Replace Brandi’s Car
Updated August 23, 2026
The short version
We're down to three cars:
- 2026 Volvo XC90 T8: the plug-in hybrid, the modern version of the car you have now
- Volvo EX90: the fully electric XC90
- Rivian R1S: still in the running, but it has to prove it fits in the garage
The two Volvos cost about the same to buy, so choosing between them comes down to one question: are you ready to give up gas entirely? The research is done; that question is now a test-drive question.
What's left is physical:
- Test drive the XC90 T8 and EX90 back to back
- Park a friend's R1S in our garage to settle whether it fits
- Take one photo of the electrical panel to price a home charger
- Try Off Road mode on your current car on the driveway (it raises the car 1.6 inches and we've never used it)
Why replace it now
Your car has run well and nothing big has broken. It's also ten years old, past the point where the factory warranty can be extended, and old enough that expensive failures are realistic. The specific worry: it has the first-generation version of the rear electric motor (Volvo calls it the ERAD), which is the least reliable version of this platform's one signature failure. A replacement runs over $10,000 at a dealer, and the whole car is only worth about $11,000 to $13,000 as a trade-in.
Waiting also has a price. The car is losing about $2,500 of value a year at this point. So there's no deadline, but the plan is to buy on our schedule, before a breakdown forces it and kills the trade-in value. Realistically that means sometime in the next 12 months.
What the car has to do
- Three rows. You haul the kids plus their friends.
- About the same size as the XC90.
- Clears our driveway without scraping.
- Electric or hybrid.
- CarPlay is Joe's preference. Both Volvos have it, wireless and standard. Rivian has said it will never add it. Since it's your car and you care less, this stayed a preference and didn't cut anything.
The buying rules we're using
- Avoid the first year of a new model (let someone else absorb the bugs).
- Avoid the last year before a redesign (it's instantly old).
- The sweet spot is years 2 to 4 of a generation.
- It's new-model-year turnover time right now (August), so discounts on leftover 2026s are live.
What we measured
The garage is the real constraint. Your car is 194.9 inches long with about 6 inches to spare, so anything much over 201 inches doesn't fit. That eliminated the Lexus TX and the Toyota Grand Highlander outright. The Rivian is 200.8 inches, a tenth of an inch inside the line on paper, which is why it gets a real-world fit test instead of a spec-sheet verdict.
The driveway eliminated nothing, which was a surprise given it was the requirement we started with. The realization that got us there: Joe's Macan uses the same driveway, sits lower than anything we'd buy, and only occasionally scrapes, so it's the true worst case. Every candidate clears the driveway better than the Macan does. The scraping also happens at the bottom, where the ramp meets the street, which means the front approach angle is what matters and ground clearance was the wrong thing to shop on all along. Full data in Appendix A.
One bonus finding: your current car has an Off Road mode that raises it 1.6 inches at speeds under 25 mph, which is exactly the driveway situation. If using it stops the scraping, the driveway is easier than we thought.
What got cut
| Car | Why it's out |
|---|---|
| Kia EV9 | The value play of the group (about $60k with the strongest warranty), but Joe's call: "pretty sure my wife isn't going for a Kia." It also had the worst predicted reliability of anything we looked at. |
| Hyundai Ioniq 9 | Same brand call, and it's a first model year anyway. |
| Lexus TX 550h+ | Too long for the garage (203.1 in). |
| Toyota Grand Highlander Hybrid | Too long for the garage (201.4 in). |
| Audi Q7 / Q9 | The Q7 is the last year of an old generation with no plug-in or electric option here; the Q9 is a first model year and too long. A redesigned Q7 plug-in around 2028 could be interesting, but that's outside our window. |
The three cars
2026 XC90 T8: the known quantity
The case for it: dimensionally identical to your car, the seat you already like with a fully modernized interior, and zero change to how you live with it. It keeps charging on the regular garage outlet, and its 32 miles of EPA electric range covers most of your average day (34 miles), so it runs mostly electric with gas for the long days.
- Price: $85k to $91k sticker loaded, but dealers are discounting 2026s hard as the 2027s arrive. Realistic price after negotiating: $79k to $85k.
- One timing wrinkle: the 2027 gets a new powertrain (more electric range, a little less power). So a 2026 is the last year of a proven setup and a 2027 is year one of a new one. Neither is clean under our buying rules; the discounted 2026 looks like the stronger side.
- Reliability: the mechanical story is good after ten years of refinement. The rear electric motor that's the worry on your 2016 was redesigned in 2022 and techs now call it very reliable. The complaints on new ones are software: infotainment freezes and audio glitches. Roughly $850 a year to maintain, $2,000 a year to insure.
- Resale: the best of the three. Worth roughly $38k after five years.
EX90: the electric XC90
The case for it: the same size and feel as what you like, fully electric, and it physically fits better than the spec sheet suggests (narrower with mirrors out than your car, and it clears the driveway with margin).
