DGActual Market Brief — Week of September 8, 2026

Happy Tuesday. Labor Day bumped us a day, and the market did not take the weekend off: the model moved up again, Ford posted an ugly August, Honda quietly hit a million U.S. sales in eight months, Volkswagen approved another 50,000 job cuts, and Jaguar Land Rover is cutting 4,000. Closer to home, used retail prices fell for a ninth straight week, the FSBO board flipped one big signal, and we finished a full refresh of Google ratings for 17,000 dealerships, which produced the number I cannot stop thinking about: America's dealers are collecting 11,500 Google reviews a day. Every day. Including Sunday.

Overview: a strong August that nobody felt

New vehicles sold in August at a pace that would add up to 16.8 million a year if it held, according to NADA. That is a healthy number, and our own forecast model caught up to it this morning, moving from 16.38 to 16.49 million for the third quarter. Yet the brand-level August reports read like a different market. Ford sales fell 10.3 percent to 170,681 vehicles, with its electrified lineup down 41.5 percent. Toyota, Hyundai, and Mazda were also down. Honda, Kia, and Subaru were up, and Honda crossed one million U.S. sales for the first eight months of the year on the strength of hybrids.

How can the industry have a good month while half the brands have a bad one? Mix. Hybrids surged, and incentives did the heavy lifting: the average new-car discount hit $3,451 in July, up 8.1 percent, per NADA. The customers showed up. They just bought different cars, from different stores, with more of the factory's money on the hood. For your store, the traffic is real and the margin is thinner than the headline.

Sales and demand: the model ticks up, and we can explain about half of it

The DGActual model's third-quarter base rose 0.11 million to 16.49 this morning. About half of that move is simple: the model leans on the most recent sales pace it can see, and August's strong print (17.19 million on the broad measure that includes heavy trucks) just entered that window next to July. The other half comes from the model switching from a finished-quarter base to a quarter-in-progress base as September fills in. We owe you the exact line-by-line breakdown and it will be in Monday's SAAR brief. We learned in August what happens when we guess at attribution, so we are not doing that.

For readers new to this: SAAR is the seasonally adjusted annual rate, the industry's way of saying 'if this month's pace held for a year, we would sell this many.' Our model forecasts it from how consumers feel, how well they are paying their loans, unemployment, and last quarter's pace. Two of those inputs came back to life this month: the Fed's second-quarter loan delinquency read (2.62 percent, slightly better) and consumer sentiment (55.2, up from the high 40s). All summer, sentiment had been stuck on a placeholder value while we waited for the data. It is real now, and it is part of why the model has climbed two weeks running.

Credit is still the quiet drag. The Fed's own August notes describe auto sales softening and loan demand mixed, and Chair Warsh said late last month that rate hikes may still be needed. Translation for the desk: no rate cut is coming to rescue the monthly payment. Whatever payment wall you are hitting today is the wall for the rest of the year.

Labor and production: Volkswagen, JLR, and a GM win in Canada

Volkswagen's supervisory board approved a plan for roughly 50,000 additional global job cuts on top of about 50,000 already under way, with four German plants under review and the lineup potentially shrinking by half by 2035. One of those plants, Osnabruck, is being sold for defense production, which preserves about 1,400 jobs but takes auto capacity off the board. Jaguar Land Rover opened a buyout program aimed at roughly 4,000 jobs over two years and 1.7 billion pounds in savings, blaming tariffs, falling sales, and Chinese competition.

On the other side of the ledger, Unifor members ratified a GM deal that secures more than C$1.1 billion for Ontario plants, including next-generation heavy-duty Sierra production. Stellantis halted Fiat 500 production at Mirafiori again, September 2 through 4, on an engine-component shortage, one week after restarting. And Honda is asking its suppliers to cut prices roughly 30 percent as part of a $9.4 billion savings push through 2030. When an automaker squeezes suppliers that hard, it shows up at your parts counter eventually, one way or another.

China and EV makers: BYD goes global, VinFast retreats

BYD now makes more money outside China than inside it. Overseas revenue hit 53 percent of first-half sales, a first, even as first-half profit fell 21 percent. August sales rose 18 percent with exports setting a fifth straight record while China sales fell 16 percent, and the company is now targeting more than 2.5 million overseas sales in 2027. NIO's second-quarter revenue jumped 69 percent with losses narrowing. VinFast paused three India programs and asked suppliers to stop work.

