Tesla, Inc. TSLA

Key data

The business

Tesla sells electric cars, and most of the money still comes from them: the mass-market Model 3 and Y, the pricier S and X, and the Cybertruck. Bolted onto that are three other things. A fast-growing energy business that sells battery packs, the Megapack the size of a shipping container for power grids and the Powerwall for homes. A software business that sells Full Self-Driving as an option worth about $8,000 on the car you already own. And two bets with no revenue yet: a paid robotaxi pilot and a human-shaped robot called Optimus. The moat is the brand, the charging network, and the manufacturing scale, wrapped around more real-world driving data than anyone else has.

The numbers

Revenue is growing again while profit falls away. The last quarter was the strongest top line in a year, and the least the company earned on it.

Q2 revenue of $28.2B was up 25% on the same quarter last year, but operating profit fell 57% to $398M, a margin of 1.4%, and the $0.32 earned badly missed the $0.50 the Street looked for. The squeeze from fading regulatory credits and rising costs has deepened, which tells you it is structural, not a one-quarter dip.

Revenue compounded about 15% a year from 2021 to 2025, but that hides the turn: it peaked at $97.7B in 2024 and slipped for the first time in 2025. Profit is the real tell. Net income peaked at $15.0B in 2023 and fell to $3.8B two years later, a 75% drop, and diluted EPS went from $4.30 to $1.08. The market pays 300 times that trough because 21 analysts model earnings nearly tripling to $3.14 by 2028. What the filings do not contain is a single dollar of robotaxi or Optimus revenue, and the car business has given up operating margin every year since 2022. What settles it is the first quarter automotive margin stops sliding, or the first disclosed robotaxi revenue line.

Read the middle column against the first. Gross margin fell from 25% to 18% in the price-cut years and has held there since. Operating margin kept sinking to 2.6% anyway, because the money the factory makes is being spent, on autonomy, before it reaches the bottom line. That is a choice, not a collapse. Whether it is a good one is the whole question.

Management

Elon Musk put $523.8M of his own money into the stock over the past year, seven open-market purchases, the largest being three purchases totaling $340M on September 12, 2025. That is the loudest insider signal on this name and a genuine one: a founder buying at these prices believes the payoff comes. Against it, 66 insider sales totaled $112M, plan status not disclosed on the filings, small enough and scattered enough to read as routine officer selling rather than a verdict. Pay is the milestone-tranche CEO award. There are no buybacks of size; the company spends instead, capex of $8.5B last year and climbing, funded by $14.7B of operating cash flow. It stopped disclosing unit deliveries by model in the quarterlies, so the outside read leans on revenue.

How it fails or surprises you

The bankroll runs dry. Automotive operating margin has fallen from 17% in 2022 to 2.6% for the first half of 2026. If price cuts and vanished credits push the car business toward breakeven, the cash that funds the autonomy bet dries up before it pays off. First print: automotive gross margin excluding credits below 10% for two straight quarters.

Robotaxi converts (right tail). Tesla runs a paid robotaxi pilot and books no revenue for it. If it scales to real fleets across several cities inside 18 months, a service with software margins gets bolted onto a company still valued on cars. The market owns the option and has nothing to model. First print: a disclosed robotaxi revenue line.

The 25% growth is not cars. Revenue rose 25% while profit stalled, the fact this read explains least. If that growth is energy storage and FSD rather than more cars sold, the car franchise is shrinking faster than $28B a quarter suggests. What proves the bear wrong: automotive revenue, not total, rising on genuinely higher unit volume.

Closing thoughts

The payoff here depends on two prints you can watch: whether automotive margin stabilizes and whether robotaxi revenue appears. Roughly $43B of cash and short-term investments against $7.7B of long-term debt means the left tail is a re-rating, not a bankruptcy. If the recovery to $3.14 does not come and no autonomy revenue appears, 300 times earnings drifts toward the 17 times the market pays other carmakers, and that is a very long fall. If autonomy converts, the multiple looks cheap in hindsight. You lose valuation, not the company, and that asymmetry is the reason to hold a name this expensive at all.

The bet is still that Tesla becomes a robot and software company before the car business stops paying for it. What breaks it is the pair to watch: automotive margin excluding credits, and any disclosed robotaxi revenue. If margins keep sliding and that robotaxi line stays empty another year, you are holding a carmaker at a software price, and the market will eventually charge you for the difference.