Uber Autonomous – FOMO syndrome

Japanese investors have a bad case of FOMO syndrome.

• Fear of missing out (FOMO) is the only rational reason I can come up with for why Toyota and Denso would pour $667m into such an awful asset.
• I think that the only hope for this billion not to disappear into a black hole is for Uber to buy one of its plentiful and substantially superior, self-driving rivals.
• Uber has announced that it has raised $1bn from a group of largely Japanese investors by selling shares in its autonomous driving unit at a mind-melting valuation of $7.25bn.
• I say mind-melting because, based on the data that is available, I think that Uber’s autonomous driving unit is virtually worthless.
• Toyota and Denso are together putting in $667m with Softbank’s Vision Fund putting in the balance ($333m).
• The issues with Uber’s autonomous driving solution are legion but here are the top three:
  • First and foremost, performance: The available data points to an offering that was bad in 2016 which worsened materially in 2018.
  • In 2016, Uber trailed Waymo by a factor of 5,000x with roughly one disengagement for every mile driven (see here).
  • Uber has travelled backwards in the last 2 years as in 2018 (see here), Uber’s performance deteriorated by a factor of 3 with its vehicles suffering 3 disengagements for every mile driven.
  • Uber’s offering now trails Waymo by 31,570x.
  • Second, intellectual property: Events indicate that the quality of Uber’s IP in autonomous driving is very poor.
  • Uber was the subject of a 2017 court case where Waymo alleged that Uber stole some of its IP.
  • The case was settled out of court in 2018 with no blame attached to either side but it is obvious that Uber was on the losing side.
  • Uber paid Waymo $245m in Uber stock and promised not to use its hardware or its software in its autonomous products.
  • At the time, I was of the opinion that Waymo’s IP had not really helped Uber’s offering as it was by far the worst performing company that had been testing in California (see here).
  • However, it looks like the main reason for Uber’s substantial deterioration in performance is the removal of Waymo IP from its offering.
  • This strongly implies that Uber’s own IP in autonomous driving is far worse than I thought and maybe be almost worthless.
  • Software IP is the most valuable piece of any autonomous driving offering as most of the hardware is off-the-shelf and widely available.
  • Therefore, in any rational investment assessment where there are no revenues, software IP will be the foundation upon which any valuation argument is put forward.
  • How this IP is valued at $7.5bn in a complete mystery especially compared to its betters (see below).
  • Third, competition: There are many autonomous driving offerings being worked on today.
  • Some are part of large companies like Google, Baidu or BMW while others are standalone start-ups like Zoox or Pony.ai.
  • During 2018, 29 companies tested autonomous driving in California alone and all of them performed at least 3x better than Uber.
• Based on current performance and trajectory, one would have to conclude that if Uber was to operate an autonomous fleet of vehicles as a service it would be a disaster.
• Hence, something needs to change and I continue to believe that the only option is an acquisition.
• Therefore I make 2 conclusions:
  • First, FOMO: the reported valuation of $7.5bn makes no sense whatsoever.
  • Zoox which RFM has rated as No. 3 in autonomous driving raised money in July 2018 at a valuation of $3.2bn.
  • How IP that is demonstrably materially worse can be worth more than twice as much is nonsensical.
  • NIO, Yandex & pony.ai are all demonstrably superior to Uber’s offering but their autonomous driving assets are also valued at a fraction of $7.5bn.
  • Consequently, I can only conclude that it is a fear of missing out (FOMO) that has convinced the Japanese investors to pay far more than this asset is worth.
  • This speaks volumes about the power of Uber’s brand despite its recent difficulties.
  • They should have looked at the other both plentiful and superior offerings in my opinion.
  • Second, acquisition: I think that the only hope for Uber to have a credible autonomous driving offering is shutter what it has built and buy one of the other players.
  • Zoox and Nuro would be my top choices to look at assuming Uber can’t buy one of the Chinese players, but there are plenty more to look at in the middle of the pack.
  • Furthermore, with autonomous driving a long way out, there is the prospect of there being a round of fire sales as the mediocre end of the race runs out of money and fails to raise more.
• Acquisition is the only rational use of proceeds as doubling down on what is the worst offering available looks like it will be a disaster.