General Motors Company (GM) · Q3 2026

S1Our Overall Take

Retail Channel: NEUTRAL | Composite: NEUTRAL | Confidence: MEDIUM

This is a read on GM's U.S. vehicle-selling channel only, not a rating on the company. General Motors sells through four brands (Chevrolet, GMC, Buick and Cadillac) and our data follows those vehicles at U.S. dealerships: how many sell, at what price, how long they wait for a buyer, and how many sit on the lots. This report covers July through September 2026.

Prices held. The average transaction price of a new GM vehicle was $54,314, up 1.2% from a year ago and flat on the second quarter. The discount from the manufacturer's suggested retail price (MSRP, the "sticker") was 5.1%, against 4.8% a year ago and 5.4% in the second quarter: a third of a point either side, inside normal noise. New vehicles that sold took 72 days from first advertisement to sale, about three days faster than a year ago.

Stock built up. GM dealers' new inventory was 587,236 vehicles on September 30, 17.6% more than on June 30 and 9.5% more than a year earlier. At the selling rate of the quarter's last 45 days, that is a Market Days Supply of 94 days, against 73 three months earlier. The increase in inventory is concentrated in full-size pickups and sport-utility vehicles (SUVs): Silverado light-duty inventory rose by 21,549 vehicles and Tahoe inventory by 8,034. GM told investors in July it would keep truck production at record levels while it launches the next-generation Silverado and Sierra in December, so part of this is very likely deliberate stocking ahead of the changeover (our interpretation). It still leaves GMC holding more days of stock than its mainstream peers, where it held fewer three months earlier, and Buick further above them.

We do not lean on the volume count. Our count of new GM sales fell 11.7% from a year ago, in a captured market down 3.2%. GM's own count of U.S. deliveries, which includes fleet, fell 5.5% (management's own disclosure). Our coverage rate of GM's figure was 85.5% this quarter against 91.5% a year ago, and the models where the two disagree most (Colorado, Canyon, Suburban, Trax and the Buick brand) are the fleet-heavy ones. We show the decline and do not use it as evidence of demand.

Why the composite is Neutral, and which way it could be wrong. More than a third of GM's guided 2026 operating profit comes from businesses our data cannot see, and in July management guided almost all of them positively: most of the remaining warranty benefit was expected in the third quarter, electric-vehicle losses were improving, and full-year pricing was guided to the high end. Tariffs, commodities and onshoring costs push the other way. With a neutral dealer channel and a positively guided invisible majority, a neutral composite is more likely to understate GM's reported quarter than overstate it.

S2The Quarter in Numbers

Market share of captured new retail (North America, GM's four brands):

The year-over-year share drop is larger than the market's own movement can explain: our capture of the whole new market fell about 3%, while our capture of GM fell almost 12%. Most of the change is our measurement of GM moving, not GM losing ground on that scale (Section 3).

S3Inventory and Pricing Risk

GM's own normal range for new Market Days Supply, on the North American basis our benchmark uses, is 72–79 days (the middle half of the last 15 quarters); this quarter reads 95 on that basis, well above it and rising. GM reported 55 days at the end of June on its own basis, which counts stock on the lot rather than every advertised vehicle; our figure for the same date ran 132% of GM's, above it as expected on our basis. Brand used Market Days Supply, across all dealers carrying GM's brands, is at the top of its 15-quarter range; the franchise dealers' part of that stock is a dealer-health reading that sits outside the verdict (Section 5b). New-vehicle discounts are steady rather than widening, so the build has not yet shown up as price cutting (Section 6).

S4Electric Vehicles, Geography, and the Two Cases

Electric vehicles (EVs). New fully electric GM vehicles sold numbered 22,463 in our data, 3.9% of new sales against 9.9% a year ago, when buyers rushed to beat the September 30, 2025 end of the federal tax credit. GM's own count fell 61.7% (management's own disclosure); ours fell 65.1%. The cars that are selling are selling closer to sticker: the average electric-vehicle discount narrowed to 2.8% from 5.8% in the second quarter, and sold days on market fell to 116 from 142. Stock is rebuilding faster than sales, though: fully electric Market Days Supply was about 98 days against 79 in June.