- Price: $86k sticker for the loaded version, $80k to $83k negotiated on a purchase. The real deal is the lease (below).
- Reliability: the 2025s had a rough first year, headlined by a cluster of onboard-charger failures that killed home charging and took months to fix. The 2026 got a redesigned electrical system, and so far it has zero complaints and zero recalls filed with the government, with the caveat that 2026s are still young. The question to put to the dealer: does the redesign actually fix the 2025 charger failure?
- The lidar mess: Volvo sold 2025 EX90s promising features from a roof-mounted sensor, the sensor supplier went bankrupt, the features are never coming, and there's now a class action. The 2026 dropped the sensor entirely, so it doesn't affect us directly, but a story like that sits on a model's used values for years.
- Resale is the weak spot: used 2025s have already lost about a third of their value in 18 months. Over five years the EX90 gives back roughly $7k to $8k more than the T8. The two cost the same to buy, so that gap is the real price of going full electric.
- The lease escape hatch: Volvo is propping the EX90 up with $7,500 of lease cash and a resale assumption the used market doesn't support, which means Volvo absorbs the depreciation risk instead of us. Roughly $1,080 a month all-in, about $39k over three years, walk away clean at the end. If there's a real chance we wouldn't keep it long, leasing beats buying.
- Needs the home charger install ($1k to $2k one time, covered below).
Rivian R1S: the one that has to fit
The case for it: the car you actually named interest in, and the only one that makes the driveway a non-issue outright. It also carries the most practical downsides of the three.
- The gate: 200.8 inches against a garage that fits maybe 200.9. A friend has one; parking it in our garage settles it.
- No CarPlay, ever. Rivian is betting on its own software instead.
- Price: $88k to $92k the way we'd option it, and Rivian barely discounts. Figure $5k to $8k more than either Volvo.
- Reliability is genuinely strange: Consumer Reports ranks Rivian dead last in reliability and first in owner satisfaction, three years running. Owners love these cars even when they break. The second-generation engineering fixes are real on paper but unproven in the data.
- The money risk is collisions: the body is made of huge aluminum panels, minor accidents have produced $40k to $54k repair quotes, and insurance prices that in at about $3,400 a year, roughly $1,400 more than the XC90.
- Resale: the worst of the three, roughly $28k after five years.
- Company risk: Rivian still loses money on every vehicle it builds, though it's well capitalized (over $5B in cash plus a big VW partnership). Analysts put bankruptcy odds around 5%. That risk is part of why the used values are so low, so some of it is already priced in.
- Warranty: actually the best factory battery and drivetrain coverage of the group (8 years).
Side by side
| XC90 T8 2026 | EX90 2026 | R1S (if it fits) | |
|---|---|---|---|
| Powertrain | Plug-in hybrid, 455 hp, 32 mi electric | Full EV, 449 hp, ~305 mi | Full EV, 533 hp, 329 mi |
| Realistic price | $79–85k | $80–83k buy; lease is the deal | $84–90k (near sticker) |
| Garage fit | Identical to yours | Fits, 2.6 in to spare | 0.1 in on paper; fit test pending |
| Driveway | Clears comfortably | Clears with margin | Solves it outright |
| Reliability risk | Software glitches | 2025 charger failures; clean 2026 so far | Worst-rated brand, happiest owners |
| Insurance | ~$2,000/yr | ~$2,650/yr | ~$3,400/yr |
| Home charger needed | No | Yes, $1–2k | Yes, $1–2k |
| CarPlay | Wireless, standard | Wireless, standard | Never |
| Worth in 5 years | ~$38k | ~$30k | ~$28k |
Hybrid or full EV: what actually changes day to day
The only daily-life difference is fueling. Your T8 charges on the regular garage outlet because its battery is small. A full EV's battery is about ten times bigger, and that same outlet adds only 3 to 4 miles of range per hour. You average about 34 miles a day, so it would technically keep up, but recovering from a road trip or a near-empty battery would take days.
So any full EV means adding a 240-volt circuit in the garage, the same kind a dryer uses. Planning number: $1,000 to $2,000 all-in, unless the electrical panel needs an upgrade, which is the one unknown. A photo of the panel gets an electrician to a real quote. It's a one-time house upgrade that carries to every future EV, so it shouldn't sway the hybrid-vs-electric decision.