Stateside, automakers urged Congress to make the ban on Chinese connected vehicles, hardware, and software permanent before year end, while Stellantis is reportedly exploring a Huawei and JAC partnership for Maserati. NHTSA opened an audit into Tesla's Cybercab after its Austin launch. Audi added a compact EV for Europe.

Safety desk

Two large Stellantis campaigns this week. Ram 1500, 2025 through 2026 model years: 239,131 trucks under campaign 26V560 for radio software that can block the rearview camera image, and this one carries do-not-drive guidance until updated. Jeep Grand Cherokee and Grand Cherokee L, 2021 through 2023: 328,381 vehicles under 26V562 for rear coil springs that may be incorrectly installed and detach while driving. Ford Mustang, 2024 through 2026: 148,663 under 26V547 for wiring-harness ground connections that may fracture. Mitsubishi Outlander and Outlander PHEV: 67,146 under 26V549 for rearview-camera display failures. Smaller campaigns cover Chrysler Pacifica and Voyager (2,017, stability-control risk), and single-digit-to-double-digit counts for Tesla Model Y and Ford Bronco Sport, F-150, and Maverick fastener issues.

If you carry Ram or Jeep, those two campaigns alone touch more than 567,000 vehicles, and one of them tells owners not to drive the truck. That is a service-lane opportunity and a customer-trust moment in the same phone call. Check your open-campaign list before the first appointment of the day.

DARPI: ninth straight week down, and a correction that is ours

DARPI is the DGActual Automotive Retail Price Index: used retail asking prices from dealer listings across 11 segments, tracked weekly and indexed to 100 at our June 8 baseline.

Week 16: retail fell 0.10 points to 96.45, the ninth consecutive weekly decline, though the smallest drop of the streak. Nine weeks is a trend. A 0.10 drop is a trend getting tired. Full Pickup led the way down at minus 0.35, then Electric Vehicle at minus 0.25 (its eighth straight decline) and Midsize SUV at minus 0.16. Minivan was the only segment to gain meaningfully, up 0.13. Versus the June baseline, Midsize SUV remains the weakest at minus 10.07, then Fullsize SUV at minus 8.04 and EV at minus 6.03. The compact segments are still the most resilient, each within 0.6 of where they started in June. Cheap cars are holding their value. Expensive SUVs are not. That is the whole used market in one sentence right now.

A correction, and it is ours. For the past two briefs we reported the retail-to-wholesale spread at record lows: negative 2.47 on August 24 and negative 2.76 on August 31. Those numbers were wrong. Our wholesale benchmark had been frozen at a mid-July checkpoint reading for six weeks instead of updating to the official monthly index, which made the gap look wider than it was. Restated against the official July wholesale reading (210.0, published August 7), the true spreads were negative 1.78 and negative 2.07. The direction was right: retail has been below wholesale every week since mid-July. The magnitude was overstated by about a third, and neither week was a record. The actual widest gap was August 2 at negative 2.54. We caught this while chasing a missing data point this week, fixed the pipeline so the benchmark now keys on publication dates rather than calendar months, and we are telling you here rather than quietly revising the chart. Where it stands now: as of this week's close the spread was negative 2.17 against the July official reading. The official August wholesale index posted this morning at 208.2, down 0.9 percent from July and up 0.4 percent from a year ago. Against that fresh reading, this week's retail sits at negative 1.33. Wholesale is finally falling faster than retail, which is the first sign in two months that the gap may be starting to close from the other side. Plain English for anyone new here: when retail asking prices sit below wholesale auction values, dealers are listing cars for less than it costs to replace them at auction. That is the squeeze this index has been describing all summer.

Segment-level model movers: in Midsize Car, the 2024 Nissan Altima leads volume at $20,054 across 4,993 dealer listings, with the 2024 Accord at $26,861 and the 2024 Camry at $27,143. In EV, the 2023 Model Y at $33,500 (2,098 listings), the 2023 Model 3 at $27,689, and the 2023 Mustang Mach-E at $30,991 anchor the segment. On near-new EVs overall, based on dealer listings, the median asking price is $31,799 at dealerships against $35,000 in FSBO listings, and dealer EVs are turning in about 28 days at the current listing.

FSBO acquisition desk: RAV4 flipped, hard

This section is about FSBO listings: for sale by owner, private sellers listing their own cars. The question is simple: for the same model, 2022 through 2025, at comparable mileage, is a private seller asking less than a dealer? When the answer is yes, that is a car you may be able to buy from a driveway for less than at auction.

The board this week, every spread mileage-matched on both sides.

Negative = the private seller (FSBO) is asking less than dealers for the same model at comparable mileage. All figures compare 2022 through 2025 model year vehicles in the 20,000 to 60,000 mile band on both sides. n = matched private listings.