Geography. Texas, GM's largest state in our data, fell 5.6%; Michigan (−20.7%), Florida (−19.0%) and California (−17.9%) fell more. Because our overall GM count is running below GM's own this quarter, we treat these as descriptions of our data rather than regional demand signals.

01Overall Verdict and Confidence

Retail Channel Signal: NEUTRAL | Confidence: MEDIUM. Built from MarketCheck data on U.S. dealer sell-through, pricing and inventory only. Two lenses move: inventory build risk reads Cautiously Bearish, and every other measurable lens reads Neutral. The volume lens is recorded as not observable this quarter, because our count of GM sales moved by six points against GM's own count between the two quarters being compared (Section 3).

Weighting. A single bearish lens does not carry the verdict. The stock build is real and large, but the price data shows no strain yet (average price up 1.2%, discount within a third of a point of both comparison quarters) and an aged count without price corroboration is a risk flag, not a verdict. It is also consistent with GM's stated plan to keep truck output at record levels through the December launch (our interpretation). Used-vehicle readings are reported in Section 5b and sit outside this verdict.

Composite Signal: NEUTRAL | Confidence: MEDIUM. The parts of GM our data cannot see were guided as follows on the July 21 call (management's own disclosure): warranty improvement of $1.0–1.5 billion for the year, with $0.5 billion realized in the first half and most of the rest expected in the third quarter; regulatory savings of $0.5–0.8 billion; electric-vehicle losses improving by $1.0–1.5 billion; GM Financial earnings before tax of $2.5–3.0 billion; software and services revenue above $3 billion; and full-year North American pricing up about 0.5%. Against them: gross tariff costs of about $0.9 billion a quarter, commodity, logistics and memory-chip inflation of $1.5–2.0 billion with more in the second half, and $1.0–1.5 billion of onshoring costs ramping into year-end. These segments are large enough to set the earnings direction on their own, so the composite cannot move past Neutral unless both tiers point the same way. They do not: the channel is Neutral. The cap is more likely to understate GM's quarter than overstate it, because the largest third-quarter items in the guidance were tailwinds.

The volume count is checked against GM's own delivery release before it is used, average transaction price leads the pricing lens, and wholesale guidance is not graded against retail data.

02Signal Summary Table

03Detailed Sales Numbers

One useful read on demand is how many vehicles leave the lots and at what price. Here the price side is clean and the count side needs care.

New vehicles

In our data, new GM retail sales were 573,867 vehicles and $30.8 billion of revenue, at an Average Transaction Price (ATP) of $54,314. Sales fell 3.5% from the second quarter and 11.7% from a year ago, while the captured U.S. new market fell 3.2% year over year. The monthly path ran 199,429 in July, 192,709 in August and 181,729 in September.

Checked against GM's own count. GM reported 670,974 U.S. deliveries for the quarter, down 5.5% (management's own disclosure), a figure that includes fleet sales to rental, commercial and government buyers. Our count was 85.5% of GM's this quarter and 91.5% of it a year ago, so the year-over-year comparison spans a six-point step in what our data captures. The disagreement sits in specific places:

Where the two counts agree (the light-duty pickups, Yukon, Escalade and electric vehicles) the retail and total pictures match. Where they disagree, the nameplates are those most exposed to fleet buyers, and GM reported its best first-half fleet result in more than five years (management's own disclosure). That points to fleet growth as at least part of the gap, but our data cannot separate it from a change in what we capture (our educated guess). For these nameplates we use price and days on market only, not our unit change.

What GM's count implies for gasoline vehicles. Taking GM's total and electric-vehicle counts together, its U.S. deliveries of gasoline and diesel vehicles were 645,501 against 643,846 a year ago, up 0.3%: flat, while the electric decline drove the whole of GM's reported 5.5% fall (derived from management's own disclosure). In our retail data, combustion-engine sales fell 5.8%.

Market share. In the North American share table (Part 1), GM's unit share fell 1.4 points year over year and 0.2 points quarter over quarter. The year-over-year move is mostly our own capture of GM moving: the captured market fell about 3%, our GM count about 12%. The sequential move is under the 0.3-point floor and is not read. By brand on a U.S. basis, Chevrolet's unit share was 9.8% (−0.7 points year over year, +0.1 points sequentially) and GMC's 3.7% (−0.4 and −0.3 points). Revenue share runs above unit share for GM (17.3% against 15.6%) because of its truck and SUV mix; that gap narrowed from 2.2 points a year ago to 1.7 points, as Cadillac's electric vehicles and GMC's heavy-duty pickups fell back.

