Tesla’s auto woes crash Elon Musk’s AI dreams

The electric-vehicle pioneer’s second-quarter numbers, reported late Tuesday, missed analysts’ expectations at most levels. The automotive gross-profit margin excluding regulatory credits—a closely watched measure of underlying profitability in the core vehicle business—came in at 14.6%, compared with consensus expectations of 16%, according to RBC’s calculation. The stock fell roughly 8% after hours.

Tesla offered financing deals during the quarter to make its vehicles more affordable. These boosted deliveries, which came in above expectations earlier this month, but at the expense of gross margin, as it now turns out. Further down the income statement, the hit to margins was exacerbated by high operating costs, but some of this will unwind. The company reported $622 million of restructuring charges for the quarter, reflecting mass layoffs following a terrible first quarter.

One bright spot was the energy business, which sells battery storage units and solar panels, where revenue doubled compared with the second quarter of last year. The “services and other" business, which includes used Tesla sales, parts, insurance and the charging network, also grew strongly, with revenue up 21%. Together these units brought in revenue of $5.6 billion, 22% of the company total and enough to help it unexpectedly report top-line growth.

While the services operation is closely related to the core automotive one, energy storage is a business with independent potential. At last, claims that Tesla isn’t just a car company have some grounding in the numbers, rather than merely in the expansive vision of chief executive Elon Musk.

Tesla’s second-quarter delivery update already made the strength of the energy business clear, so this could be one reason why investors have warmed to the stock lately. But the 35% rally since last month’s annual general meeting also seems connected to hopes for investments in artificial intelligence and autonomy. It was brought to an abrupt halt earlier this month when Bloomberg reported that Tesla was delaying a highly anticipated August robotaxi event that will likely tease details about its autonomy program. On Tuesday, Tesla confirmed the new date as Oct. 10.

How to factor businesses at radically different stages of technological development into one stock valuation is perhaps the essential problem for Tesla investors.

As Brad Cornell, emeritus professor of finance at UCLA’s Anderson School of Management, puts it, even as Tesla shows signs of turning into yet another mature car company, Musk has convinced investors it has “growth options" such as robotaxis and humanoid robots. Unlike the core business, these options are virtually impossible to pin firm values on. This helps explain why Tesla is such a divisive and volatile stock.

A UBS analysis earlier this month argued that the market has valued the company’s automotive business fairly consistently at between $200 billion and $300 billion over the past four years, with other initiatives accounting for the rest of its wildly fluctuating market value, which now stands at roughly $790 billion. The brokerage downgraded its recommendation to sell on the basis that the post-AGM rally had, at that point, taken the proportion of Tesla’s valuation represented by cars below its four-year average of 30%.

One might quibble with the details, but the basic insight seems sound: It is a good idea to unload Tesla shares when hopes for unproven businesses such as autonomy completely overshadow tangible fundamentals. Every quarter, after all, the company is forced to report actual numbers, bringing AI-crazed investors crashing back to earth.

Write to Stephen Wilmot at stephen.wilmot@wsj.com