Tesla leaves Uber plenty of road

That was about the only certainty that emerged from Tesla’s much-hyped “We, Robot” event Thursday night. The gathering was ostensibly for Tesla to unveil a robotaxi vehicle and related service, but it was light on details about the cars and included even less about a service. Tesla Chief Executive Elon Musk spent barely 20 minutes on stage (after arriving nearly an hour late).

Tesla shareholders were flummoxed. “That’s it? Disappointing lack of detail,” read the headline of a report Friday by the Morgan Stanley analyst Adam Jonas, a noted Tesla bull. Tesla’s shares slid nearly 9% on Friday. The shares had run up 45% since Musk announced in early April a plan to unveil a robotaxi this year.

Uber shares did the opposite, jumping nearly 11% by the close. That, too, can be seen as a correction to an overreaction. Musk’s announcement in April sent the ride-booking provider’s stock tumbling in the ensuing months as investors worried about a competitive threat that Uber had formally spelled out in its initial public offering prospectus in 2019.

Many analysts saw Tesla’s disappointing event as removing a final overhang on a company that has dramatically improved its operations and cash flow over the past couple of years. “Twenty-minute robotaxi event concludes six months of Uber event worry,” Justin Post of BofA Securities noted to clients. “Tesla’s toothless taxi is a best-case outcome for Uber,” John Colantuoni of Jefferies wrote in his report.

The risk isn’t zero. Musk plans to have Tesla’s robotaxi—dubbed the Cybercab—in production in a couple of years and to sell it for less than $30,000. The idea is that individual buyers would turn the vehicles into a business. “Someone who is an Uber or Lyft driver today can manage a fleet of cars,” Musk said Thursday. He also promised that the Cybercab would have a total operating cost of about 40 cents a mile, which BofA’s Post noted “would far undercut Uber’s driver costs ($2.00/mile+) and would drive industry pricing lower.”

But those new “fleet operators” would still need riders to fill their fleets and produce a return on their investment. The same logic has led other robotaxi operators to sign partnership deals with Uber. The most notable of those is with Waymo, owned by Google parent Alphabet, which is already offering commercial robotaxi service in San Francisco, Los Angeles and Phoenix but will use Uber’s app exclusively when it launches in Atlanta and Austin, Texas, next year.

View Full Image

Uber has a global base of 156 million monthly active platform customers, and its mobility business in the U.S. alone now handles more than $34 billion a year in gross bookings. Lyft, which operates mainly in the U.S., reported 23.7 million active riders at the end of the second quarter. Building a new robotaxi network from scratch would be a tall order, even if every existing Tesla owner signed on to such a service. Tesla has shipped about 6.7 million cars since 2015, according to FactSet data.

Musk, of course, actually needs to get those Cybercabs rolling first. And that also will be challenging—likely more so than he envisions. Building the cars is one thing, but Tesla also needs to be certified for full, unsupervised self-driving in the states where it plans to sell them.

That is typically a multiyear process, though Musk claimed Thursday that the company intends to have such driving available for its Model 3 and Y vehicles in California and Texas next year. In their report Friday, Bernstein’s analysts said, “We continue to struggle to see Tesla overcoming the technological and regulatory hurdles needed to leapfrog current level 4 robotaxis.”

Uber’s robotaxi road looks free of merging traffic for now.

Write to Dan Gallagher at dan.gallagher@wsj.com