Light-duty pickups, a one-off cut. Across the full-size light-duty pickup segment in our U.S. data (Silverado 1500, Sierra 1500, F-150, Ram 1500, Tundra, Titan and their electric versions), GM sold 145,353 vehicles, up 5.0% in a segment up 5.5%, for a segment share of 38.9% against 39.1% a year ago and 39.5% last quarter.

Used vehicles

Used sales and prices are read in Section 5b. GM does not earn revenue when a dealer resells a used GM vehicle; used figures matter to it through its lease book, the cost of selling new cars, trade-ins and dealer health.

04Top Models and Brands

Brands (new)

Chevrolet and GMC are 86% of GM's new revenue in our data, so they carry the read. Cadillac's 32.7% fall in our data mostly reflects its electric vehicles' rush a year ago (our interpretation); its gasoline Escalade fell 10.8% in our data (15.4% by GM's count), and Cadillac prices firmed, with the discount narrowing by 1.2 points and ATP up 8.6%.

Models (new, top 16 by units)

The two light-duty pickups are a quarter of GM's new revenue in our data, and they behaved differently. Silverado light-duty sales rose 15.4%. Its price sat 10.4% below sticker, narrower than 10.9% below sticker last quarter but deeper than 9.8% below sticker a year ago. Sierra 1500 sales were flat on a year ago and fell 11.0% from a strong second quarter. Both trucks are in the last months of their current generation; a deep sticker discount at that stage is what end-of-cycle trucks usually show, and GM said in July it had not seen the typical heavy end-of-cycle discounting (management's own disclosure). Our discount data does not measure GM's incentive spending, so it neither confirms nor contradicts that (Section 6).

Used

Used sales and prices by model are read in Section 5b. GM does not earn revenue on these resales; they matter to it through its lease book, the cost of selling new cars, trade-ins and dealer health.

05What's Sitting on the Lots (Inventory)

How much stock dealers carry, and how old it is, tells you whether today's prices can hold. A fast-growing pile of older stock is usually cleared with discounts later.

New vehicles

GM dealers' new inventory was 587,236 vehicles on September 30, 2026, up 17.6% from June 30 (499,465) and 9.5% from a year earlier (536,424). The sport-utility share of the lot was 50.3%, with pickups at 44.7%. Stock listed over 90 days was 29.0% of the lot, against 29.8% in June and 26.5% a year ago. Active days on market averaged 78 (median 49).

Where the build is:

Most of the new stock is young (the over-90-day share for the Silverado light-duty and the Blazer is below GM's average), which fits fresh production arriving rather than unsold cars aging. The Buick Envision is the exception: more than half of its stock is over 90 days old, at 199 days of supply.

Used vehicles

Dealers' used GM stock is read in Section 5b.

5bResidual Value (used vehicles, outside the composite)

GM earns almost nothing when a dealer resells a used Chevrolet, GMC, Buick or Cadillac, and used sales are not part of GM's revenue. Used figures matter to GM for four reasons. Lease book: GM Financial sets a residual value when it writes a lease and sells the car into the used market when it comes back. Cost of selling new cars: lease payments depend on that residual, so resale values feed the lease support GM provides to keep payments competitive. Trade-ins: most new-car buyers trade in a vehicle, often a GM one, and its value is part of what they effectively pay. Dealer health: franchise dealers earn much of their profit on used cars, pay interest on every car in stock and need lot space for new cars. This reading sits outside the composite verdict and does not move it.

Used prices (lease book, cost of selling new cars, trade-ins). The average used GM vehicle sold for $28,800, 2.8% more than a year ago. Used Trax (+7.4%), Traverse (+5.3%), Equinox (+2.8%) and Silverado 2500HD (+2.4%) prices rose, while the Silverado 1500, the largest used nameplate, fell 1.7%. The discontinued Malibu is draining out of the used market (−32.5%), while the newer Trax is arriving (+30.9%) at a higher price. Across the brand, the resale values behind GM's leases, its lease pricing on new cars and its buyers' trade-ins did not weaken; on the Silverado 1500 they softened.

