Tesla Q3 2026 Deliveries: A 5% Beat Built on Inventory, and an Energy Margin Down to 20.4%
Tesla delivered 22,141 more cars than it built in the third quarter, almost the whole of a 5% beat. Energy storage, the business bulls point to, earns less per dollar than a year ago and less in total than the services line.
Tesla delivered 22,141 more cars in the third quarter than it produced. According to Tesla's release, Tesla built 464,391 vehicles and handed over 486,532, which means the difference came out of stock that already existed. Friday's rally in the shares is paying for a beat that, by that measure, is mostly a clearance.
The consensus Tesla compiles for itself was 461,974 deliveries, so the headline beat is about 5%, or roughly 24,500 cars. Production alone came in about 2,400 above that number. Strip the inventory release out of the deliveries and Tesla would have landed a hair over expectations rather than well clear of them. That is a respectable quarter, short of the demand surprise the move in the stock implies, and a warehouse can only be emptied once.
The year-on-year picture agrees. Deliveries are down 2.1% from the record 497,099 of the third quarter a year ago. The second quarter was nearly as strong, with deliveries above 480,000, after a first quarter of 358,023. Tesla has now sold more than it built for two quarters running: over 480,000 delivered against over 450,000 built in the second quarter, then 22,141 more than it built in the third, roughly 50,000 cars out of stock in six months. Sales running ahead of the factory is a pattern that ends when the stock of finished cars does. In Friday morning trading the shares were up 5.08% at $372.09, a price that assumes the next quarter's volume comes from somewhere other than the lot.
Energy storage is where the bullish reading gets more ambitious, and where the numbers hold up less well. Tesla deployed 13.7 GWh in the quarter, up from 13.5 GWh in the second quarter and 8.8 GWh in the first. The 13.7 GWh is not a record: the record is 14.2 GWh in the fourth quarter of 2025, and Tesla's own second-quarter update called 13.5 GWh its "second-best quarterly deployment number." On volume the cars beat and energy only kept pace.
Volume is the lesser problem. The more telling figure is what each deployment earns. Tesla's 10-Q for the second quarter shows energy gross profit of $640 million at a 20.4% margin, against $846 million at 30.3% a year before, and 28.7% for the half. Revenue implied by those figures is about $3.14 billion, up from about $2.79 billion, so revenue rose and gross profit fell by about a quarter. The cause is price: the deployments are larger and each is earning less.
Against the rest of the company, that margin looks small. Tesla's total gross profit in the second quarter was $4.751 billion, so energy supplied about 13% of it. Services and other earned a record $648 million at a 14% margin, according to the same Q2 update. The energy business, the one with the growth story, produced less gross profit than the services line. Nobody prices a service line at a premium for being a stabiliser, and the energy line has not yet earned that treatment either.
The bull case has a name attached. Deepwater's Gene Munster says the "EV winter" is thawing and sees 15% growth in full-year 2027 deliveries. He also points to energy compounding in volume, which the series from 8.8 to 13.5 to 13.7 GWh supports. The volume argument is correct and beside the point for a margin question. A business that deploys more gigawatt-hours at a falling price per kilowatt-hour can grow revenue and shrink profit at once, and the second quarter did exactly that. Whether the price has stopped falling is the one thing the delivery release cannot tell anyone, and Tesla itself warns in the release that deliveries and deployments "should not be relied on as an indicator" of its financial results.
The car side has the same structure. Munster's 2027 growth figure is a forecast about demand, and the evidence in hand about demand is a quarter in which sales ran ahead of production. A real thaw would show production rising to meet deliveries with the lot staying full, and this quarter showed the reverse. Inventory is a useful buffer for a company that wants to show volume while a new product ramps. It becomes a problem if it is producing the volume, because the next quarter has to be built to be sold.
The two stories separate cleanly. Cars: a beat flattered by stock, with deliveries still lower than a year ago, and a growth forecast that has to be earned by a factory that has been out-sold by its own showroom. Energy: more gigawatt-hours at a margin that gave up roughly ten points in a year. A holder who bought on the energy floor argument is relying on a quarter that is not yet in the accounts: the only energy margin on file is the 20.4%.
The test arrives on Wednesday, October 21, when Tesla publishes third-quarter results after the close and holds its call at 5:30 p.m. ET. The floor argument needs energy gross margin back above 25%, and anything near the second-quarter level means price is still eating the volume. The inventory line should show roughly the drawdown the delivery release implies, and automotive gross margin will show what Tesla paid to move those cars, since a clearance that costs margin is a weaker result than one that does not. If energy margin recovers past 25% and the inventory line looks sensible, the energy argument gets its first real evidence. If it does not, Friday's buyers paid for the wrong business.