What each car actually costs per year
Everything in one place: depreciation, insurance, fuel and electricity, maintenance and tires, and the charger install, at your real mileage (about 12,400 miles a year). Assumptions and sources are in Appendix B; the two that matter most are PG&E's EV overnight rate ($0.23/kWh) and no solar, both to confirm.
| Per year | XC90 T8 | EX90 | R1S |
|---|---|---|---|
| Depreciation | $8,800 | $10,300 | $11,700 |
| Insurance | $2,000 | $2,650 | $3,400 |
| Fuel + electricity | $2,000 | $1,300 | $1,400 |
| Maintenance + tires | $1,100 | $750 | $1,400 |
| Charger install (spread over 5 yrs) | 0 | $300 | $300 |
| Total per year | ~$13,900 | ~$15,300 | ~$18,200 |
| 5-year total | ~$70k | ~$76k | ~$91k |
What the table says:
- The EX90's fuel savings are real but get eaten by its insurance. It saves about $700 a year on energy and costs about $650 a year more to insure, so the two Volvos stay roughly $1,400 a year apart, almost all of it depreciation. Going full electric costs about $7k over five years.
- The R1S costs about $4,300 a year more than the T8, and every line contributes: it costs more to buy, insure, shoe (tires are its known weak point), and it depreciates fastest.
- Solar would change this, and we don't have it. We'd like to add it eventually, but it's not on the immediate horizon, so the table stands. If solar lands partway through ownership, the EV energy lines roughly halve from that point on, which quietly favors the EVs on a long hold.
- The T8's number assumes plugging in every night. Driven as a pure gas car it burns about $2,700 a year in premium instead of $2,000 blended.
How long we keep it changes the math
The five-year table above is the worst case for the EVs, because years one through five are where they depreciate hardest. Held ten years, the way we've held your current car, the two Volvos come out essentially even all-in: the EX90's fuel savings, given ten years to compound, pay back its early depreciation. The R1S stays roughly $17k more either way.
So the second question after gas-or-no-gas: is this another ten-year car? If yes, the Volvo choice is genuinely free on money and comes down to the test drives. If there's real doubt, the T8 is clearly cheapest, and the EX90 should be leased rather than bought.
The warranty plan
Both Volvos come with a short 4-year/50,000-mile warranty, and the expensive failure-prone parts (the rear electric motor, air suspension, and the onboard charger) live under that short warranty, not under the longer 8-year battery warranty. The plan:
- Say no to everything in the dealer's finance office. Volvo's factory extended plan (VIP Platinum) can be bought any time while the factory warranty is active, from any dealer, and online Volvo dealers sell the identical contract for $1,000 to $2,000 less than the finance office quotes.
- VIP Platinum runs about $4k for 10 years of coverage and covers exactly the parts listed above. The decision can wait years.
- Skip third-party warranties. Most exclude the high-voltage parts, and Consumer Reports' data shows the average buyer loses money on them.
- If the Rivian wins: its extended plan (Rivian Care, 10 years/unlimited miles) exists but runs 2 to 4 times the cost of Volvo's.
Deal timing
Current offers expire August 31, but September deals on leftover 2026s should hold or improve as 2027s land, so there's no urgency spike. The federal $7,500 EV credit ended in September 2025; all the money is in manufacturer programs now, and Volvo's is concentrated on the EX90 lease. Re-pull the numbers before any dealer visit.
What's coming that we'd miss
Checked 2026-08-23, since buying now means passing on whatever's next:
- XC90: the all-new generation around late 2028 is reported to be an extended-range EV: roughly 100 miles of electric range with a gas engine that only works as a generator. At 34 miles a day that car would be effectively electric with zero road-trip compromise, so it's the one future vehicle that directly answers our gas-or-no-gas question. It's also a first model year on a new platform, which our buying rules skip, making it realistically a 2030 purchase. Holding your 2016 that long means three more years of $2,500-a-year decay plus the motor-failure risk this whole project exists to get ahead of. Not worth waiting for, but it does suggest the car after this one might be a Volvo again.
- EX90: nothing big announced or credibly rumored for 2027. The major upgrade (the 800-volt system with much faster charging, plus the software stabilization) already landed on the 2026 we'd buy.
- R1S: the 2027s are already out (launched August 2026) and modest: trim renames, efficiency software tweaks, and one real feature, second-row captain's chairs for a 6-seat layout. If captain's chairs appeal to you, that's 2027-only; for hauling the kids plus friends, the 7-seat bench is probably the point anyway. Rivian's attention is on the cheaper R2 launch, so no near-term R1 generation change is expected.
The takeaway: nothing coming argues for waiting inside our 12-month window.
Selling your car through Carvana
The likely path for the XC90 (probable plan as of 8/23). One California fact makes this easier: CA gives no sales-tax credit for trade-ins, tax is charged on the new car's full price either way, so trading in at the dealer buys nothing except convenience. Selling to Carvana usually beats the trade-in number, and a private sale would net $2k to $3k more in exchange for real hassle.
How it works:
- Get the offer online. Enter the plate or VIN and answer the condition questions; the offer comes back in minutes and is good for 7 days. Offers move week to week, so run it when we're actually ready and again a few days later if the first number disappoints.
