Financial Report Preview | Roadster takes off and profits fall, can Tesla (TSLA.US)'s “narrative valuation” hold up?

The Zhitong Finance App learned that after the market on October 21, Tesla (TSLA.US) will hand over its financial report for the third quarter. The company's third-quarter deliveries fell 2.1% year over year, but currently, its dynamic price-earnings ratio is over 200 times. The high valuation reflects expectations for autonomous driving software and energy storage technology rather than expectations for current operating margins in the automotive industry.

Prior to that, on the evening of October 15, at SpaceX's McGregor test site in Texas, Tesla released the second-generation Roadster, which had been delayed for many years: the vehicle climbed and hovered upside down on a giant U-shaped suspended runway, then vacated with an air-conditioned propeller jointly developed with SpaceX, finished rolling the barrel in the air, then landed steadily and returned to the demonstration runway. Musk called it “the greatest product demonstration in history,” and left a widely quoted phrase: “It turns out there are indeed flying cars in the future.” The demo came out six days before the earnings report, paving a new forward-looking guide for Musk during the conference call; in Wall Street's view, it was not the sports car that would turn back on its head.

A textbook-level demonstration, and a no-buy capital market

The technical details of the demonstration were indeed rare: the chassis fan generated adsorption force to hang the car upside down on the runway, the air-conditioning propeller completed eviction and attitude control, there was no driver in the car throughout the process, and the movement was completed remotely by a pre-set program (Dong Flush and Dongfang Wealth quoted live reports). Tesla claims 0.81 seconds of acceleration from 0 to 60 mph, and offers three versions: Performance, Sport with additional rear propeller, and Rocket, which has a full-body distributed propeller and can fly stably for a short period of time. Pre-order thresholds were announced simultaneously — Performance requires a deposit of $50,000, Thrust Edition is $250,000, and Rocket Edition is an invitation-only single-seater model.

Commercialization is the problem. The Roadster is targeted for mass production in 2028, and the promised delivery date of the car was 2020 (J.P. Morgan Rajat Gupta) when it first launched in 2017. What's more critical is compliance: the Rocket version with a full flight kit does not allow access to public roads and is limited to closed grounds, and the Performance version, which can go on the road, does not include the full flight kit — “what can fly can't be bought, can't fly.” The European Aviation Safety Authority also stated that any means of transport operated as an aircraft must meet applicable aviation safety requirements. The author of the original article, Brett Ashcroft Green, put it more bluntly: even the Rocket version can't have the same capabilities as in the demo. “DOT and NHTSA are clearly not ready for cars that can overcome traffic jams.”

The stock market's reaction confirmed this reservation. On the first trading day after the presentation (October 16), Tesla rose more than 2% in the intraday period and closed down 2.13% to $354.72. The cumulative decline for the whole week was 7.31%, and the decline increased to about 21% during the year. During the same period, the Nasdaq 100 index fell 5.49% and Nvidia fell 6.61% in the same week — the cooling of AI trading dragged down high-multiple growth stocks, and was the dominant force this week.

Accounts are already on the desktop: delivery exceeds expectations, profit expectations have been lowered all the way down

Third-quarter deliveries were announced on October 2:486,532 vehicles, up 1.3% from the previous quarter, better than market expectations for two consecutive quarters, but still 10,657 fewer than 497,099 vehicles in the same period last year, down 2.1%. Structurally, Model 3/Y delivered 478,237 vehicles, which was basically the same year on year. Other models had only 8,295 units, a sharp drop of 47.9% year over year. Cybertruck popularity declined mainly because of the decline in popularity of Cybertrucks; the high base for the same period last year also cannot be ignored — it was the rush season before the $7,500 federal tax credit expires on September 30, 2025. The installed energy storage capacity was 13.7 GWh, up 9.6% year on year and 1.5% month on month.