Used electric vehicles (lease book). Dealers sold 16,448 used fully electric GM vehicles, up 21.4% from a year ago, at an average price of $41,044, up 27.6%, and in 38 days on average. The mix is shifting from older Bolts toward newer off-lease Ultium models (our interpretation) (Section 8). Returned electric vehicles are selling at higher prices and clearing quickly, an input to how those leases settle.

Dealers' used stock (dealer health). Brand used Market Days Supply, across all dealers carrying GM's brands, was 59 days, against 51 in the second quarter and 51 a year ago, at the top of its 15-quarter range. Dealers are carrying more used stock against their sales than a year ago; for the franchise dealers among them, that is stock they pay interest on and must clear to make room for new cars. Across all U.S. dealers, franchise and independent, used GM stock was 561,692 at quarter-end, up 13.3% from a year ago, and the share listed more than 90 days rose to 26.7% from 24.1%. The average list price of that stock was $30,115; by price, 33.0% were under $20,000 and 13.3% over $50,000; by mileage, 22.9% had under 30,000 miles and 30.4% over 100,000. Independent dealers' stock matters to GM only through the prices it clears at, read above.

Used sales (dealer health, lease book). Dealers sold 879,827 used GM vehicles, 3.3% fewer than in the second quarter and 7.9% fewer than a year ago, against a used market down 3.5%; that is 1.5 used sales for every new GM sale in our data, against 1.5 a year ago. These are the dealers' sales, not GM's: they describe dealers' used business and, through how fast cars clear, the market that returned lease vehicles are sold into. Because our new and used counts for GM both run below the market this quarter, we read the decline as direction only (our interpretation).

06Price Position: Sticker Discount and Asking Price

The discount from sticker shows how close buyers are paying to the price GM sets. It is a demand and pricing-power indicator only: it does not measure GM's incentive spending, which mostly runs through subsidized financing and leases that never change the advertised price, and it says nothing about profit.

The average discount narrowed by 0.3 points from the second quarter and is 0.3 points deeper than a year ago: both small moves. Franchise dealers began including mandatory fees in advertised prices from spring 2026 following Federal Trade Commission (FTC) enforcement, so part of any price change across April–May 2026 reflects how prices are advertised rather than what buyers pay; the size of that part is not estimated. Our benchmark bands treat a 2–5% discount as normal incentive activity and 5–10% as aggressive; GM's lineup sits at the line, its light-duty pickups beyond it (Silverado light-duty 10.4% below sticker, Sierra light-duty 9.7% below sticker), and Cadillac near sticker.

Section 6c (dealer's own discount on Silverado 1500 and Sierra 1500) is not shown this quarter: both nameplates failed the quarterly self-check that the split must pass before publication (Section 12).

Used: discount from first listed price. For each used GM vehicle sold, we compare its final price with the first price on its listing. These are readings on used GM prices, which matter to GM through its lease book, the cost of selling new cars and trade-ins (Section 5b); they sit outside the verdict.

The average discount per car narrowed by 0.8 points from a year ago, but that is partly the fee change: the share of used cars that sold above their first asking price rose from 5.3% to 8.7%, the signature of fees added to listings after they went up. The share discounted at all rose to 51.9%, and the value-weighted discount is close to a year ago. The asking premium on unsold stock (quarter-end used list price against the quarter's used ATP) was 4.6%, below the 10% screening threshold.

07Days to Sell and Market Days Supply

These are two different measurements: days on market describes how long each vehicle took to sell; Market Days Supply describes how long the stock on hand at quarter-end would last at the average daily selling rate of the 45 days ending that day.

7aDays on Market (how long each vehicle takes)

New vehicles that sold took 72 days from first advertisement to sale, faster than a year ago and level with the second quarter. The typical car (the median) took 48 days in both years: the gap between mean and median is the long tail of slow sellers. Used vehicles are taking a little longer to sell, and the unsold used stock is older than a year ago; for GM these are readings on dealers' used business and on how fast the market that returned lease vehicles are sold into is clearing (Section 5b).

