Tesla, Inc. (TSLA.US) Q3 Earnings Report: Savings on the Books, but the AI Story Drifts Further Away

Tesla, Inc. (TSLA.US) Q3 Earnings Report: Savings on the Books, but the AI Story Drifts Further Away After Tesla, Inc. (TSLA.US) released its Q3 earnings report, the stock fell more than 3% in after-hours trading. Looking at the data alone, this report is not bad: profit, cash flow, and capital expenditures beat Wall Street expectations, while revenue was only slightly below. The real turning point came during the earnings callfull-year capital expenditure guidance was cut from $25 billion to $20 billion. Management framed "spending less" as discipline, but multiple analysts interpreted it as projects being pushed back. Q3 results were mixed: cash flow improvement came from spending pace, while profit did not keep up The report shows Tesla, Inc.'s third-quarter revenue was $28.05 billion, flat year over year and slightly below the market expectation of $28.13 billion. Adjusted earnings per share were $0.48, above the market expectation of $0.44; adjusted EBITDA was $3.90 billion, above the market expectation of $3.66 billion. Cash flow was the most closely watched part of this report. Last quarter, free cash flow was negative $1.092 billion, so the market expected negative $4.99 billion this quarter; the actual result was positive $1.53 billion, turning from negative to positive. But breaking it down, operating cash flow was $6.263 billion, only a little more than $20 million higher than the same period last yearessentially flat; what really made the difference was capital expenditures. Although $4.732 billion doubled year over year, it was still $3.2 billion less than the market expectation of $7.95 billion. The earnings call then cut full-year capital expenditure guidance from $25 billion to $20 billion, with the CFO calling it "efficient management of capital deployment." In other words, cash flow beat expectations mainly because of the spending side; the beat came from spending less, not earning more. The market did not fully buy it. Roundhill Investments CEO Dave Mazza put it bluntly: for most companies, spending less and generating more cash is a good thing, but Tesla, Inc. is the exceptionthe market does not value it on the basis of a traditional automaker. Capital expenditure is the only hard indicator investors have for judging how fast Optimus and Cybercab are being built. It was raised by $5 billion just six months ago and is now cut by $5 billion again"it looks less like discipline and more like future capacity ramp-up slipping." Analyst disagreement was concentrated here as well. StoneX's Mickey Legg was cautious: the Q3 spending shortfall "reflects both some projects slipping and the overall plan getting smaller." Morningstar's Seth Goldstein was more optimistic: lower capital intensity means Tesla, Inc. can "self-fund without needing to issue debt or equity." But the optimistic view has a preconditionthat projects are not delayedand the earnings call precisely did not prove that. The income statement tells another story. Revenue was flat year over year but slightly below expectations; operating income was $718 million, down 56% year over year; operating margin was 2.6%, down 322 basis points from a year earlier. Gross margin rose 160 basis points year over year to 19.6%, adding about $440 million in gross profit; but operating expenses rose $1.35 billion year over year to $4.776 billion, which not only offset that improvement but also reduced operating income by $906 million year over year. GAAP net income was $787 million, down 43% year over year; non-GAAP net income was $1.697 billion, down only 4%. The roughly $900 million gap between the two mainly came from stock-based compensation, digital assets, and unrealized gains and losses on SpaceX equity investments. In other words, the sharp GAAP decline included a large amount of volatility unrelated to the core automotive business. The cash flow statement looks better than the income statement; the income statement is more honest than the cash flow statement. The cash flow improvement is real, but it came from the pace of spending, not from better core operations; the income statement did not keep up, with expense expansion far exceeding gross profit improvement. Automotive and energy storage: inventories and the buffer both thinning Data shows Q3 automotive revenue was $20.56 billion, down 3% year over year; deliveries were 486,532 units, down 2% year over year but above the market expectation of 463,000. This "beat" needs a discount: production was 464,391 units, and deliveries exceeded production by 22,141 units. Some of the cars sold in Q3 were inventory left over from Q2. Regional performance was about offsetting each other, not a synchronized recovery. The U.S. weakened after the federal EV tax credit expired; Cox Automotive estimates Tesla, Inc. sold about 144,000 vehicles in the U.S. in Q3, down nearly 20% year over year, but still holding about a 54% share of the U.S. EV marketmore than all other brands combined. Shanghai wholesale sales in September were 95,366 units, the best month of 2026, with exports up 58% and domestic sales down 9%. European registrations rose 43.3% from January to August. The product mix deserves more attention. Models other than Model 3/Y delivered only 8,295 units, nearly halving year over year. Roadster is still in the "design and development" stage, with capacity undetermined and mass production not until 2028. Deepwater's Gene Munster called it an "engineering showpiece," and Morningstar expects annual sales of fewer than 2,000 units. It is a brand symbol, not a volume product. Energy storage deployments were 13.7 GWh, up 10% year over year, but energy revenue fell 16%. Management attributed part of this to tariff-related customer refunds and acknowledged a decline in energy gross profit. This former profit buffer is thinning. Services and other business was a bright spot: revenue grew 33%, with a 15.9% gross margin, driven by charging, insurance, and parts, but the scale is too small to offset declines