R2 Ships on Time and the Beat Is Real, But Regulatory Credits Did the Margin Work and They Leave in the Second Half

Results vs. Consensus

Rivian reported after the close on Thursday, July 30. On the headline metrics this was a clean beat across the board, and the market's initial reaction reflected that before reversing hard the following session.

Quality of Beat/Miss

• Revenue (+7.7% beat): Revenue of $1,658M grew 27.2% year over year, split between automotive at $1,143M (+23.3%) and software and services at $515M (+37.0%). Deliveries of 12,194 rose 14.4%, so roughly half of the automotive growth is volume and the balance is the regulatory credit line. Automotive revenue per delivery was $93.7K including credits; strip the credits and the underlying figure is closer to $84.9K, which is down modestly on the prior year. The CFO attributed that to "a decline in average selling prices from a higher mix of commercial van and R2 deliveries." Revenue quality is therefore mixed: the volume is genuine, the price is softening, and the credits are temporary.
• Margins (10.8%, best in company history): Consolidated gross margin of 10.8% is the highest Rivian has ever printed and compares with (15.8)% in the prior-year quarter, a 2,661 basis point swing. Underneath it, the automotive segment lost $(36)M against $(335)M a year ago. That headline improvement is real but heavily assisted. Inside the $(36)M sit roughly $108M of credits and an IEEPA tariff refund receivable that the CFO confirmed was booked in cost of revenues and sized at "roughly half" of a total previously described as being in the tens of millions of dollars. Working against those was approximately $100M of R2 ramp cost. Net of credits and adding back the ramp charge, underlying automotive gross loss is in the region of $(44)M to $(70)M depending on where the tariff refund lands, which is not materially better than the $(59)M the segment lost in the fourth quarter of 2025 on 20% fewer deliveries.
• EPS (beat on both bases): GAAP loss per share of $(0.63) beat a $(0.78) consensus; the Street's adjusted $(0.47) beat a $(0.65) number. Rivian does not publish a non-GAAP EPS, so the adjusted figure is a Street construct. The $0.16 gap between the two is almost exactly stock-based compensation: $226M of SBC on 1,325M weighted-average shares is $0.17 per share. Worth noting for the year-over-year comparison: weighted-average shares rose 14.7%, so the improvement in loss per share understates the improvement in absolute net loss, which narrowed from $(1,117)M to $(833)M.
• Free cash flow (deteriorating): The one unambiguous negative. Free cash flow of $(849)M was more than double the $(398)M of the prior-year quarter, despite a $284M improvement in net loss. Operating cash outflow of $(487)M compares with a $64M inflow a year ago. This is the line where the R2 ramp is actually being paid for, and it is moving the wrong way even as the P&L improves.

Year-over-Year Comparison

The production comparison flatters. Rivian produced only 5,979 vehicles in the prior-year quarter because the Normal plant was down for the R2 line installation, so the 111% increase measures the absence of a shutdown rather than a capacity gain. The delivery number, up 14.4%, is the honest volume comparison.

Sequential Comparison

The sequential picture is the more useful one, and it is genuinely encouraging on the operating lines. Deliveries up 17.6%, revenue up 20.1%, adjusted EBITDA loss narrowing $93M, and operating expenses essentially flat at +1.5% is real operating leverage. The cash balance rose $480M sequentially only because of the first-half equity issuance; on a free cash flow basis the quarter still consumed $849M.

Segment Performance

Automotive

Automotive revenue of $1,143M on 12,194 deliveries was the segment's best quarter since the third quarter of 2025, and the gross loss of $(36)M is the narrowest the segment has ever reported. Deliveries beat the company's own 9,000 to 11,000 expectation by a wide margin, helped by growth in both the commercial van and R1 lines alongside the first external R2 units.

"Delivery results topped our 9,000-11,000 vehicle expectations due to robust growth quarter-over-quarter in EDV and R1, coupled with the introduction of R2 deliveries." — Claire McDonough, CFO

The composition of that $(36)M is where the analytical work is. Three items move it materially, and only one of them is operating progress. First, automotive regulatory credit revenue of roughly $108M, up about $103M year over year, which drops through at close to a full margin. Second, an IEEPA tariff refund receivable, which the CFO confirmed under questioning was booked in cost of revenues rather than revenue. Third, working against both, approximately $100M of incremental cost of revenue from the R2 ramp.

"In the quarter, we recognized approximately $100 million in incremental cost of revenue due to the ramp of R2 production as compared to the production at more normalized levels." — Claire McDonough, CFO

The unit economics were laid out with unusual specificity on the call. Cost of goods sold per delivered vehicle was $96,700. Removing the $100M ramp charge, which is $8,200 per unit across 12,194 deliveries, gives $88,400, an improvement of roughly $5,000 per unit against the first quarter's $93,584. That improvement is attributable primarily to volume absorption and secondarily to the tariff refund, which means it is partly a fixed-cost leverage effect that should persist and partly a one-time credit that will not.

