year over year, as the R2 ramped in its first full quarter. Management raised 2026 guidance to 65,000–70,000 deliveries in July and reaffirmed it in October.
2. Rivian is now the number 2 US EV brand, but the growth is vans and R2. It took 7.3% of US EV sales in Q3 2026, up from 3.0% a year earlier. Delivery vans and the new R2 explain all of that growth; sales of the premium R1 fell 31%.
3. The gross profit is real, but it is not yet automotive. Q2 gross profit of $179 million came entirely from software & services ($215 million). Automotive lost $36 million even with $108 million of regulatory credits and a tariff refund. Excluding credits alone, that is a loss of about $11,800 per vehicle. Sixty percent of software & services revenue was Volkswagen joint-venture engineering fees, which are milestone-driven.
4. Funding is the binding constraint. My base case burns about $15 billion cumulatively, with stock-based compensation counted as a cost, before free cash flow turns positive in 2032. Management cites over $14 billion of available and targeted capital, much of it conditional. I expect at least one more equity raise; July’s $1.3 billion sale at $15.50 is unlikely to be the last.
5. At $12.95 the market already pays a little more than my base case. A reverse DCF needs 1.3 points more automotive margin than my base path. The upside is a bull case worth nearly 3x, built on licensing, autonomy and the Uber robotaxi program. The downside is a one-plant survival case worth about $5.
6. Neutral, fair value $14. The expected return of about 11% sits just under the 12.6% cost of equity and well inside the 10-point band my rule requires for Outperform or Underperform. I would turn buyer below about $11.70, or on evidence that automotive gross profit turns positive in Q4 2026 without credits.

Where I differ from the market

Consensus sits at roughly $19, with 22–28 analysts split between Hold and Buy and targets from $10 to $25. My $14 is below that for three reasons, each a deliberate choice rather than a difference in optimism.

The business today

Rivian builds the R1T pickup and R1S SUV (roughly $70,000–$100,000), commercial delivery vans for Amazon and others, and the smaller R2 SUV, built since April 2026 and delivered since June (launch trim $57,990, with a $44,990 version due in summer 2027). Everything is built in Normal, Illinois, which has capacity for about 215,000 vehicles a year after the R2 line. A second plant in Georgia, financed by a $4.5 billion Department of Energy loan, is under construction for up to 300,000 units, with production targeted for late 2028.

Software & services has become the second engine. It includes the 50/50 technology joint venture with Volkswagen (up to $5.8 billion in total commitments), vehicle service and repair, used vehicles, and the Autonomy+ driver-assistance package ($2,500 or $49.99 a month). In March 2026 Uber agreed to invest $300 million, and up to $1.25 billion on milestones, toward a fleet of autonomous R2 robotaxis starting in 2028.

Reported results, USD billions (deliveries in thousands)

Sources: Rivian 8-K earnings releases for Q4/FY2025 (Feb 12, 2026), Q1 2026 (Apr 30, 2026) and Q2 2026 (Jul 30, 2026). FY2025 free cash flow is operating cash flow of ($0.779B) less capex of $1.710B.

Where Rivian builds and engineers

Every Rivian sold so far has come from one plant in Normal, Illinois. Software and electronics engineering is spread across California, Canada and Europe, much of it now inside the Volkswagen joint venture. A second plant in Georgia is the company’s volume option, and the main use of its government loan.

1. 1

   Normal, Illinois

   Plant, operating · owned

   • About 7.5 million square feet. Builds the R1T, R1S, Commercial Van and, since April 2026, the R2 (deliveries from June).
   • Capacity of up to 215,000 vehicles a year on multiple shifts. 2025 output was 42,284.
   • A 1.2 million square foot supplier park across the street feeds the line through a tunnel.