7bMarket Days Supply (how long the whole lot would last)

The used row covers all dealers carrying GM's brands and is read against its own history only; its franchise part is a dealer-health reading (Section 5b).

On its own history, GM's new Market Days Supply is well outside its normal band and moving up: 95 days on the North American basis against a normal range of 72–79, and at the top of the 54–95 range seen over the last 15 quarters. GM's normal range is mid-pack among manufacturers; the only fair comparison across brands is make by make, on the same basis. 100% means the GM brand is in line with franchise dealers of its peer makes; below 100% means it carries less supply than they do:

Chevrolet still carries less stock than its mainstream peers, but less of an advantage than before. GMC and Buick are both now longer than their peers and above GM's own range: the combination that marks real build risk rather than a difference in how a brand stocks. Cadillac moved from very lean to slightly below its luxury peers.

Against GM's own figure. GM reported 511,000 U.S. dealer units and about 55 days of supply at June 30 (management's own disclosure). Our June 30 count was 499,465, very close, but our Market Days Supply was 73, 132% of GM's. GM's figure counts stock on the lot against a one-month selling rate; ours counts every advertised vehicle, including vehicles advertised before they arrive, against the 45 days to quarter-end.

Coherence check. Sold days on market divided by Market Days Supply is 77% for new vehicles, inside the 69%–108% range we treat as consistent. For used vehicles it is 80%, just below the 82% low end: fresh used units are selling while older stock sits, which fits the older used inventory in Section 5b.

08Electric Vehicle Sales

GM told investors in July to expect lower electric-vehicle volumes and a resumption of building "to demand" in the second half (management's own disclosure). Two questions therefore matter: is the decline running at the planned pace, and are the electric vehicles GM still sells clearing cleanly?

The decline. New battery-electric sales were 22,463 in our data, down 19.4% from the second quarter and 65.1% from a year ago, when the expiring federal tax credit pulled buyers forward. GM's own count fell 61.7% to 25,473 (management's own disclosure). Electric vehicles were 3.9% of GM's new sales in our data, against 4.7% last quarter and 9.9% a year ago. This is the planned outcome, not a demand verdict.

The remaining portfolio.

Pricing on the remaining fully electric portfolio improved: the average price sat 2.8% below sticker, against 5.8% below sticker last quarter, and average sold days on market fell to 116 from 142. Cadillac's four electric models sell within about half a point of sticker. The weak spots are small: the Blazer EV and HUMMER EV take more than 230 days to sell. The open question is stock: fully electric inventory rose to 26,215 from 24,157 in June while sales fell. Fully electric Market Days Supply rose to about 98 from 79.

Used electric vehicles. Dealers sold 16,448 used fully electric GM vehicles, up 21.4% from a year ago, at an average price of $41,044, up 27.6%, and in 38 days on average. The mix is shifting from older Bolts toward newer off-lease Ultium models, and the market is absorbing them quickly (our interpretation). Returned electric vehicles clearing faster and at higher prices is an input to how GM Financial's electric leases settle (lease book, Section 5b).

09The Positive Case

Pricing held through the build. The single most useful price number we have, the average transaction price, rose 1.2% from a year ago and was flat against the second quarter, while the discount from sticker stayed within a third of a point of both comparison quarters. That is consistent with what GM reported in July: pricing at the high end of its guidance and an incentive environment it described as "fairly consistent" (management's own disclosure). Dealers took on 17.6% more new stock in three months without visibly cutting price to move it.

The franchise truck is selling. Silverado light-duty sales rose 15.4% in our data and 13.3% by GM's own count: the two meters agree, and the Silverado is GM's largest single source of new revenue. Across light-duty pickups GM kept its place in a segment that grew 5.5%, and its pickup stock is young: under a quarter of Silverado light-duty listings have been up more than 90 days. If the build is stock laid down ahead of the December launch, it is the kind that sells (our interpretation).

Resale values did not weaken. Used GM prices rose 2.8% and used GM electric vehicles sold for 27.6% more than a year ago, in 38 days. Those prices are inputs to GM Financial's lease book, to GM's lease pricing on new cars and to its buyers' trade-ins; the reading sits outside the verdict (Section 5b).