Assessment: The automotive segment is closer to breakeven than it has ever been on a reported basis and further from it than the headline suggests on a clean basis. Strip the credits and add back the ramp charge and the underlying gross loss is roughly $(44)M to $(70)M, which sits alongside the $(59)M the segment lost in the fourth quarter of 2025 on 9,745 deliveries. Put differently: 25% more volume, a new higher-priced product, and a full quarter of cost work have not yet produced a durable improvement in underlying automotive economics. The fourth quarter, with a second shift running and the credits gone, is the first clean read the Street will get.

Software & Services

Software and services delivered $515M of revenue at a 41.7% gross margin, generating $215M of gross profit. That single segment produced more gross profit than the entire company reported, and has done so in every quarter since the third quarter of 2025. Growth of 37% year over year came from vehicle electrical architecture and software development services performed for the Volkswagen joint venture, repairs and maintenance, and the Autonomy+ subscription, partly offset by lower remarketing revenue.

"$308 million or 60% of software and services revenue was attributable to our joint venture with Volkswagen Group." — Claire McDonough, CFO

Concentration in the joint venture rose this quarter. Volkswagen accounted for 60% of segment revenue against roughly 54% for full-year 2025. The margin expansion, from 34.3% in the prior-year quarter to 41.7%, was explicitly attributed to the joint venture engineering work, which means the segment's profitability is becoming more dependent on a single counterparty rather than less.

There is a cash-flow footnote that deserves more attention than it received on the call. In the first half of 2026, deferred revenues were a $(362)M use of cash. In the first half of 2025, they were a $403M source. That is a $765M swing, and it is the single largest driver of the deterioration in operating cash flow. Deferred revenue balances cover several things at Rivian, including service plans and charging, but a segment whose revenue is 60% joint-venture engineering and whose deferred balance is now draining rather than building is recognizing revenue it collected in prior periods. The P&L benefit is real; the cash has already been banked.

Assessment: Software and services remains the reason Rivian's consolidated gross profit is positive at all, and at 41.7% margins it is a genuinely valuable asset. Two things have changed for the worse this quarter, though neither is yet alarming. Volkswagen concentration increased, and the deferred revenue balance flipped from a cash source to a cash use. Investors underwriting the segment as a durable annuity should watch the deferred revenue line in the third-quarter cash flow statement more closely than the revenue line in the income statement.

Key Operating Metrics

Quarterly regulatory credit figures are derived from the company's disclosed $164M first-half total and the stated $103M year-over-year increase in the second quarter; only the first-half aggregate and the year-over-year delta were given directly.

Key Topics & Management Commentary

Overall Management Tone: Confident on product and deliberately hedged on economics, a combination that was consistent rather than contradictory. Management led with the R2 launch and the reservation-conversion surprise, then spent the financial commentary pre-committing the Street to a weaker third quarter and a back-loaded fourth. The posture is measurably less celebratory than the fourth-quarter 2025 call, where the framing was that the inflection had arrived; here the framing is that the inflection is being built and the bill is being paid in the interim. Where management was least convincing was on disclosure: no R2 unit count, no reservation figure, no conversion rate, and no sizing of the tariff benefit, each declined in response to a direct question.

1. R2 Shipped on Time, But Not the R2 the Thesis Was Built On

Rivian committed in February to external R2 customer deliveries in the second quarter of 2026. It delivered on that. External deliveries began in June, the vehicle has been well reviewed, and the CEO is driving one daily. In a sector where launch slippage is the norm, hitting the date matters and management earned the credit it took.

"In the second quarter, we began deliveries of R2 vehicles to external customers." — RJ Scaringe, CEO

The complication is which R2 is shipping. The vehicle in customers' hands is the Launch Edition, described on the call as a $58,000 configuration and the most expensive R2 variant. The February thesis, and the mass-market total addressable market argument that underpinned the upgrade, was built on a roughly $45,000 vehicle. The trims that reach that price point are now an early-2027 event.

"we are looking forward to introducing the other trims, our mid-spec and our base or standard spec, as we look at the early part of 2027." — RJ Scaringe, CEO

Assessment: Launching on time with a high-trim, low-complexity configuration is the correct manufacturing decision and management explained the logic well. It is also a deferral of the investment case. What has been de-risked is production start; what has not been tested is whether demand exists at volume for a $45,000 Rivian, because that vehicle will not be on sale for another two to three quarters. Anyone underwriting R2 as a mass-market product is still underwriting an untested proposition, just with better manufacturing evidence than they had in February.

2. The Regulatory Credit Mask

Automotive regulatory credit revenue was approximately $108M in the quarter, an increase of about $103M year over year, and $164M across the first half. Credits carry essentially no associated cost, so they translate almost directly into gross profit. Against a consolidated gross profit of $179M, they are the majority of the number.

Management was transparent about this when pressed on the second-half bridge, and the disclosure is the most important thing said on the call.