   Source: FY2025 10-K, Items 1 and 2

2. 2

   Stanton Springs North, near Social Circle, Georgia

   Plant, under construction · leased land

   • About 1,700 acres. Ground broken September 2025; buildings go up in 2026.
   • First phase raised to 300,000 vehicles a year in April 2026, from 200,000. Production targeted for late 2028.
   • Will build R2 and later R3, plus Uber robotaxis. Financed by a Department of Energy loan of up to $4.5 billion.

   Sources: FY2025 10-K; Rivian, Apr 30, 2026

3. 3

   Irvine, California

   Headquarters

   • Corporate headquarters at 14600 Myford Road: leadership, design and vehicle engineering.
   • Also a site of RV Tech, the Volkswagen joint venture.

   Source: FY2025 10-K cover page; Rivian careers

4. 4

   Palo Alto, California

   Software, autonomy and the joint venture

   • Software and autonomy engineering, including the in-house Rivian Autonomy Processor.
   • Headquarters of RV Tech, the 50/50 Volkswagen venture for electrical architecture and software.

   Sources: FY2025 10-K; Rivian careers

5. 5

   Plymouth, Michigan

   Engineering

   • Vehicle engineering and prototyping, close to the Detroit supplier base. Rivian’s original home.

   Source: Rivian careers

6. 6

   Vancouver, British Columbia

   Software engineering

   • Software and digital engineering. Canada is one of three countries where Rivian leases principal facilities.

   Sources: FY2025 10-K; Rivian careers

7. 7

   Queen Creek, Arizona

   Key supplier · LG Energy Solution

   • Makes the 4695 cylindrical battery cells for R2 under a five-year, 67 GWh agreement. Early cells come from South Korea.
   • US-made cells cut tariff and shipping exposure on the most expensive part of the car.

   Source: LG Energy Solution and Rivian, Nov 2024

8. 8

   Belgrade, Serbia

   Software engineering

   • Software engineering for Rivian and for RV Tech. Rivian’s main European engineering site.

   Sources: FY2025 10-K; Rivian careers

9. 9

   Berlin, Germany

   Joint venture engineering

   • RV Tech engineering site serving Volkswagen brands, as reported in 2025–2026. Volkswagen funds much of the joint venture’s cost.

   Source: press reports; not in the 10-K

Rivian also owns or leases service centers, delivery hubs and Rivian Adventure Network charging sites across North America; the 10-K does not list them individually. Map data: Natural Earth (public domain).

How a Rivian is built in Normal

Rivian designs its own electric motors, gearboxes, inverters, battery packs, vehicle electronics and chassis systems. The packs and drive units are assembled in the Normal plant itself. Cells and most other parts come from outside suppliers, many of them single-source.

Sources: FY2025 10-K (in-house components, supplier park); Rivian, “Building for R2” (March 2025): the R2 expansion added a 1.1 million square foot body shop, general assembly and end-of-line building, a battery pack building and a pretreatment and e-coat paint building.

Why the footprint matters for the stock

• Utilization is the margin story. Normal built 42,284 vehicles in 2025 against capacity of up to 215,000, about one-fifth. The plant’s fixed costs fall on too few cars, which is most of why automotive gross profit is negative. R2 exists to fill the plant.
• One plant means one point of failure. Until Georgia opens, a shutdown in Normal stops all revenue. In 2024 a planned retooling and a shortage of one drive-unit part each cut output.
• Georgia is the bull case’s factory. My base case passes Normal’s capacity in 2030, so every vehicle beyond that depends on Georgia arriving on time. It is also why my capital spending stays at $2.0–2.6 billion a year from 2027 to 2030.
• Volkswagen pays for part of the engineering. The software sites in Palo Alto, Irvine, Belgrade and Berlin increasingly sit inside the joint venture. That shares the cost, but it also means Rivian’s software edge is partly shared with a much larger automaker.

Market share: where Rivian wins, and where it does not

The US electric vehicle market shrank after the $7,500 federal credit ended in September 2025. Third-quarter 2026 EV sales were 262,649, down 39.5% from a year earlier, and EVs were 6.4% of all new vehicles sold, against more than 10% a year before. Rivian grew 46% into that falling market.