Outside our data, the tailwinds are back-loaded into this quarter. GM said most of the remaining $0.5–1.0 billion of its warranty improvement would land in the third quarter, and that electric-vehicle losses, regulatory credits and software revenue all continue to improve (management's own disclosure).

10The Negative Case

The stock build is the clearest risk. New Market Days Supply rose to 94 on our U.S. basis (95 on the North American basis against GM's normal 72–79), and GMC and Buick now carry more days of stock than their mainstream peers. The full-size SUVs are the weak spot: Tahoe stock rose 51.0% in the quarter while Tahoe sales fell 12.1% in our data (2.6% by GM's count), and Suburban sales fell 25.0% in our data (11.1% by GM's). If the outgoing trucks and SUVs do not clear before the new pickups arrive, the fourth quarter could need heavier discounting on top of the 35,000-unit year-over-year volume headwind GM already flagged (our interpretation). Today the evidence is stock, not price: discounts have not widened.

The light-duty pickups are already discounted deeply. The Silverado light-duty sold at 10.4% below sticker and the Sierra light-duty at 9.7% below sticker, both about 0.6–0.7 points deeper than a year ago. That is a demand and pricing signal on the advertised price only; it does not measure GM's incentive spending or margin.

Dealers' used business is softer than the market. Dealers' used GM sales fell 7.9% against a used market down 3.5%, used GM stock across all dealers rose 13.3% with more of it aging, and brand used Market Days Supply across all dealers is at the top of its range. For GM the franchise part is a dealer-health reading: stock dealers pay interest on and must clear to make room for new cars. Used prices are holding for now (Section 5b, outside the verdict).

Outside our data, the costs ramp in the second half. GM guided commodity, logistics and memory-chip costs higher in the second half than the first, tariffs at about $0.9 billion a quarter, and onshoring costs rising into the fourth quarter (management's own disclosure). None of these is visible in transaction data.

11ADid Management Get It Right Last Quarter?

Claims from the July 21, 2026 earnings call, tested against this quarter's data. Several of GM's claims are measured on wholesale shipments to dealers or on internal cost lines; those cannot be graded against retail sales and are marked not observable, with our retail reading shown for context only.

The gradable pattern is simple: the inventory claim held, and the demand claim holds for pickups but not yet for full-size SUVs.

11BWhat to Watch on the Earnings Call

1. September 30 inventory and its make-up. Our count of advertised new GM stock rose 17.6% in the quarter to a Market Days Supply of about 94 days. Ask for GM's U.S. dealer inventory and days of supply against its 50–60 day target, and how much of it is outgoing-generation pickups and SUVs built ahead of the December launch.
2. Full-size SUVs. Tahoe and Suburban sales fell in our data while their stock rose sharply. Ask whether full-size SUV demand is as strong as described in July, and whether the build reflects the capacity increase GM announced.
3. Fleet versus retail. GM's U.S. deliveries fell 5.5%; our retail count fell 11.7%, with the gap concentrated in fleet-heavy nameplates. Ask for the third-quarter fleet share (22.3% in the second quarter, from 17.8% a year earlier).
4. Electric-vehicle stock. Retail electric-vehicle sales fell 19% from the second quarter while GM planned to resume building to demand. Ask how the second-half wholesale increase compares with retail sell-through, and for the Equinox EV and Blazer EV plans.
5. The fourth-quarter bridge. GM flagged a 35,000-unit year-over-year volume headwind and higher launch costs for the fourth quarter. Ask whether the third-quarter stock build changes that, and how much of the warranty improvement landed in the third quarter as expected.

11CPost-Print Accuracy Summary (GM Q3 2026 Results, October 20, 2026)

What this section is: This section was added after General Motors reported Q3 2026 results on October 20, 2026. Everything above it is the report as written before those results, unchanged.

Scorecard: Of 34 calls tested against what GM reported, 15 were confirmed, 7 were partly right and 1 was missed; 11 could not be tested because GM does not disclose them (days to sell, used prices) or because the report had already said they could not be checked on retail data. Both tiers of our verdict were Neutral, and Neutral was right: revenue was a record $50.0 billion, up 2.9%, adjusted earnings before interest and taxes (EBIT-adjusted) rose 21.6% to $4.1 billion, the North American margin reached 8.7%, 2.5 points higher than a year earlier, and adjusted earnings per share (EPS) of $3.70 beat the consensus estimate by one cent. GM narrowed its full-year range around an unchanged midpoint, and the shares fell about 5% on its warning about rising commodity, logistics and memory-chip costs, all of which sit outside vehicle transactions.