"The bigger driver is really the absence of regulatory credits in second half versus first half, and then the ongoing investments in R&D and SG&A." — Claire McDonough, CFO

Assessment: This is the quarter's central fact and the reason the stock fell on a beat. The margin trajectory that the market has been extrapolating since the fourth quarter of 2025 has been substantially credit-assisted, and the assistance stops. The clean automotive gross loss, excluding credits and adding back the R2 ramp charge, is not visibly better than it was two quarters ago. The company is not being deceptive about this; it disclosed the numbers and answered the question directly. But the Street was modeling a margin curve, and a meaningful part of that curve was a subsidy line item.

3. The $100M Ramp Charge and What "Normalized" Means

The $100M of incremental cost of revenue attributed to the R2 ramp was probed carefully, and the answer was more useful than the headline number. It comprises two distinct things: genuine inefficiency such as expedited freight and short-term supplier premiums, and period-expensed cost that at higher volumes would have been absorbed into inventory.

"the $100 million reflects a combination of ramp-related inefficiencies such as expedited freight cost out of the gates or short-term premiums with suppliers that we anticipate coming off as we increase our production volumes over the course of the next couple of quarters. It's also cost that we expensed in period that would have naturally been absorbed into inventory, for example, if we were at more of a Q4 run rate of production volume." — Claire McDonough, CFO

Management also clarified that the reference point for "normalized" is fourth-quarter 2026 run-rate production, not fully ramped R2 production. That distinction matters for anyone trying to reverse-engineer a breakeven volume: the $100M is measured against a level that itself remains well below the plant's capability.

Assessment: The decomposition is credible and the transparency is welcome. The practical implication is that the ramp drag does not disappear cleanly. The expedite-and-premium portion should decay through the fourth quarter, but the absorption portion only reverses when volume rises, which makes it self-correcting but slow. Only about two-thirds of the quarter carried the R2 cost structure, so the third quarter will absorb a full quarter of it plus second-shift labor, which is precisely why management guided third-quarter automotive gross profit lower.

4. Second Shift, and Why the Fourth Quarter Is the Whole Ball Game

R2 production started on a single shift. The second shift is targeted for the end of the third quarter, and management was explicit that it contributes nothing material to volume until the fourth quarter.

"We are not expecting any material contribution to volumes with the second shift in this quarter, but we will see it in quarter four." — Javier Varela, COO

The delivery arithmetic makes the concentration plain. Full-year guidance of 65,000 to 70,000 against a first-half actual of 22,559 requires 42,441 to 47,441 units in the second half, and management confirmed the mix is weighted toward the fourth quarter. With the third quarter carrying single-shift output plus a full quarter of ramp cost, the fourth quarter has to deliver both the volume and the margin proof point in a single twelve-week window.

Assessment: The entire 2026 investment case now rests on one quarter. That is a poor risk structure for a shareholder, independent of whether management executes. A two-week supplier problem in October, a software validation delay, or a slower second-shift training curve does not just cost volume, it removes the only clean margin datapoint the year was going to produce. The asymmetry here is unattractive: hitting the fourth quarter confirms what management has already told you, while missing it invalidates the exit-rate thesis entirely.

5. Input Cost Inflation Is a New and Unquantified Headwind

Memory, raw materials and logistics costs were named repeatedly as offsets to the volume and credit benefits, and they were the reason the adjusted EBITDA guidance improved by only $50M at the midpoint despite better credits and 3,000 additional units.

"Current macro and geopolitical factors are creating added complexity, cost, and uncertainty, our team continues to work hard to manage supply chain risks and increasing commodity and memory costs." — Claire McDonough, CFO

Memory is the specific worry. Rivian's vehicles are compute-heavy by design, and the autonomy roadmap increases silicon and memory content per vehicle rather than reducing it. The company is buying into a DRAM market where hyperscaler demand is absorbing supply, which is an unusual position for an automaker to be in and a direct consequence of the software-defined-vehicle strategy.

Assessment: This is a genuinely new risk that did not feature in the February call, and it was not quantified. It also cuts against the R2 cost-down story in a specific way: the vehicle's bill of materials advantage over R1 was supposed to come partly from architectural simplification, and memory inflation erodes exactly that. Management deserves credit for flagging it unprompted, but an unsized headwind on a business with negative automotive gross margin is not something a model can absorb gracefully.

6. The Capital Stack: $7.2B Available, "Over $14 Billion" Advertised

Rivian ended the quarter with $5,310M of cash and short-term investments and sold 86.25M Class A shares in July for approximately $1.3B, bringing pro forma available liquidity to $7,163M. Management framed the total capital picture considerably more expansively.

"Our total available liquidity and targeted future capital is over $14 billion to support Rivian's investments in growth initiatives." — Claire McDonough, CFO

The gap between $7.2B and $14B is entirely conditional capital: an up-to-$4.5B Department of Energy loan tied to the Georgia plant, $1B of non-recourse debt from Volkswagen, and $250M of equity from Uber, with roughly $1.1B beyond that unspecified. The Volkswagen and Uber tranches were both described as "subject to the completion of certain conditions." The Department of Energy facility requires Rivian to pre-fund equity commitments and reserve accounts before drawing, which is what the July equity raise was partly for, and the financing roadmap in the earnings materials places the first draw in early 2027.