- Be blunt in the condition questions. They re-inspect at pickup and cut the offer on the spot for anything undisclosed: dents, scratches, warning lights, worn tires. A disclosed ding is priced in; a discovered one costs more.
- Schedule pickup at the house. A rep comes to us, verifies the car matches what we entered, checks VIN and mileage, and the handoff takes about 20 minutes.
- Payment is a check at handoff or direct deposit that lands within a day or two.
What to have ready:
- The title, with every owner on it present with a driver's license
- Current registration
- Both key fobs and the charging cable set
- Plates stay with the car in California, and no smog check is needed when selling to a dealer
- After handoff, file the DMV Notice of Transfer and Release of Liability online (takes five minutes and ends our responsibility for tickets or accidents from that moment)
Two sequencing notes:
- It's your daily car, so the pickup gets scheduled after the new one is home. With a 7-day offer window and pickup available within days, the gap is small.
- Run CarMax and one dealer trade number in the same week for leverage; Carvana isn't always highest, and all the offers run on similar 7-day clocks.
Expected number: the 8/1 KBB ballpark says trade value of $10.7k to $13.1k, and instant offers usually land between trade and private-party value. At ~120k miles, expect the low end.
What happens next
- Park the friend's R1S in the garage and note how much room is actually left
- Test drive the 2026 XC90 T8 and EX90 back to back; ask the EX90 dealer whether the redesigned electrical system eliminates the 2025 onboard-charger failures
- Photo of the electrical panel (main breaker amps plus free slots) to price the charger install
- Try Off Road mode on the driveway in your XC90
- For margin, not gating: measure garage depth, width at the parking spot, and the door opening
- Check which PG&E rate plan the house is on today; any EV purchase should probably come with a switch to EV2-A (utility and no-solar confirmed 8/23)
- Research and compare the candidates (done 2026-08-15; resale research added 2026-08-18; all-in cost table added 2026-08-23)
Open questions
- Gas or no gas? The Volvos are close enough on price that this is most of the decision between them, and it's yours to make after the test drives.
- Is this another ten-year car? Decides whether resale matters and whether the EX90 should be leased instead of bought.
- Does the R1S actually fit?
- How to sell yours: current lean is Carvana (section above). Private sale would net $2–3k more for real hassle; trading in buys nothing in California.
- Budget check: we paid about $85k for your car in 2016. All three land inside that.
The research detail behind everything above:
Appendix A: dimensions and driveway data
The spec-sheet trap. Volvo publishes width with mirrors folded; Kia, Hyundai, Lexus, and Toyota publish body width without mirrors. Compared directly, that makes the Volvo look wider than it is and every rival narrower. Everything below is normalized to mirrors-folded (how it sits parked) and mirrors-out (how it fits through the door), from factory spec sheets.
The baseline: her 2016 XC90 T8 (factory spec)
| Measure | in | mm |
|---|---|---|
| Length | 194.9 | 4950 |
| Body width (no mirrors) | 75.7 | 1923 |
| Width, mirrors folded | 79.1 | 2008 |
| Width, mirrors out | 84.3 | 2140 |
| Height (incl. shark fin) | 69.9 | 1776 |
| Ceiling needed to open the tailgate | 84.5 | 2146 |
| Wheelbase | 117.5 | 2984 |
| Ground clearance (standard) | 9.4 | 238 |
| Ground clearance (air suspension range) | 8.9 to 10.5 | 227 to 267 |
Constraint priority (set 2026-08-14): length first (about 6 in of slack, ceiling roughly 200.9 in), driveway second (later resolved: filters nobody), width third, height not a concern.
All candidates vs. the baseline (deltas vs. 2016 XC90; parked width = mirrors folded)
| Car | Length | Parked width | Mirrors out | Height | Tailgate open |
|---|---|---|---|---|---|
| XC90 2016 (hers) | 194.9 | 79.1 | 84.3 | 69.9 | 84.5 |
| XC90 2026 | 195.0 (+0.1) | 79.1 (0) | 84.3 (0) | 69.7 (-0.2) | 84.5 (0) |
| Volvo EX90 | 198.3 (+3.4) | 80.3 (+1.2) | 83.2 (-1.1) | 68.8 (-1.1) | 85.7 (+1.2) |
| Kia EV9 | 197.4 (+2.5) | 80.6 (+1.5) | 89.1 (+4.8) | 70.1 (+0.2) | n/v |
| Hyundai Ioniq 9 | 199.2 (+4.3) | n/v (body 78.0) | n/v | 70.5 (+0.6) | n/v |
| Rivian R1S | 200.8 (+5.9) | 81.8 (+2.7) | 88.4 (+4.1) | ~73.3 std / 71.5 low | n/v |
| Lexus TX 550h+ | 203.1 (+8.2) | n/v (body 78.3) | n/v | 70.1 (+0.2) | n/v |
| Toyota Grand Highlander Hyb. | 201.4 (+6.5) | n/v (body 78.3) | n/v | 72.1 (+2.2) | n/v |
Length filter (~200.9 in ceiling)
| Car | Length | Slack left |
|---|---|---|
| XC90 2026 | 195.0 | 5.9 in ✅ |
| Kia EV9 | 197.4 | 3.5 in ✅ |
| Volvo EX90 | 198.3 | 2.6 in ✅ |
| Hyundai Ioniq 9 | 199.2 | 1.7 in ✅ |
| Rivian R1S | 200.8 | 0.1 in ⚠️ (fit test pending) |
| Toyota Grand Highlander Hyb. | 201.4 | over ❌ |
| Lexus TX 550h+ | 203.1 | over ❌ |
Driveway: how the analysis evolved
The first pass measured everything against the 2016 XC90 on approach, breakover, and departure angles plus ground clearance, and concluded every candidate was worse, with the EX90 looking disqualified at -5.0° of approach. Two corrections rebuilt it (2026-08-14):
- The Macan is the real floor. Joe's 2019 Macan (8.1 in clearance, 16.9° approach, steel springs) uses the same driveway and only occasionally scrapes. The XC90 is the comfortable end of the range; the Macan is the bar.