What we got right: The report’s most important decision was not to read our own volume count. GM sales in our data fell 11.7% while our capture of GM fell, so the volume lens was set to not observable and the share table carried a caution that most of the share move was our measurement. GM reported U.S. sales down 5.5%, all of it in electric vehicles, North American shipments to dealers up 3.5% and U.S. share down 0.7 points, so a volume or share-loss call on our count would have been wrong. Pricing read as firm but not rising, which matches GM’s year-to-date pricing gain and its full-year guide. The pickup, fleet and electric-vehicle readings all tied to what GM reported.

What we missed and why: One call was wrong: for the second quarter running, the deep sticker discount on the outgoing Silverado and Sierra was carried as a negative. End-of-cycle trucks carry factory cash in the advertised price, and our split of that cash from the dealer’s own discount did not pass its check this quarter, so the figure could describe the trucks but could not show the dealer was cutting harder. GM said truck incentives were several hundred basis points below prior changeovers and that pickup mix lifted profit. The stock build we flagged was real, but management called it planned ahead of the pickup changeover, and the risk we attached to it is a fourth-quarter test rather than a third-quarter result. Two smaller points: the year-over-year price rise of 1.2% should have carried, beside the figure itself, the caveat that franchise dealers began including fees in advertised prices in spring 2026, a step of about the same size; and about 40% of the fall in our capture of GM, measured against GM sales that include fleet, was GM’s own shift toward fleet rather than our capture.

Bottom line: Pricing, pickups, fleet and electric-vehicle readings were reliable, and the decision to withhold the volume read was right. Read our year-over-year GM unit counts only beside GM’s own sales release, and read sticker discounts on outgoing trucks as description, not signal.

12Methodology and Glossary

Where the data comes from

Primary data source: MC Compass data was the primary data source: new and used retail sales and the daily inventory panel, filtered to oem_stock_ticker = GM (Chevrolet, GMC, Buick, Cadillac). Sales cover July–September 2026 (current quarter), April–June 2026 (prior quarter) and July–September 2025 (year-ago quarter). The inventory snapshot is September 30, 2026, the true quarter-end; the panel's data edge is October 9, 2026, and September 30 is a stable day. Geographic figures reflect the physical location of the vehicle.

U.S. scope. GM figures in this report are U.S.-only, matching GM's own U.S. reporting; Canadian records are excluded so Canadian-dollar prices do not mix into U.S. averages. Two figures use the North American basis because their benchmarks are built on it: the market-share table and the new Market Days Supply read against GM's own range (95 days North American, 94 U.S.).

Compass totals, Q3 2026 (U.S.): new 573,867 units and $30.8B; used 879,827 units and $21.9B; combined $52.8B. These are dealers' retail sales; the used figures are not GM's sales or revenue.

Revenue basis. Revenue is the sum of captured selling prices and is not units times average price. About 1.1% of new and 13.5% of used sales this quarter carry no captured price.

Basis line. Inventory basis: Compass panel. DOM clock: opportunity start (sold and active days on market are measured from the first day the vehicle was advertised). Market Days Supply: active units on September 30 divided by the average daily sales of the 45 calendar days to quarter-end, August 17 to September 30 (ds_denominator = t45); baselines installed October 19, 2026. GM's own days-of-supply figure uses a one-month selling rate and stock on the lot; ours uses the 45 days to quarter-end and every advertised unit, and the two are compared directly with that difference noted. New-vehicle Market Days Supply counts every advertised unit, including units advertised before they reach the dealership; company-reported figures count stock on the lot and are not directly comparable. GM's own reference range for new Market Days Supply is 72–79 days (middle half of 15 quarters, North American basis); brand used Market Days Supply (all dealers carrying the brand) is read only against its own history.