There is a continuity point here. In February, management guided the Street to expect $2B of additional Volkswagen capital during 2026. What is now described is $1B of non-recourse Volkswagen debt plus $250M of Uber equity, both conditional, and a $1.3B equity raise happened in the interim.

Assessment: The liquidity position is adequate for the next eighteen months and there is no near-term solvency question. The framing is what deserves scrutiny. Presenting $14B as the relevant capital figure when roughly half is milestone-gated and the largest single component cannot be drawn until 2027 is optimistic accounting of a balance sheet that is currently funded by dilution. Shareholders paid for this quarter's liquidity, and at the current burn rate they will likely be asked to pay again.

7. Dilution Is Now a Material Part of the Return Equation

Shares outstanding went from 1,240M at December 31, 2025 to 1,362M at June 30, 2026, and the July offering adds 86.25M more, taking the pro forma count to approximately 1,448M. That is a 16.8% increase in seven months. First-half financing activities included $1.3B of capital stock proceeds before the July raise added roughly another $1.3B.

Set that against free cash flow. The first half consumed $1,924M, compared with $2,489M for all of 2025. Operating cash outflow alone was $(1,190)M against $(124)M in the prior-year first half, driven principally by a $765M adverse swing in deferred revenues and a $323M swing in inventory as R2 stock was built.

Assessment: A company burning cash faster than last year while issuing equity to fund it is transferring the cost of the ramp from the income statement to the share count, where it is less visible and permanent. The improvement in adjusted EBITDA is real, but adjusted EBITDA is not what funds a factory. Any per-share valuation framework needs to work off 1,448M shares and assume further issuance, not off the 1,325M weighted average that flattered this quarter's loss per share.

8. Autonomy: A Credible Roadmap on a Long Fuse

Autonomy consumed more airtime than any other topic and is where the CEO's conviction is highest. The roadmap as laid out: point-to-point capability by the end of 2026 at what was described as Level 2++, hands-off with eyes on the road; hands-off and eyes-off Level 3 in 2027; Level 4 in 2028, including robotaxi variants operated with Uber. In-house RAP1 silicon and lidar arrive with third-generation hardware at the end of this year.

"our in-house silicon RAP1 is on track. As you know, that underpins what we call our Gen 3 autonomous hardware platform." — RJ Scaringe, CEO

The architectural argument is the strongest part of the pitch: a single end-to-end trained large driving model spanning Level 2 through Level 4, with sensor topology and inference capability differing across variants but the model shared. That collapses what used to be two separate engineering programs into one, and Rivian's growing R2 fleet feeds the data flywheel.

Monetization has started. Autonomy+ has been a paid service since April and take rates were described as trending positively, though no figure was given. R&D is guided to step up in the second half on GPU sourcing for autonomy training.

Assessment: The strategy is coherent and the vertical integration argument on cost is real. It is also expensive, back-end loaded, and being funded by shareholders during a period when the core automotive business does not cover its own cost of goods. The 2028 Level 4 target sits three years out on a company with under two years of unconditional liquidity. Treat autonomy as option value that justifies patience, not as a valuation input for 2026 or 2027.

9. Amazon Scales, Everyone Else Still Does Not

The commercial van business is quietly the healthiest part of the automotive segment. Amazon now runs over 40,000 Rivian electric delivery vans, the platform has passed one billion cumulative miles, and Rivian is developing larger-battery and all-wheel-drive variants at Amazon's request. Van volume growth was cited as a contributor to the delivery beat.

Third-party fleet adoption remains the gap, and the CEO was candid about it.

"Now, that has taken longer than we would have expected, or we would have hoped for in terms of other fleets adopting a strategy towards electrification and a strategy of adopting not just an electric vehicle, but a highly connected and intelligent vehicle." — RJ Scaringe, CEO

Assessment: Four years after the Amazon relationship began, it remains effectively the entire commercial business. The total cost of ownership argument is evidently persuasive to the customer that co-designed the vehicle and has not yet been persuasive to anyone else at scale. The van mix also dilutes automotive average selling price, which management named as a driver of the year-over-year decline. This is a good business with one customer, and the concentration is not improving.

10. Cost Structure Versus Chinese Competition

Asked what Rivian can learn from Chinese electric vehicle platforms, the CEO gave an unusually direct answer that manufacturing technique is not the differentiator.

"I think the big difference that exists between U.S. cost structure and Chinese cost structure is the input cost. You have a much, much lower labor cost in China, much, much, much lower capital cost structure, where in many cases, the capital cost is zero, meaning it's being provided by the local government." — RJ Scaringe, CEO

Assessment: This is an honest and somewhat uncomfortable framing. Rivian is saying the gap is structural and policy-dependent rather than something engineering can close, which is almost certainly correct and also means the company's cost position depends on trade policy holding. It reads as a reasonable answer to a fair question, but a bear will note that management has just described its own cost disadvantage as not addressable through execution.