- The contact point is the bottom transition, where the street meets the ramp. That makes approach angle the only relevant spec; ground clearance, breakover, and wheelbase drop out.
Result, ranked on approach angle vs. the Macan floor (16.9°):
| Car | Approach | vs. Macan |
|---|---|---|
| Volvo EX90 | 18.8° | +1.9 ✅ |
| Kia EV9 | 20.0° | +3.1 ✅ |
| Hyundai Ioniq 9 | 20.1° | +3.2 ✅ |
| XC90 2026 | 21.0° | +4.1 ✅ |
| XC90 2016 (hers) | 23.8° | +6.9 ✅ |
| Rivian R1S | 35.6° | +18.7 ✅ |
Every candidate clears better than the Macan does today, so the driveway filters nobody. Two earlier conclusions were withdrawn on 2026-08-14: "the EX90 is the casualty that matters" (it clears, with air suspension to 9.8 in in reserve) and "air suspension is a requirement" (back to a preference).
Full three-angle data from the original pass, kept for reference:
| Car | Approach | Breakover | Departure | Ground clearance | Wheelbase |
|---|---|---|---|---|---|
| Macan 2019 base (the floor) | 16.9° | ~16.6° est. | 8.1 in | 110.5 | |
| XC90 2016 (hers) | 23.8° | 21.3° | 23.3° | 9.4 in std / 10.5 air max | 117.5 |
| XC90 2026 | 21.0° | 20.9° | 23.1° | 8.1 to 8.5 in* | 117.5 |
| Volvo EX90 | 18.8° | 17.7° | 19.9° | 8.5 in / 9.8 air max | 117.5 |
| Kia EV9 | 20.0° | n/v (~16–17° est.) | 22.8° | 7.8 in (no air option) | 122.0 |
| Hyundai Ioniq 9 | 20.1° | n/v (~14–15° est.) | 22.9° | 6.9 in (no air option) | 123.2 |
| Rivian R1S | 35.6° | 29.7° | 34.4° | up to 14.9 in | 121.1 |
* The 2025/26 XC90 clearance is quoted at curb weight + 1 person while the 2016 is at curb weight, so those two numbers aren't directly comparable; the angles are. Published breakover figures run 1 to 3 degrees better than the geometric estimate since the belly low point is rarely at the wheelbase midpoint; estimates adjusted accordingly.
Appendix B: resale research detail (2026-08-18)
Numbers below are labeled observed (actual listings or transactions) or forecast (residual models). Percentages from iSeeCars and CarEdge are measured against original MSRP; dollar losses are computed against realistic negotiated prices.
Per-vehicle summary
| XC90 T8 2026 | EX90 | Rivian R1S | |
|---|---|---|---|
| Realistic negotiated price | ~$82,000 | ~$81,500 | ~$87,000 |
| 5-yr retained (% of MSRP) | 42.5% (observed) | ~35% (segment proxy) | 32.3% (observed) |
| Value at year 5 | ~$37,800 | ~$30,000 | ~$28,400 |
| 5-yr depreciation cost | ~$44,200 | ~$51,500 | ~$58,600 |
| Value at year 8 | ~$21,900 | ~$18,900 (proxy) | ~$17,000 |
| Value at year 10 | ~$14,400 | ~$12,000 (proxy) | ~$13,800 |
| Data quality | Real 5-yr history | 18 months only | Real 4-yr history |
XC90 T8
- Observed (iSeeCars, 2026): XC90 PHEV depreciation 41.2% at 3 years, 57.5% at 5, 71.2% at 7, 83.8% at 10. The non-plug-in XC90 is better at every mark (34.4 / 54.2 / 67.5 / 79.5), so the T8 badge costs about three points at year five. The T8 wins the three-way comparison because the EVs depreciate worse, not because plug-in hybrids hold value well (hybrids as a category lose 35.4% at five years; the T8 loses 57.5%).
