Tesla’s AI gamble collides with a profit collapse

Tesla’s record revenue masks a profit collapse and negative cash flow, leaving the shares priced for an AI outcome still years away.

Mark Rogers Mark Rogers

Hold

Tesla’s (TSLA) shares have fallen 18.9% in the past week, closing at $313.03 on 24 July 2026, a long way down from the $498.83 high the stock touched over the past year.

The drop followed second-quarter results released on 22 July, which triggered a 12% single-day slide and wiped out more than $140bn in market value, as investors who had waited years for evidence that Tesla’s AI bets could pay off were instead handed a profit warning dressed up as a delivery record.

Revenue climbed 26% year on year to $28.24bn, pushing trailing twelve-month sales past $100bn for the first time, while deliveries hit a second-quarter record of 480,126, roughly 18% above the analyst consensus of 406,000 units.

Yet very little of that growth reached the bottom line, since operating income fell 57% to just $398m, dragging the operating margin down to 1.4% from 4.1% a year earlier, while adjusted earnings per share of $0.33 missed Wall Street’s roughly $0.53 estimate by a wide margin.

Spending Surge Puts Margins Under Pressure

The car business itself held up reasonably well, with automotive gross margin slipping only modestly to 16.9% from 17.2%, which shows the damage came from elsewhere.

Operating expenses jumped 47% to $4.35bn as Tesla poured money into Optimus, Cybercab, the Nevada Semi factory and its Austin chip fab, and free cash flow turned negative for the first time in years, at minus $1.1bn, after capital expenditure more than doubled sequentially to $5.79bn.

Management has guided for full-year capex above $25bn, with further increases expected over the next two to three years, which is why this spending plan sits at the centre of the whole investment case.

Robotaxi and Optimus Carry the Valuation

Tesla trades at a price-to-earnings ratio above 300, a multiple that only makes sense if robotaxis, Optimus and AI infrastructure eventually generate returns nobody can yet measure.

Robotaxi now operates unsupervised in seven US metros, including newly launched services in Miami, Orlando and Tampa, with cumulative paid robotaxi miles approaching 2.4 million, though Musk described the pace as gated by the “march of nines” needed for safety rather than by demand, which pushes timelines further out than bulls had hoped.

Optimus sits further behind, with construction of the first production line only beginning in Fremont this quarter and external deliveries not expected before 2027, and Musk called it the hardest manufacturing challenge Tesla has ever attempted, with no existing supply chain to lean on.

Energy Business Shows Stronger Momentum

Energy storage offered a genuine bright spot, with deployments up 41% to 13.5 GWh and record EMEA volumes, while services and other revenue grew 50% and FSD subscriptions rose 56% to 1.48 million.

These are the parts of the business actually growing profitably, but they remain too small to offset the pressure building in the car division.

Bulls Still Bet on Tesla’s Long-Term Vision

Wall Street analysts are split rather than despondent, with Cantor Fitzgerald and TD Cowen maintaining price targets near $485 to $510, implying upside above 50% from current levels, while JPMorgan and Morgan Stanley sit on hold ratings closer to $400 to $445, and that spread suggests the models still credit Tesla for a future that hasn’t shown up in the numbers yet.

A net cash position of $43.5bn gives Tesla room to keep funding this build-out without existential risk, but management has already flagged that capex will keep rising for two to three more years, meaning the cash drain shown this quarter is closer to the start of the cycle than the end of it.

Musk’s own framing on the call, that robotaxi expansion is limited by safety validation rather than demand, cuts against any hope of a quick reacceleration in the numbers that matter to the share price.

Holding through this stretch means betting that Optimus and robotaxi economics arrive roughly on the timeline management describes, at a valuation that leaves almost no cushion if they don’t. That is a different bet to the one investors were making a year ago, when vehicle margins alone still did most of the work.