Guidance & Outlook

Three of the four quantitative revisions went the right way, and the capital expenditure cut is the most underappreciated positive in the print. Reducing capex $250M while raising delivery guidance is the opposite of what a company struggling with a ramp does, and management attributed it to project efficiencies and timing of spend.

The language change on automotive gross profit is the item to mark. In February the framing was that 2026 would be "a transformational year for our automotive gross profit." On this call it became a transition year.

"As a reminder, we believe this is a transition year for the automotive segment's path towards long-term profitability as we scale R2." — Claire McDonough, CFO

Implied second-half ramp: First-half deliveries of 22,559 against full-year guidance of 65,000 to 70,000 requires 42,441 to 47,441 units in the second half, which management stated as approximately 42,400 to 47,400 and confirmed is weighted toward the fourth quarter. At the midpoint that is roughly 22,500 units per quarter against a first-half average of 11,280, so second-half quarterly run-rate must approximately double.

Implied second-half profitability: First-half adjusted EBITDA was $(851)M. Full-year guidance of $(2,000)M to $(1,800)M implies a second-half loss of $(949)M to $(1,149)M, a midpoint of $(1,049)M. The second half is therefore guided roughly 23% worse than the first on close to double the volume. The bridge, per management, is the absence of the $164M of first-half regulatory credits, continued R2 ramp inefficiency in the third quarter, and a step-up in research and development driven by GPU procurement for autonomy training.

Implied second-half capital expenditure: First-half capex of $734M against full-year guidance of $1,700M to $1,800M implies $966M to $1,066M in the second half, roughly 1.4x the first-half rate. Combined with the EBITDA path, second-half free cash flow burn is unlikely to improve much on the first half's $(1,924)M.

Exit-rate commitment: Management reaffirmed that R2 achieves positive gross profit as part of the 2026 exit rate, enabled by higher volumes and fixed-cost absorption. This is the single commitment that matters and the one to grade in February.

Guidance style: Rivian has guided conservatively and hit its numbers throughout 2025 and the first half of 2026, and the second-quarter delivery beat against a 9,000 to 11,000 expectation continues that pattern. The delivery guidance looks achievable. The margin commentary is where management has changed its own framing twice in six months, which argues for discounting the exit-rate language more heavily than the volume language.

Analyst Q&A Highlights

Whether the R2 ramp is supply-gated or demand-gated

The opening question went straight to the structural issue: over the next year, is R2 volume limited by how fast Rivian can build or by how many people want one. Management answered the supply half thoroughly, describing supplier synchronization as the binding constraint and the reason for starting on a single shift, then answered the demand half with enthusiasm but no metrics. The reservation-to-order conversion on the Launch Edition was described as meaningfully above internal expectations.

Q: "As you think about R2 over the next year, do you think the pace of deliveries is going to be gated more by supply or demand?"
— Mark Delaney, Goldman Sachs

A: "The biggest category of focus for us has been supply base and making sure that all of our suppliers are ready to continue ramping to much higher levels as we look at the back half of the year, and in particular, as we start to plan for running, fully utilizing the plant and running across two shifts."
— RJ Scaringe, CEO

Assessment: The answer is that it is supply-gated, and that is the better of the two problems to have this early. Supply constraints resolve with time and capital; demand constraints do not. The caveat is that demand has not actually been tested, because the only variant on sale is the top trim to a pre-existing reservation base. The genuine demand test arrives with the standard trim in 2027.

The tariff refund and the second-half EBITDA bridge

The sharpest exchange of the call paired two observations that did not obviously fit together: an unquantified tariff benefit inside second-quarter gross profit, and full-year guidance implying a worse second half despite a stated positive gross margin exit rate. Management sized the tariff item only relatively, as roughly half of a total previously described as being in the tens of millions, and then gave a clear answer on the bridge that identified regulatory credits as the dominant factor.

Q: "I noticed in the second quarter gross profit commentary, you mentioned an IEEPA tariff refund benefit. I'm curious how much that was because I didn't see it in the full-year outlook. Just relatedly, it looks like the guidance implies a steeper EBITDA loss in the second half versus first half, despite the positive gross margin exit rate commentary. Just trying to bridge those two aspects."
— Rajat Gupta, JPMorgan

A: "As we think about the second question on the EBITDA loss being steeper in the second half of the year, one of the big differences there is we did have $164 million of regulatory credits that benefited our gross profit outlook in the first half."
— Claire McDonough, CFO

Assessment: This is the exchange that reframed the quarter, and it is the most likely proximate cause of the next session's decline. The question forced management to state on the record that the first half's margin performance was substantially credit-funded and that the support is not there in the second half. The answer was honest and complete on the bridge; it was evasive on the tariff sizing, which remains the one number in the automotive gross profit walk that cannot be independently reconstructed.