Honda CR-V is now three weeks confirmed: private sellers ask $2,502 less than dealers on comparable miles. Tesla Model Y's private discount widened to $1,687, though the matched sample fell from 156 to 45 listings this week, so treat the size of the move with some care. Nissan Rogue holds at $910 under.

The story of the week is the Toyota RAV4, and it is a trap. On raw numbers, private RAV4s look $2,827 cheaper than dealer RAV4s, the biggest apparent bargain on the board. Match the mileage and it flips: private sellers are asking $2,855 MORE than dealers for comparable cars. The private RAV4s listed right now simply carry far more miles than dealer stock, and the low sticker is the mileage talking. That is a $5,700 swing between the lazy read and the honest one. If you bought off the raw number this week, you overpaid. This is exactly why we refuse to publish a private-seller comparison without mileage on both sides. F-150 and Grand Cherokee remain anchored well above dealer pricing. Patience files.

Google Reviews Report: 11,500 reviews a day

We completed a full refresh of Google ratings for every franchise dealership we track this weekend, 17,096 stores with 27.8 million reviews between them, and for the first time we can compare a full month against the prior one. The headline is not about ratings. It is about volume.

Across 17,750 dealers present in both months, review counts grew by 344,542 in 30 days. That is 11,485 new Google reviews per day, about 19 per store per month, or one every 7.5 seconds nationally. Meanwhile, 93 percent of dealers saw no change in their star rating at all: 718 moved up, 552 moved down, 16,480 stayed exactly where they were, and the national average sat at 4.48 both months. Here is the reframe: your Google star rating is a five-year reputation. Your review count is a monthly scoreboard. Stores obsess over the first number and ignore the second, and the second is the one that actually tells you who is winning right now.

Brands with 100 or more stores present in both the August and September pulls. Source: DGActual Dealer Reputation Scoreboard, reviews.dgactual.com.

A Toyota store adds 47 reviews a month. A Jaguar store adds fewer than two. Kia stores are out-collecting Honda stores per location, which is the surprise in this table, and Lexus is the only luxury brand in the top tier. Review velocity is two things multiplied together: how many customers walked through the door, and how systematically someone asked them. A store that trails its brand average on this number has a process problem, not a customer problem.

Brand rating moves, limited to brands with 100 or more stores: Porsche slipped 0.019 (275 stores), Mitsubishi and Lexus 0.014 each, Chrysler 0.012. Up: Buick 0.014 (497 stores), Dodge 0.011 (570), Cadillac 0.009 (570). These are hundredths of a point, real but small, and we are not going to dress them up as trends after one month.

The 5.0 problem, quantified: 442 dealers hold a perfect 5.0 rating. Their median review count is 3. Only 6 of them have more than 500 reviews. The actual elite, 4.9 or better on 3,000 or more reviews, is a group of 68 stores, led by Honda of Concord (4.9 on 19,364 reviews), Family Toyota of Arlington (16,734), Serra Toyota (15,391), and Lindsay Lexus of Alexandria (13,925). A perfect score on three reviews is a rounding error with a trophy. When you benchmark reputation, weight by volume or the 3-review stores will win every time.

Full segment data at data.dgactual.com. Dealer reputation across 17,096 stores at reviews.dgactual.com.

Sources: NADA Market Beat, August 2026. Federal Reserve (DRCLACBS, TOTALSA, Beige Book August 2026). University of Michigan Surveys of Consumers. Reuters (Volkswagen restructuring and Osnabruck sale, Jaguar Land Rover, GM Canada, BYD first-half results and August sales, Stellantis Mirafiori halt and Huawei-JAC talks, Honda supplier cost targets, Fed Chair Warsh remarks via AP). BBC News (JLR job cuts). CNBC and Ford Authority (Ford August sales). Automotive News (August brand results). Honda Newsroom (one million sales). NIO SEC Form 6-K. Detroit News (connected-vehicle ban). Motor1 (Cybercab audit). NHTSA recall filings 26V560, 26V562, 26V547, 26V549, 26V561, 26V558, 26V550. Cox Automotive Manheim Used Vehicle Value Index (July official 210.0 published August 7; August official 208.2 published September 8). DGActual Retail Price Index Week 16, SAAR model auto-update September 8, FSBO acquisition pull (MarketCheck API) September 6, and Dealer Reputation Scoreboard September 2026 refresh (Google Places). DGActual Market Intelligence. Week of September 8, 2026.