- Observed cross-check: a 2021 XC90 Recharge (~$68,500 blended MSRP) lists at $27,800–28,700 today, a 55% five-year loss. KBB (56–60%), iSeeCars (57.5%), and CarEdge (50.3%) all land in the same band.
- Observed direction: used 2023 XC90s average $42,119 across 2,215 listings, up 3.5% year over year (CarGurus, Aug 2026). The used XC90 market is stable to firming while the EX90's is falling.
- Forecast: no Volvo won a 2025 or 2026 ALG / J.D. Power residual award, so there's no third-party backing for a "Volvo holds its value" claim.
EX90
- Observed: 738 used 2025 EX90s average $54,182 (range $41,894–95,140), down 18.2% year over year (CarGurus, Aug 2026). Against a ~$82k average original sticker that's ~66% retained at 15–18 months. KBB has the 2025 Twin Motor Plus down $31,490 (38%) in one year.
- Used 2026 listings average $87,231, but that pool is mostly untitled dealer inventory, so it isn't a residual signal yet.
- The lidar timeline: Volvo ended the Luminar supply agreement 2025-11-14; Luminar filed Chapter 11 on 2025-12-15; a 2026-06-12 owner notice confirmed the promised features will never arrive; US owners were offered ~$1,500 in optional benefits; a class action was filed 2026-07-31 in LA County. A feature set advertised, sold, and withdrawn from cars still under warranty is the kind of story that sits on a nameplate's used values for years. The generation, not just model year one, appears to carry the penalty.
- No observed 3- or 5-year residual exists (oldest cars are 18 months old). CarEdge's forecast curve is internally inconsistent and was discarded; the 5-year figure above uses the luxury electric large SUV segment observed rate as a proxy.
R1S
- Observed (iSeeCars): 67.7% depreciation at five years (32.3% retained), vs. 65.2% for luxury electric large SUVs and 41.5% for all vehicles. The R1T truck holds up much better (54.5%), a truck-segment effect.
- Observed: a 2022 R1S First Edition ($91,800 MSRP) averages ~$52,730 in current listings, a 42.6% loss in four years. Named causes: Gen-1 quality issues, Rivian's own 2023–24 price cuts, Gen 2 making Gen 1 look dated, the credit expiration, and a 36% sales drop in 2025.
- A forum trade-in comparison (indicative only) showed a 2023 Quad Large offered $67,850 vs. a 2025 Dual Large offered $69,500, a $1,650 gap. If Gen 2's fixes were being priced in, the gap should be wider.
- Company financials (Q2 2026): revenue $1.658B up 27% YoY; automotive gross profit negative $36M, carried by software/services ($215M at 42%); net loss ~$835M; $5.3B cash plus $7.16B pro forma after a July raise; VW JV committed up to $5.8B through 2027; DOE loan renegotiated down to $4.5B with Georgia pushed to late 2028. AlphaSpread's credit model: 4.6% bankruptcy probability. The Fisker precedent shows the tail (Ocean Ones at under a third of MSRP in two years), but Rivian is far better capitalized, and the R1S's current used values already assume a lot of bad news.
- Collision cost shows up in insurance rather than any published total-loss rate: the Feb 2026 $53,736 quote (~250 labor hours), an $11,000 bumper tap, a $42,000 rear-end bill. The certified repair network covers ~25% of US zip codes vs. Tesla's 40%, which drives the ~$3,400/yr insurance figure.
The EV segment backdrop
iSeeCars 2026 study (950,000 five-year-old vehicles, inflation-adjusted):
| Segment | 2025 study | 2026 study |
|---|---|---|
| Trucks | n/v | 34.2% |
| Hybrids | 40.7% | 35.4% |
| All vehicles | 45.6% | 41.8% |
| EVs | 58.8% | 57.2% |
Every segment improved; EVs least. The EV-to-hybrid depreciation gap widened from 18.1 to 21.8 points. Used-EV price momentum (Cox, Recurrent) is positive, but that strength is concentrated in cars under $20k and helps whoever buys the car from us, not us; for a new purchase the cumulative-loss number is the relevant one.
The federal credits are all dead (new $7,500 and lease loophole ended 2025-09-30; home charger credit ended 2026-06-30), and 243,000–330,000 EV leases expire during 2026, triple 2025's volume. That lease-return wave lands on used EV values through 2027, exactly the window where a new EX90 or R1S takes its steepest losses.