Share by segment

Total EV share hides the real picture. Rivian leads two small segments, competes in a third, and is only starting in the one that matters most for the thesis.

What my forecasts require

Turning deliveries into market share is the simplest test of whether a forecast is reasonable. Rivian sells almost entirely in North America, mostly in the US, so I compare its deliveries with the US EV market.

Source: Cox Automotive / Kelley Blue Book, Q3 2026 EV Sales Report, estimates as of October 13, 2026 (PDF, commentary). Shares are computed from Cox’s corrected unit estimates; Cox’s printed share column (Tesla 54.0%, Rivian 7.2%) uses its earlier, uncorrected total. Today’s market size is Q3 2026 sales times four.

Four things that decide the stock

Forecasts and scenarios

The model is driver-based: deliveries times average selling price for automotive, a separate software & services line, operating expenses in dollars, then unlevered free cash flow through 2035. Stock-based compensation stays in expenses. Accumulated losses shelter cash taxes through 2035; a 21% tax applies in the terminal year.

Base case, USD billions

Losing money, or investing?

A common defense of Rivian is that it loses money because it is building plants and new products, not because the business is broken. The 2025 accounts let me test that claim. The answer decides which valuation method fits, so I settled it before choosing one.

Spending on future capacity and products Running today’s business Inflow Result

Three tests

1. 1
   Is the spending building something?

   Yes
2. 2
   Does the car make money once it is built?

   Not yet
3. 3
   Would the losses stop if the investing stopped?

   No

How this shaped the method

• Investment has to be charged when it is spent. A discounted cash flow subtracts capital spending in the year it happens and credits the cash it produces later. Earnings ignore most of it, so an earnings multiple would misjudge an investing company in either direction.
• The payoff has to come from the drivers. My model lets gross margin rise only as deliveries fill the plant, so the investment pays back only if volume arrives. A sales multiple would value the revenue whether or not it ever turns profitable.
• The payoff has to beat the cost of the money. Discounting at 11.5% asks whether roughly $15 billion of further investment earns more than investors could get elsewhere at the same risk. In my base case it only just does, which is why the rating is Neutral.

Sources: Rivian Q4 and FY2025 earnings release (8-K Ex. 99.1, Feb 12, 2026): income statement, cash flow statement, and quarterly stock compensation and depreciation by line, summed by me. “Cash” R&D and SG&A exclude stock compensation and depreciation. Back-office systems, such as the SAP S/4HANA software Rivian uses for finance and procurement (AWS case study), sit inside SG&A. Rivian does not disclose them separately, and they are not a material driver of the loss.

How I valued Rivian, and why this method

Rivian has no earnings: in my base case EBITDA stays negative through 2029 and operating profit through 2030. A price-to-earnings or EV-to-EBITDA multiple would divide by a negative number. What the stock is worth depends on how fast the losses close and how much money it takes to get there. That calls for a cash-flow model built from the operating drivers, run under several scenarios. The previous section explains why: Rivian is partly investing and partly losing money, and only a cash-flow model can tell the two apart.

1. 1Operating drivers

   Deliveries × price, software & services revenue, gross margins and operating costs, year by year.

   Why: the thesis is unit economics. Volume and margin are what each quarterly report updates.

2. 2Free cash flow, 2026–2035

   Operating profit plus depreciation, less capital spending and working capital. Stock compensation stays a cost.

   Why: Rivian’s risk is running out of cash, not accounting profit. Ten years reaches a mature margin.

3. 3Discount rate

   Cost of equity 12.6% from a measured beta of 1.48. WACC of 11.1%, plus 0.4 points for execution risk, gives 11.5%.

   Why: the risk charge comes from how the stock actually trades, not from a guess.

4. 4Terminal value

   Ten times 2035 EBITDA, valued at the end of 2035 and cross-checked against 3% perpetual growth.

