SpaceX Didn't Cause Tesla's Slow Summer. It Was Inside the Earnings.

The story of Tesla's slow summer is told as two bad things stacked up: first came the SpaceX IPO, then came the earnings rout. Everyone assumes the two are connected, and that both are bad for the stock. One of those assumptions is backward, and the other never had a mechanism behind it at all.

Start with the headline anybody can verify. As of early September, Tesla shares were down roughly 19% for the year and about 27% off their 52-week high near $499. The worst moment came on July 24, when the stock fell 14% in a single day to its lowest close in about a year, right after the company reported second-quarter results. That day is the "earnings rout." The slow drift around it is the "SpaceX aftermath." Two labels, one lazy summer.

Read the earnings release, though, and the word "rout" does a lot of unfair work. Tesla delivered a record 480,126 vehicles in the quarter, up 25% from a year earlier, and revenue of about $28.2 billion beat forecasts. The miss was entirely elsewhere. Adjusted earnings came in around $0.33 a share, more than a third below what Wall Street wanted, as operating margin collapsed to 1.4% and free cash flow turned negative for the first time in two years. Record volume, worse profit. That combination is usually the signature of a company choosing to spend, not of one losing its customers.

Record Deliveries, Worse Business

The distinction matters, because the crowd's reflex is to read an earnings miss as proof of weakness in the car business. The car business was the healthiest part of the quarter. The profit miss came from the decision to pour money into the future. Tesla spent $5.8 billion in a single quarter on AI and robotics, let operating expenses jump 47%, and told investors capital spending would top $25 billion for the year — with more to come after that. Elon Musk called it a "massive capex year" and said he wanted to spend "as fast as we can without it being too wasteful."

Now look at what that does to the number that actually governs a stock's value, which is not deliveries and not revenue but what the company earns on the capital it ties up. Return on invested capital was roughly 7% in fiscal 2024. By the first half of 2026 it had fallen to about 3.3%. The celebrated numerator kept climbing — more cars, more revenue, record quarters — while the denominator that converts those into shareholder value quietly deteriorated. That is the textbook wrong-metric trap: the market reads the rising headline and misses that the economics underneath got worse, because the company is converting a profitable cash engine into a capital-devouring ambition machine.

The SpaceX Culprit Points the Wrong Way

Now the other scapegoat. The comfortable story is that SpaceX's gigantic June IPO soaked up attention, wallets, and narrative oxygen, leaving Tesla to sulk. The finances suggest the opposite direction of causality.

SpaceX went public on June 12 at $135 a share in the largest IPO ever, raising tens of billions while selling only a sliver of itself — and Elon Musk cannot sell any of his shares for a year, so the founders' wallets were not part of the story either. But here is the detail that should stop the narrative cold: Tesla itself owned a piece of SpaceX, and that stake showed up in Tesla's own second-quarter profit as a roughly $1 billion unrealized gain — one of the cushions that made the reported quarter look better than the operating numbers did. In other words, the very IPO that headlines blamed for dragging Tesla down was standing inside Tesla's income statement propping the "earnings" up. Investors were told the two were enemies; the accounting says they were already sharing a household.

The linkage was always thinner than the phrase "SpaceX aftermath" suggests. When the IPO order book closed in June, Tesla shares actually jumped, and even then the analysis was hedged by noting the selling pressure was a story "if it were ever a thing in the first place." SpaceX shares have since fallen back below their IPO price and given back much of their first-day pop — which is precisely the kind of mark-to-market that could turn next quarter's cushion into next quarter's drag, without Tesla selling a single additional car.

The Rout Didn't Make Anything Cheap

Here is where the contrarian reading has to be honest about what the summer actually was. The stock did get knocked down: from a high near $499 to the mid-$300s. But a 27% fall does not automatically make an expensive stock cheap. At roughly $365, Tesla still traded near $1.4 trillion in market value against about 378 times trailing earnings, roughly 455 times forward earnings, and about 130 times EV-to-EBITDA — with free cash flow negative. The "slow summer" did not return the stock to reason. It returned an extraordinary valuation to merely very expensive.

Which leaves the honest version of the argument, and it is uncomfortable for both the buyers and the skeptics of the dip. The bulls are right that the miss was a funding decision rather than a demand failure — record volume proves the car engine still works. But the entire valuation now rests on the promise that robotaxis, robots, and energy repay the spending, not on the car business. And the evidence for that promise is still thin. The purpose-built Cybercab fleet that is meant to carry the story is small enough that analysts count it in the dozens of vehicles — estimates cluster around 25 to 50 units — against rivals' autonomous fleets that number in the thousands, and management itself called revenue from robotaxis and other AI projects negligible. Full Self-Driving subscriptions are real and growing — past 1.4 million — but at this valuation that is the appetizer being priced as the whole meal.

So the thing that would prove the crowd wrong — the failure condition for the summer-bear's case — is not one more bad headline. It is a quiet one: robotaxi and robot scale stays tiny while the drain on cash continues, and the next SpaceX mark-to-market flips from a cushion into a rebate. The company may win the transition and still hand shareholders nothing if the price was set for a future that had to arrive on schedule, funded by a car business that no longer returns its capital.

Everyone is right that Tesla had a hard summer. The market problem begins one step later: the rout was the cost of buying the promise, the SpaceX blame was pointed at the wrong house, and at 378 times earnings there is no such thing as a cheap dip. Being with the crowd on "buy the robot dip" protects a portfolio's excuses, not its returns.