The profit side is another curve. The consensus was on adjusted earnings of $0.43—0.45 and revenue of $27.8 billion to $28.1 billion (Zacks was $0.44/$27.8 billion; AlphaStreet combined 24 analysts were $0.45/$28.11 billion; EPS estimated range of $0.30-0.70), down 10% — 14% from $0.50 in the same period last year. The consensus was reduced by 3 cents in the past 30 days; if you look at the 90-day window, the EPS consensus was reduced by 18.2% (AlphaStreet) from $0.55; the author of the original article counted 18 cuts. The lesson from the last time is still in sight: the adjusted EPS for the second quarter was $0.33, 34% lower than the consensus, and the biggest loss in nearly eight quarters. The options market implied a two-way fluctuation of 7.43% in this financial report, which is higher than the average absolute fluctuation of 5.95% after the previous four financial reports.

Profit Margin: The Real Battleground

The problem in the second quarter was “sell more, earn less for every car sold”: after excluding regulatory points, the gross margin of cars fell from 19.2% to 16.3%, and gross energy margin fell from 39.5% to 20.4% (the latter included a historical warranty adjustment of US$240 million and the loss of tariff revenue). Management explained that after excluding the one-off factor of about US$230 million, the gross margin of cars remained roughly flat month-on-month; Morgan Stanley expects the gross margin of cars excluding points to be 16.4% in the third quarter.

What's even more impressive is the operating margin: only 1.4% in the second quarter, and every dollar of revenue was only about 1.4 cents in operating profit. According to business quant data as of June 30, 2026, Toyota 7.86%, Honda 8.76%, GM 3.04%, Ford 1.32%, and Ferrari 31.22% for the same period.

Comparing Tesla with Ford, GM, and Toyota will draw two conclusions — it is still leading in terms of gross margin, falling behind Toyota in terms of operating profit margin; and the market clearly doesn't price it according to car companies. His evidence is the equity curve: the other three companies can only use operating cash flow to pay employees and have almost no ability to raise additional capital without breaking stock prices. Tesla's share capital expansion path is therefore “very different,” and it was able to integrate equity capital and basically avoid long-term debt during periods of high stock prices — it is still the only car company whose cash and short-term investments far exceed long-term debt, which gives it room for trial and error.

It should be added that Toyota's own operating profit margin is also declining: FY2025 is 10.0%, and FY2026 sources range from 6.3% to 7.9%, mainly due to US tariffs and the situation in the Middle East. While Tesla “lost to Toyota,” Toyota also retreated.

Imagination Checklist: FSD, Robotaxi, Energy Storage, and a $25 billion bill

The author lists three high-margin businesses that have yet to be realized: FSD subscriptions (he likens the Apple App Store model, but reminds consumers that the entry threshold is much higher than mobile phones subsidized by operators), Robotaxi scale-up (he believes the real risk is a capacity bottleneck — “spend an hour more to build this, make another model in less than an hour”), and energy storage (which he judges is more about revenue growth than high gross profit, and may self-eat the lithium supply needed for vehicles).

The latest progress is: FSD has 1.48 million paying users, an increase of 56% over the previous year, and the North American selection rate given by Morgan Stanley is 55%. On the Robotaxi side, after the launch of Cybercab in September, the Austin fleet grew 8 times to more than 300 cars in 30 days. The cumulative unsupervised mileage exceeded 1 million miles (about 380,000 miles in the second quarter earnings report), and the network expanded from 3 cities to 7. Dallas was the next stop.

Costs are rising at the same time: capital expenditure guidance for 2026 exceeds US$25 billion, and the company has arranged up to US$30 billion in debt financing; capital expenditure doubled year on year to US$5.79 billion in the second quarter, free cash flow was about -1.1 billion US dollars, and capital expenditure in the first half of the year was US$8.28 billion, more than doubling year on year. Musk called this stage “the fastest industrial scale expansion since World War II.” Another regulatory variable hanging over the financial report: NHTSA has asked Tesla to make a sworn response on how the pedalless Cybercab was certified according to federal safety standards until October 30, nine days after the call.