Defining the $100M ramp charge and the meaning of "normalized"

A follow-up sought the baseline against which the $100M of incremental cost was measured, specifically whether it assumed one shift or two and whether mix was normalized. The answer separated the charge into transitory inefficiency and volume-driven absorption, and pinned the comparison to fourth-quarter run-rate production rather than fully ramped R2 output. Pressed on whether that allows the Street to triangulate a breakeven volume, management accepted the framing while narrowing it.

Q: "Just maybe, the comment about the $100 million in incremental cost of revenues because of ramp of R2 compared to normalized levels. Just maybe some context, what are you classifying as normalized? Is that one shift, two shifts, and are you also assuming a more normalized mix in that analysis?"
— Joseph Spak, UBS

A: "That's generally a fair way to think about it. As we talked about it being really at sort of Q4 normalized levels, that's not the full extent of R2 fully ramped up."
— Claire McDonough, CFO

Assessment: Management gave more than it had to here, and the qualification is the useful part. Anchoring "normalized" to fourth-quarter volumes rather than full ramp means the $100M understates the eventual absorption benefit but also means the reference point is one the company has not yet achieved. The charge is not a one-time item that vanishes in January; it decays as volume rises, which makes the fourth-quarter delivery number the direct input to the margin outcome.

The cost-per-unit bridge and what else moved

An analyst noted that sequential cost of goods per unit rose by only a few thousand dollars despite absorbing the $100M charge, implying underlying improvement elsewhere. The response supplied the full arithmetic, and a follow-up established that the tariff refund had been booked in cost of revenues rather than as revenue, which is the detail that makes the reported automotive gross loss reconcilable.

Q: "To only see COGS per unit up a few thousand dollars quarter-over-quarter, despite that incremental $100 million, says that there's probably something incrementally that's going on quarter-over-quarter on the cost side, whether it's within EDV or R1."
— Dan Levy, Barclays

A: "As you look at the quarter, we had a $96,700 a unit of COGS. If you were to back out the $100 million on a per-unit basis, that's about $8,200 a unit. That would've gotten us to $88,400. That's about a $5,000 per unit improvement on a quarter-sequential basis relative to where we were in Q1."
— Claire McDonough, CFO

Assessment: The most useful disclosure on the call and a genuine improvement, but the attribution matters. Management named volume absorption as the primary driver and the tariff receivable as a secondary one. Volume absorption persists and compounds; the tariff credit does not. The $5,000 per unit is therefore a mix of durable and non-durable, and the durable portion is the number that determines whether the fourth-quarter exit rate commitment holds.

Why the guided margin dip did not appear, and the missing R2 unit count

One line of questioning combined two observations that the company had left uncomfortable: that the gross margin deterioration management had warned about did not materialize, aided by tariff and credit income, and that Rivian still had not said how many R2s it actually delivered. Management addressed the timing question directly, explaining that R2 cost structure hit only about two-thirds of the quarter. The unit-count question went unanswered.

Q: "I think you still haven't told us exactly how many R2s you delivered. It's just maybe a handful. You had warned us about the risk to gross margin as you ramp up. In fact, I know there was a bit of outside help in Q2 from IEEPA, from ZEV income. The dip in the gross margin has actually not really happened."
— Philippe Houchois, Jefferies

A: "As we think about the second quarter, you can think about us converting from a pre-production environment to our start of production then start of our cost structure, hitting our cost of goods sold for roughly 2/3 of the quarter. It was not a full quarter's impact."
— Claire McDonough, CFO

Assessment: The most pointed question of the call, and it named the credit and tariff assistance explicitly before management volunteered it in that context. The two-thirds-of-a-quarter explanation is a satisfying answer to the margin question and makes the third-quarter guide-down internally consistent. The silence on R2 unit volume is the more telling half. A company delivering a number it was pleased with would disclose it.

Research and development trajectory and autonomy compute

Research and development was roughly flat sequentially at $466M, which sat oddly against repeated statements about escalating autonomy investment. The explanation was a mix shift within the line: engineering design and development spend fell as R2 moved into production, masking a significant increase in autonomy spend underneath. A second-half step-up was guided, tied specifically to graphics processor procurement for model training.

Q: "R&D, I think, was flat sequentially. I know you're making a lot of investments in autonomy. Curious how would you think about R&D kind of trending in the second half of the year."
— Itay Michaeli, TD Cowen

A: "As we look at the second half of the year, we do anticipate seeing a step-up that's predominantly related to GPU sourcing for our autonomy training that will continue to ratchet up over the course of the second half of the year."
— Claire McDonough, CFO

Assessment: The flat research line was flattering and the company said so, which is to its credit. The forward implication is less comfortable. Autonomy compute is a fixed cost that scales with ambition rather than with vehicle volume, it is being procured into a tight market, and it lands in the same half in which regulatory credits disappear. Operating expenses have been remarkably well controlled at +1.5% sequentially; that discipline is guided to break in the second half.

Long-term financial targets and the absent Investor Day

A question observed that the company's long-term financial targets date to the period around its initial public offering and asked whether updated targets might be forthcoming. Management distinguished between the technology-focused event planned for later this year and a broader financial event that remains unscheduled.