in automotive and energy storage. There is also an overlooked detail: regulatory credit revenue continued to decline and was listed by management as a drag on revenue and operating income. AI: where the $20 billion is going, and where the return timetable is The earnings call spent the most time on Optimus, but it was not about progressit was about bottlenecks. When an analyst asked what the biggest problem with the current design was, Musk's answer was crisp: the hands. He explained that humans have self-healing ability, while Siasun Robot&Automation does not; fingers are soft, and the skin is full of sensors. This explains why coming off the production line does not equal commercializationthe hardest part of the hardware has not been fully solved. The timetable was also extended. Musk said it would take another year to deliver Siasun Robot&Automation to customers. This matches earlier statements that "a million units annualized, 20,000 units per week at least another year away, possibly longer." In other words, it will be hard for Optimus to make a material contribution to revenue before 2027. The contradiction is right here: if Optimus is an important destination for the $20 billion in capital expenditure, why was spending cut before the production line had even begun to ramp and the hand problem had been solved? Integrity Asset Management's Joe Gilbert said the capex cut "is inconsistent with their ambitions," and it would not be surprising if next quarter's number were revised upward. Zacks' Andrew Rocco tried to offer a neutral explanation: the cut may be due to deployment timelines; the company is in a huge transition period, and the key is for management to find a balance between future growth and CKH HOLDINGS' current cash flow. Cybercab's situation is more concrete and more urgent than Optimus's. As of mid-October, more than 300 Cybercabs were in service in Austin, with the fleet growing about 8x in 30 days, and Dallas confirmed as the next city; but Tesla, Inc. did not disclose how many of those vehicles were operating fully without safety monitors, nor did it provide paid ridership or unit economics. On the regulatory side, there is a hard constraint unaffected by capital expenditure: NHTSA issued 21 special inquiries about the Cybercab, which has no steering wheel or pedals. Tesla, Inc. requested an extension, and NHTSA gave only 30 days, with a deadline of October 30. There is also a timing gap worth notingTesla, Inc. plans to deliver Cybercab to consumers by the end of 2026, but unsupervised FSD will not enter consumer vehicles until the second quarter of 2027. That means the Cybercab consumers receive will, for as long as half a year, be a vehicle with no steering wheel and no usable driverless software. Terafab is a longer bill. Musk said equipment costs are "tens of billions of dollars" and the long-term total cost is "more than $100 billion," shared by Tesla, Inc. and SpaceX. Most orders for the first set of equipment have been placed, with the goal of producing the first test chip in May or June next year. This means the long-term pressure on capital expenditure will not disappear just because this year's figure was cut to $20 billionthis year's cut looks more like moving spending later rather than canceling it. Karobaar Capital's Khurshid gets to the core: if spending is merely deferred, the cash flow improvement is temporary; if the long-term project timetable is unchanged, future capital expenditure will come back. Merger: committee established, zero incremental information In the SpaceX earnings call the day before, Musk had just confirmed that both sides had set up independent special committees of the board to evaluate a merger, and said the Tesla, Inc. side needs 50% approval of all shares, a "fairly high threshold." The market had been waiting for the Tesla, Inc. earnings call to provide more details. There were none. The earnings report did not mention it, and the analyst call did not mention it either; management has now said twice that it would not discuss the matter. That itself is a signal. The merger went from "committee established" back to "no new information," and the 50% all-share threshold means the transaction could drag on for a long time. The enthusiasm from the day before cooled the next day. Analysts also noticed the mood on the call. SLC Management's Mullarkey said Musk "looked less engaged" and "seems bored with the auto industry, and has for some time." That mood does not change the numbers, but it affects market confidence in whether management can strike a balance between future growth and CKH HOLDINGS' current cash flow. Conclusion To be precise, this was not a report that beat expectations across the board. Revenue missed expectations; profit, cash flow, and capital expenditure beat expectations, but breaking it down, the profit beat included accounting items and expense structure, while the cash flow and capex beats came from "spending less." None of it came from better core operations. That determines the nature of this report: the short-term numbers can pass muster, but the narrative supporting the valuation has not strengthened. The improvement in free cash flow came from a capital expenditure cut, and multiple analysts interpreted that cut as projects slipping or the plan shrinking, not operating leverage. Optimus's hand problem, the one-year-away delivery timetable, and Terafab's more-than-$100 billion long-term bill all show that the return timetable for the AI business is being pushed back, while spending pressure has not disappeared. The merger had heat the day before and fell silent the next day, cooling market sentiment. The 3% after-hours decline is less a rejection of the earnings report than a mild vote that "these good numbers are still not enough to support the current narrative." Next quarter's test will be tougher: after the inventory buffer is used up and capital expenditure stops being revised down further, where will profit come from; after the Optimus and Robotaxi timetables are pushed back further, what will support the AI premium?