   Why: Rivian is still growing in 2035, so a steady-state formula understates it. I publish the cross-check anyway.

5. 5Three weighted scenarios

   Bear 30%, base 50%, bull 20%, each a full model with its own story.

   Why: the outcomes are lopsided, from survival to platform. One point estimate would hide that.

6. 6Rating rule set in advance

   Outperform if the expected return beats the cost of equity by 10 points or more; Underperform if it trails by 10.

   Why: the call follows the numbers, and anyone can check it.

Methods I considered

Valuation

I discount unlevered free cash flow for the second half of 2026 and for 2027–2035 at an 11.5% WACC, using mid-year discounting, and capitalize 2035 EBITDA at 10x as a value at the end of 2035. The cost of equity is 12.6%: a 5.2% risk-free rate plus a beta of 1.48, measured from two years of daily returns against IVV, times a 5.0% equity risk premium. With a 78% equity weight and 6.0% debt cost, WACC computes to 11.1%; I add 0.4 points for execution risk. A 10x exit multiple sits above mature automakers and far below Tesla. As a check, 10x implies 6.7% perpetual growth, which is aggressive, and the same base case on a 3% perpetuity is worth only $4.71. That gap is the honest measure of how much of Rivian’s value depends on long-duration growth.

What the price implies

Holding every other base assumption fixed, the automotive gross margin path that justifies $12.95 is 1.3 points above mine, reaching about 20.3% by 2035. That is above where Tesla has run recently, achieved at less than a tenth of its volume. The market is not ignoring the risks; it is paying for a base case that goes right.

Management, owners and partners

Rivian is still run by its founder, its factory is now run by an automaker veteran, and about a third of its shares sat with strategic partners at the April 2026 record date. The finance chief who built its funding plan leaves the day after the Q3 results.

Who owns Rivian

Ownership as of April 23, 2026, from the 2026 proxy statement, before the July 2026 offering. The July offering added 86 million shares, so each stake is now slightly smaller. Board members with manufacturing backgrounds include John Krafcik, a former Waymo CEO and Hyundai Motor America CEO who began in product development at Ford, and Sanford Schwartz, a former CEO of Cox Automotive.

How management changes my view

• Positive: a founder paid only for a large share-price gain, and an experienced automaker COO in charge of the R2 ramp.
• Positive: a third of the shares sit with Amazon, Volkswagen and Abdul Latif Jameel. That is part of why my bear case values Rivian at a strategic floor rather than zero.
• Negative: the CFO who built the funding plan leaves just as Rivian needs to raise more money.
• Watch: CEO pay that bypassed a shareholder vote, and dual-class voting. Both are small today, but they matter if results disappoint.

Sources: Rivian 2026 proxy statement (SEC DEF 14A); Rivian, “Rivian Announces CFO Transition Plan” (newsroom, Aug 27, 2026); GE Vernova appointment (EVWire); Varela appointment (WardsAuto); CEO award terms (Reuters via MarketScreener, Autoblog); joint-venture leadership (InsideEVs); Global Oryx (Bloomberg).

Catalysts

Risks and what would change my mind

What I expect on October 29

Deliveries are known: 19,248. Here are my estimates for the rest, so the call can be checked against the result.

• Revenue: about $2.0 billion. Automotive about $1.5 billion, using Cox’s model mix: R2 about 18% of deliveries at $58,000, and R1 and vans at about $85,000 excluding credits. Software & services about $0.55 billion.
• Automotive gross profit worse than Q2’s ($36) million. Credits fall away and R2 launch costs peak, as management flagged.
• Consolidated gross profit near breakeven. Software & services carries it.
• Adjusted EBITDA loss of about $0.5 billion. Wider than Q2’s ($379) million, consistent with management’s guide for a heavier second half.

A clean beat on automotive gross profit would be the most important signal for the thesis, more than any revenue number.