Q: "It's been a while since the IPO and long-term financial targets were put out. Could this be something where things have changed for better or for worse, for the better now? Could we see updated long-term targets for people to anchor around?"
— Chris Pierce, Needham

A: "For the Autonomy & AI Day, it will be more focused on the technology roadmap. We're working over the coming quarters to think about the right time to set up a broader-based Investor Day as well, that would be more focused on longer-term financial targets."
— Claire McDonough, CFO

Assessment: A reasonable question deferred politely. The company will happily host an event about technology and will not yet commit to one about economics, at a moment when it is asking shareholders to fund a 2028 autonomy roadmap and has just diluted them 17% in seven months. "Over the coming quarters" is not a commitment. The absence of refreshed long-term targets is a governance gap, not merely a communications preference.

What They're NOT Saying

1. R2 units delivered: The single most obvious metric of the quarter, asked directly, and not answered. Rivian disclosed total deliveries of 12,194 and that external R2 deliveries began in June, but not the split. Given that only about two-thirds of the quarter carried R2 production cost and the ramp began on one shift, the number is likely small. Withholding it prevents the Street from computing an R2 contribution margin.
2. R2 reservations and conversion rate: A sixth consecutive quarter without a demand metric. Management said conversion was "meaningfully higher" than internal expectations and then explicitly declined to quantify it. An unusually good number is normally disclosed; a qualitative superlative in place of a figure invites the assumption that the base is small.
3. The size of the IEEPA tariff benefit: Described only as "roughly half" of a total previously sized in the tens of millions, and confirmed to be booked in cost of revenues. Without the figure, the automotive gross profit walk cannot be fully reconstructed, which conveniently obscures the underlying run-rate.
4. Second-half regulatory credit revenue: Management said credits are largely absent in the second half without giving a number. Given that credits appear to be roughly $108M of a $179M consolidated gross profit, the precise second-half assumption is a material modeling input that was left to inference.
5. R2 average selling price and trim pricing: The Launch Edition was identified at $58,000. No pricing was given for the mid-spec or standard trims arriving in early 2027, and no target average transaction price. The mass-market thesis depends entirely on a price point the company has not confirmed.
6. Autonomy+ take rate: Described as "trending positively" since monetization began in April, with no rate, no subscriber count, and no revenue contribution. Autonomy is the largest single research and development commitment and the least measurable.
7. Full-year revenue guidance: Still not provided, for a third consecutive year. With a new vehicle at a new price point, lumpy joint-venture revenue recognition, and a regulatory credit line that swings by $100M between halves, the dispersion in Street revenue models is entirely self-inflicted.
8. Updated long-term financial targets: Deferred to an Investor Day that has no date. The operative targets still trace to the period around the initial public offering.

Market Reaction

• Pre-print setup: Closed at $16.83 on July 30, down 14.6% year to date against the S&P 500's +8.7%, down 3.0% over the trailing thirty days and up 29.2% over the trailing twelve months. The fifty-two week closing range entering the print was $11.64 to $22.45, so the stock was in the middle of its range and had given back most of the post-fourth-quarter rally.
• After-hours move: Traded higher on the print, roughly 1.6% to 4% depending on the timestamp, consistent with a beat on every headline metric.
• Next-day session (July 31): Opened at $17.01, a 1.1% gap up, touched a high of $17.18, then declined through the day to close at $15.22, down 9.6% or $1.61. The session low of $15.14 came near the close.
• Volume: 54.3M shares against a thirty-day average of 37.5M, or 1.4x normal.
• Context: The S&P 500 rose 0.7% on the session, so the move was entirely idiosyncratic. Lucid fell 6% on no company-specific news, which reads as sympathy rather than sector information.

The shape of the session matters more than the magnitude. This was not a gap-down: the stock opened above the prior close, near the session high, and then sold off continuously to finish within eight cents of its low on 1.4x volume. That pattern is distribution during the cash session, not an overnight repricing. The after-hours buyers reacted to the press release; the sellers reacted to the call.

The specific thing that changed between the press release and the close was the second-half bridge. The release showed a beat on every line. The call established that roughly $108M of the $179M gross profit was regulatory credits, that those credits are largely absent in the second half, that third-quarter automotive gross profit is guided down again, and that full-year guidance implies a second-half adjusted EBITDA loss roughly 23% steeper than the first half despite close to double the deliveries. A reader of the release would model an accelerating margin curve. A listener to the call would not.

Two structural pressures compounded it. The July follow-on offering of 86.25M shares was already an overhang, and the print did nothing to suggest the equity would not be needed again. And the stock had re-rated into the event, which removes the cushion that usually absorbs a mixed message. A 9.6% decline on a comprehensive beat is a repricing of the margin path, not a verdict on the quarter.

Street Perspective

Debate: Does the fourth-quarter exit rate arrive, and does it prove anything?