Lease vs. buy detail
- EX90 (CarsDirect, offers through 2026-08-31): 2026 Twin Plus at $849/mo, 36 months, $8,369 due at signing, 55% assumed residual, $7,500 lease allowance. Effective ~$1,081/mo, ~$38,900 total. Observed 2025 EX90s are already at 56–60% retained at 15–18 months, and wholesale runs 10–15% under retail, so for Volvo's 55% residual to hold at 36 months the car would need to stop depreciating. Volvo Car Financial is holding the residual above the market and eating the difference, which caps our downside at their expense.
- R1S: CarsDirect flags both current leases as overpriced ($1,159/mo and $1,599/mo), no manufacturer lease cash. Rivian prices its leases near the real depreciation risk, so an R1S lease buys no protection.
- XC90 T8: ordinary incentives ($2,000 purchase allowance or 2.99%/72 + $2k bonus; $3,750 lease allowance + $2,000 loyalty, which we qualify for). Best residual of the group, so no structural reason to lease it.
Battery worry, mostly discharged
- Volvo: 8yr/100k battery warranty, guarantees 70% capacity, transfers with the VIN. Rivian: 8yr/150k battery (Dual Motor), 8yr/175k drivetrain, transferable.
- Recurrent (1B+ miles): average EV retains ~97% of range at 3 years, ~95% at 5; battery replacement rate 0.3% for MY2022+. Geotab: 2.3%/yr average degradation, 81.6% retained at 8 years. For a home-charged car doing ~34 mi/day, degradation shouldn't meaningfully touch resale.
The all-in annual cost math (2026-08-23)
Inputs behind the per-year table in the main body. Mileage: 34 mi/day from the Mercury policy, 12,400 mi/yr.
Energy prices. PG&E EV2-A off-peak (12 a.m. to 3 p.m.): $0.23/kWh, from PG&E's EV rate page, August 2026. California regular gas averaged $5.58/gal mid-August 2026; the T8 recommends premium, assumed at $5.95. Road trips assumed 10% of miles at DC fast-charging prices ($0.50/kWh). Confirmed 2026-08-23: PG&E is the utility, no solar, and none planned near-term (wanted eventually). Still to check: which rate plan the house is on today, and switching to EV2-A with any EV purchase. Future solar cuts the EV energy lines roughly in half; every 5 cents/kWh moves an EV's annual cost by about $150 to $190.
Consumption. EX90 Twin: ~85 MPGe EPA combined, about 0.40 kWh/mi from the wall, ~$1,300/yr blended. R1S Dual Large: 78 MPGe, 0.43 kWh/mi, ~$1,400/yr. XC90 T8: 32 mi EPA electric range against a 34 mi/day average, modeled as 60% electric miles (at ~0.55 kWh/mi from the wall, 120V charging losses included) and 40% gas at 27 mpg, ~$2,000/yr; as a pure gas car it would run ~$2,700/yr.
Insurance. T8 ~$2,000/yr (carried from the 8/15 pass). EX90 ~$2,650/yr (Recharged and CarEdge national full-coverage estimates; no real quote yet). R1S ~$3,400/yr (carried from 8/15).
Maintenance + tires. T8: RepairPal ~$851/yr plus a tire set every ~40k, call $1,100/yr, partly offset by Volvo's 2yr/20k complimentary maintenance. EX90: light routine service plus heavier EV tire wear, ~$750/yr estimated (no fleet data exists yet). R1S: routine service is cheap ($300–500/yr per owner guides; KBB $623/yr with wear items) but tire wear is the known weak point, with owner reports ranging from 6k to 25k miles a set at ~$2,000 installed; modeled at ~$1,400/yr combined.
Depreciation is the 5-year table earlier in this appendix divided by five. Charger install: $1,500 midpoint spread over five years for the EVs.
Omitted as noise: registration differences (~$30/yr between the cheapest and priciest), the VIP Platinum warranty purchase (same ~$4k on either Volvo path), and financing costs (equal across cars at equal price; the R1S's higher price adds interest the table already reflects as principal).
The ten-year view. Extending depreciation to the year-10 values, energy and insurance at flat rates, and maintenance rising in the out-years: the two Volvos land within about $2k of each other all-in (~$120k each), because ten years of fuel savings pays back the EX90's early depreciation. The R1S lands ~$17k higher. This is rougher math than the five-year table (no one has ten-year data on any of these) but the direction is stable.
The trade math on the 2016
KBB through 2026-07-10: T8 trade-in $10,730–13,080, private party $13,420–16,270, low end likely at ~120k miles. Roughly 18–22% retained against the ~$73k original MSRP at ten years. The visible decay rate is ~17.4% of remaining value per year, so waiting costs roughly $2,000–2,500 a year. Set against a >$10k rear-motor failure that exceeds the entire trade value, that's the case for buying inside the next 12 months without defending any specific month. Ground truth when we're serious: run a Carvana/CarMax instant offer.