SpaceX merger: the ultimate engine of the story, and the least uncertain piece

The author of the original article made no secret of his judgment: “I think this (merger) will happen.” The logic is: the Roadster demo itself shows the collaboration between the two companies; the propeller can be used for rovers in the future; Starlink has become a key asset in the US defense system during recent military operations; if the two were merged, Tesla would be like an “early version of General Electric” — an industrial and defense company, too big to collapse. He also included Optimus in the same narrative.

He's not alone in the market betting. Barclays analyst Dan Levy suggested in a report from October 14 to 15 that the two companies may announce a merger before Tesla's annual shareholders' meeting (which must be held on December 6 at the latest by law), and that SpaceX will acquire Tesla on an all-stock basis. The value is not a traditional car collaboration but vertical integration with AI's Grok and Macrohard — combining Tesla's Robotaxi and Optimus with SpaceX's Grok and Macrohard. Wedbush's Dan Ives gave the probability of completing the merger by 2027 is over 80%. The contract probability of “announcing a merger agreement by March 2027” on Kalshi rose from 29% on October 14 to 56% on the morning of October 16.

Up to now, neither company has officially announced any merger plans. The predicted market price reflects trader pricing rather than the existence of an agreement. The Barclays report is based on analysts' scenarios rather than company disclosure. In the same period, Polymarket's price for “completing the merger by the end of 2026” was only about 24%, and Kalshi's other caliber was about 19%, indicating that the market's differences over “recent implementation” far outweigh the differences over “what will happen in the long run.” The two companies already have substantial links: SpaceX acquired xAI in February of this year, Tesla invested 2 billion US dollars to acquire less than 1% of SpaceX's shares in March, and the two sides jointly promoted the Terafab semiconductor project; Ostian estimates that GAAP net profit of about 1 billion US dollars in the second quarter came from book income from holding SpaceX shares. Even as it progresses, the obstacles are specific: two shareholder votes, an antitrust review, and a national security review triggered by a defense contractor absorbing a car company with deep Chinese exposure.

What do agencies think, and a “gentler” alternative

In terms of price targets, sellers are less divided but the direction is conservative: Investing.com compiled 38 analysts and gave a 12-month average price target of $397.83, including J.P. Morgan $415 (neutral), Morgan Stanley $400 (neutral), Goldman Sachs $360 (neutral), and UBS $391 (holding).

Morgan Stanley's Andrew Percoco takes a constructive view, citing demand resilience, FSD's broader approval to boost demand and gross profit, and undervaluation of the energy business, but is cautious about Optimus's commercialization schedule, and emphasizes short-term upward restrictions on cash consumption and valuation. His attitude towards Robotaxi best sums up the current divide: scaling up is a good thing, but “utilization, safety performance, and unit economy are the next tests.”

The analysis points out that this quarter will not look good, and may even fall short of the revised EPS expectations. The company needs a new deliverable product to catch the gap left by Cybertruck; as SpaceX is already listed and merger expectations gradually rise, the alternative Tesla needs to provide is the Baron Partners Fund (BPTRX) — the fund gives Tesla and SpaceX a total exposure of about 46%. Ron Barron said he will hold both for a long time, and the volatility is lower than that of any of them alone; the author also Disclosing that you and your clients hold long positions on BPTRX and TM should be considered a matter of interest.

There is also a reverse set of figures at the valuation level: according to Ostian's estimate, Tesla's forward-looking non-GAAP price-earnings ratio is about 215 times (about 15 times the industry median), and the forward PEG 11.5 (industry 1.25) is still around 174 times the 2027 estimate; the 2027 forward-looking price-earnings ratio is 207.84 times, and it is also said that only a few analysts dare to predict the profit trajectory after 2028 — those who dare to predict a price-earnings ratio of about 18.87 times ahead of 2032 means Profits will increase approximately 11 times in six years. This is the essence of this financial report: the market is still paying for imagination, and the October 21 statement examined the distance between imagination and cash flow.