Bull view: Management has hit every operational commitment it has made in the last three quarters, including an on-time R2 launch that few sell-side models fully believed. The second shift is on schedule, the capex guide was cut while the delivery guide was raised, and the per-unit cost improvement of roughly $5,000 sequentially shows the absorption math working. Positive R2 gross profit on the fourth-quarter exit rate is the natural output of doubling volume against a largely fixed cost base.

Bear view: The exit rate is a single-quarter, single-product measurement that management defines and reports, arriving in the same quarter it is promised, with no external verification and no disclosed R2 unit volume to check it against. Even if achieved, it says nothing about the full-year 2027 economics of a mass-market trim priced $13,000 below the Launch Edition currently shipping.

Our take: The bears have the better of this. Positive R2 gross profit at fourth-quarter volumes on a $58,000 vehicle is a meaningfully easier test than positive gross profit on a $45,000 vehicle at scale, which is what the investment case actually requires. We expect management to clear the bar it set and expect the market to discover in 2027 that clearing it proved less than it seemed. The metric to demand is automotive gross profit excluding regulatory credits, which no one has yet committed to disclosing.

Debate: Is the software and services segment a durable asset or a Volkswagen contract?

Bull view: The segment generated $215M of gross profit at a 41.7% margin, more than the whole company earned, and has grown 37% year over year. It validates the platform strategy, gives Rivian a second revenue engine independent of vehicle volume, and makes the partner deeply invested in the relationship's continuation. Consumer software monetization through Autonomy+ has only just begun.

Bear view: Sixty percent of segment revenue came from a single joint venture this quarter, up from roughly 54% for full-year 2025, so concentration is worsening rather than improving. The margin expansion was explicitly attributed to that same joint-venture engineering work. And the deferred revenue line swung $765M against the prior-year first half, which means the segment is increasingly recognizing revenue whose cash was collected earlier.

Our take: Both are right about different time horizons, and the bear point is underweighted by the market. The segment is genuinely valuable and genuinely concentrated, and the cash-versus-recognition divergence is the part almost no one is watching. Engineering services revenue against a fixed development program is not an annuity; it has a completion date. Until consumer software becomes a visible share of the segment, this is contract revenue wearing a platform multiple.

Debate: Is $15 a discount to a de-risked story or fair value for a company that keeps issuing equity?

Bull view: At $15.22 the stock trades at roughly $22.0B on approximately 1,448M pro forma shares, against 2026 revenue tracking toward $7.3B to $8.0B. That is under 3x forward revenue for a business growing deliveries more than 50% with a launched product, a cut capex guide, an improving EBITDA line and $7.2B of liquidity. The 9.6% decline on a comprehensive beat is an overreaction that creates the entry point.

Bear view: Free cash flow burn is accelerating, not moderating: $1,924M in the first half against $2,489M for all of last year. The share count is up 16.8% in seven months and further issuance is likely. Half the advertised $14B of capital is conditional and the largest tranche cannot be drawn until 2027. Buying a revenue multiple on a company that funds itself by issuing the shares you are buying is a losing structure.

Our take: The revenue multiple is not demanding and the bull arithmetic is sound as far as it goes, but it prices the top line while the risk sits in the funding line. The distinguishing question is not whether Rivian is cheap on 2026 revenue; it is how many shares exist when the automotive segment finally earns a gross profit. On the current trajectory that is a 2027 event at the earliest, funded by shareholders in the interim. That combination supports fair value, not a premium.

Model Update Needed

Valuation impact: At the July 31 close of $15.22 on approximately 1,448M pro forma shares, Rivian carries a market capitalization of roughly $22.0B, or about 2.9x the midpoint of our revised 2026 revenue range. That multiple is not demanding for the growth rate, and it is the reason this is a Hold rather than an Underperform. What it does not capture is that the company will consume $3.7B to $4.0B of cash this year and has funded itself with roughly $2.6B of equity in seven months. Our fair value estimate is $17, implying 11.7% upside from the current price, which is not sufficient compensation for the execution concentration in the fourth quarter and the probability of further issuance. We would revisit toward the high end of a $17 to $19 range on a clean fourth-quarter automotive gross profit print, meaning positive excluding regulatory credits, and toward $12 to $13 if the second shift slips or the exit-rate commitment is withdrawn.

Thesis Scorecard Post-Earnings

Overall: Weakened against the standing thesis. The operational pillars held or improved, and the launch execution was better than we expected in February. The financial pillars did not. The automotive gross profit inflection that justified the February upgrade has not occurred on a clean basis, the mass-market vehicle that anchors the addressable market argument has slipped two to three quarters, and the funding cost has risen materially through dilution. Three of the four bull pillars moved from ON TRACK to AT RISK; two of three bear points escalated and a fourth was added.

Action: Reduce to a market weight. This is not a broken story, and the February thesis was directionally right about the product. It was wrong about the timing and the price shareholders would pay to get there. We would rather own this after the fourth-quarter print, with a clean automotive gross profit number in hand and the equity need clarified, than pay for the exit-rate commitment in advance.