Appendix C: research on rejected candidates
Kia EV9 (rejected on brand, 2026-08-15). Pricing at rejection: Land AWD $70,545 sticker, transacting ~$59–62k on up to $10k factory cash; 0%/60 + $5k as the APR alternative; strong leases (Wind AWD $449/mo). Reliability: bottom-tier CR prediction (below the R1S); ICCU charging-module failures hit 2–10% of Hyundai/Kia EV owners vs. ≤1% for other EVs; 9 recalls including two loss-of-drive campaigns; repairs free under the 10yr/100k warranty (original owner) but dealer parts waits ran 8 days to 4 months. Depreciation ~42% in 2 years on a 2024 Land. Insurance ~$2,650/yr.
Hyundai Ioniq 9 (rejected on brand, 2026-08-15). Was already tripping the first-year rule and had the weakest driveway numbers (6.9 in clearance, though its 20.1° approach cleared the Macan floor). Width never verified.
Lexus TX 550h+ / Toyota Grand Highlander Hybrid (rejected on length, 2026-08-14). 203.1 and 201.4 in, both over the ~200.9 ceiling. Not researched further.
Audi (checked and rejected, 2026-08-15). Q7 is the last year of a 2017-era generation with no US PHEV/EV and a tight third row; Q9 is a first model year, gas-first, and ~204+ in. No 3-row Audi EV exists. Future watch: a ~2028 Q7 PHEV in year 2 of its generation would be a legitimate candidate, outside this project's timeline.
Rivian R1S, briefly. Ruled out on 2026-08-15 (CarPlay plus the 0.1 in of length slack), reinstated the same day: CarPlay is Joe's preference rather than Brandi's, and the length question deserves real data over spec-sheet math.
Appendix D: research log
2026-08-01. Project started. Requirements captured; buying posture and model-cycle rules set; initial candidate list named. KBB ballpark on the 2016 T8: trade ~$10.7–13.1k, private ~$13.4–16.3k.
2026-08-14. Factory dimensions pulled and normalized (found the spec-sheet trap: Volvo publishes mirrors-folded width, others body-only). Length cut the Lexus TX and Grand Highlander. First driveway pass wrongly disqualified the EX90 and promoted air suspension to a requirement; both conclusions withdrawn after the Macan-floor and bottom-transition corrections. Confirmed the 2016's air suspension behavior is normal; found the unused Off Road mode (+1.6 in below 25 mph).
2026-08-15. Charging: full EVs need a 240V install, $1–2k planning number, panel capacity the only unknown. Pricing pass: EV9 ~$59–62k, EX90 money on the lease side, XC90 T8 in run-out discounting, R1S near sticker. Reliability pass: no safe harbor anywhere (T8 = software annoyances, EX90 = 2025 charger failures with a clean-so-far 2026, EV9 = worst predicted reliability, R1S = last-place brand with the happiest owners). Corrected an earlier claim that the 2026 EX90 has working lidar; it has none. Warranty pass: VIP Platinum covers the named failure modes, buyable any time, cheaper online. CarPlay ruled the R1S out and got walked back the same day. Audi checked and rejected. Kia and Hyundai out on brand. Field closed: XC90 T8 vs. EX90, R1S conditional on the fit test. Doc restructured from append-order into case order.
2026-08-18. Resale and depreciation pass, run from Joe's voice capture asking how the three finalists hold value. The finding that changed something: the Volvos cost the same to buy but diverge $7–8k over five years on resale, so full-EV isn't free; over a ten-year hold all three converge and resale stops mattering. Second finding: Volvo's subsidized EX90 lease transfers the depreciation risk to Volvo; Rivian offers no equivalent.
2026-08-23. Doc rewritten for readability ahead of sharing with Brandi: decision framing up front, plain-language car sections, research detail moved to appendices. No facts changed. Same day, added the all-in annual cost table (depreciation + insurance + energy + maintenance + charger), closing the gap that the doc had never priced fuel and electricity. Findings: the EX90's ~$700/yr fuel savings are canceled by ~$650/yr higher insurance, so the Volvo gap holds at ~$7k over five years; over a ten-year hold the Volvos converge to roughly even and the R1S stays ~$17k more. EX90 insurance firmed to ~$2,650/yr (national estimates). PG&E confirmed as the utility, no solar (wanted eventually, not near-term); remaining input is the house's current rate plan. Also added the Carvana selling process for the 2016 (probable disposal path per Joe): 7-day offers, at-home pickup, re-inspection risk on undisclosed condition, no CA trade-in tax credit so trading in buys nothing. Also added a future-models check: the ~2028 XC90 redesign is reportedly an extended-range EV (~100 mi electric plus a gas generator), tempting but a first-year 2030 buy in practice; no big EX90 changes rumored for 2027; 2027 R1S is modest (captain's chairs the one real addition). Nothing argues for waiting.