Market share denominator. Market share is computed against MarketCheck Compass-captured North American retail transactions for the same period, not against total US registrations or company-reported deliveries. Coverage of new-vehicle retail runs approximately 80–95% of the actual market and varies by brand, so these figures are a consistent relative measure across periods, not an absolute share of the true market. Revenue share on this basis includes Canadian rows at their Canadian-dollar prices, as the reference band does. On a U.S.-only basis GM's unit share was 15.4% this quarter and 16.8% a year ago.

What this signal is built on, and what it does not adjust for

Volume, revenue and share in this report are what MarketCheck captured, not GM's results. The direction stated is our reading of the transaction evidence, not a forecast of GM's reported numbers.

Coverage record. Coverage here is our count of new U.S. GM sales divided by GM's reported total U.S. vehicle sales, which include fleet and dealer-use vehicles. The error runs below 1.00: our count is smaller than GM's, partly because fleet sales are rarely advertised at retail.

The year-over-year comparison spans a coverage step of six points, located on both sides; the sequential comparison spans a 2.3-point step. The volume lens is therefore recorded as not observable, confidence is capped at Medium, and the call rests on per-vehicle measures (prices, discounts, days on market), proportions and inventory, which do not depend on how much of GM we capture. Coverage gaps of this kind should be assumed to recur; their size should not be assumed. Eight nameplates were checked against GM's own release; six of them, plus the Buick brand, diverge by more than 10 points and are used for price and days on market only.

Precision. Our sample tracks direction better than level; absolute Market Days Supply and unit figures are best read against their own history.

What our quality checks found

Inventory snapshot date. September 30, 2026, the true quarter-end. Quarter-end new stock 587,236; used 561,692 (U.S.).

Section 6c method. For the Silverado 1500 and Sierra 1500 only, we infer how much factory cash dealers build into their advertised prices from price steps at month boundaries, when one cash program expires and the next posts, and split the discount from sticker into factory cash and the dealer's own discount. MarketCheck does not observe the cash line directly. Each quarter the split must pass two checks before publication; both nameplates failed the first this quarter and are not shown.

Lenses. All manufacturer lenses were run. The volume lens is not observable this quarter (coverage step above). Brand equity (used values) is read in Section 5b and sits outside the composite: GM earns almost nothing on a used sale, and used figures are inputs to its lease book, the cost of selling new cars, trade-ins and dealer health, which our data does not see.

Signals unlikely to be checkable on the call. GM does not disclose retail days on market, the age of dealer stock, used-vehicle prices or retail electric-vehicle pricing; those lenses are context for the call, not scored predictions.

What we cannot see

Outside MarketCheck's scope for GM: production and wholesale shipments to dealers; fleet sales not advertised at retail; GM Financial (captive finance; $0.6B earnings before tax in the second quarter, $2.5–3.0B guided); OnStar, Super Cruise and other software revenue (more than $3B guided; $6.3B deferred revenue); warranty costs ($1.0–1.5B guided improvement); regulatory credits ($0.5–0.8B guided savings); electric-vehicle restructuring charges ($2.3B in the second quarter; $10.9B since mid-2025) and loss improvement ($1.0–1.5B guided); tariff costs ($2.5–3.5B gross guided); onshoring costs ($1.0–1.5B); GM International and China equity income; GM Defense and GM Insurance; and sales outside the United States. Commodity and raw-material exposure (not observable in MarketCheck): steel, aluminum, palladium and platinum, copper, plus memory chips; guided as a $1.5–2.0B 2026 headwind with $0.6B in the first half and more in the second. Hedging programs partially offset short-term volatility; the net commodity headwind or tailwind on adjusted operating profit is disclosed in management's quarterly bridge and is structurally larger for truck- and SUV-heavy manufacturers.

Sources of figures. Every figure in this report comes from MarketCheck data unless the text names another source, such as a company filing or management's comments on its earnings call. Only figures from those other sources carry a source label.

Plain-English Glossary

Prepared by MarketCheck Research using the Anthropic Claude LLM. This report describes North American retail vehicle transaction activity only and is not investment advice.

We welcome your feedback. To offer comments, suggestions for future coverage, or to inquire about licensing this report or the MarketCheck data behind it, including vehicle-level sales, inventory and pricing data, contact Bryant Sheehy at bryant.sheehy@marketcheck.com.