Evaluating car companies’ claims about self-driving cars is getting harder, rather than easier. Now Tesla CEO Elon Musk is casting doubt on his own company’s progress.
Musk tweeted out Monday the latest versions of Tesla’s (ticker: TSLA) full self driving software–the best version of his company’s driver assistance features–is “actually not great.”
It’s a surprising–and possibly odd–revelation. Manufacturers are racing to build the best autonomous vehicles, but for people without advanced engineering degrees, judging who has the best systems has become nearly impossible. Autonomous-driving features are just too new and the systems are changing too rapidly.
The job for investors, right now, isn’t to dissect tweets. It is to understand exactly who is saying what about car makers’ competing autonomous-driving systems. Learning the bull and bear arguments is enough for now. Figuring out who is leading and what it means for the stocks of the car makers can come later.
The highly anticipated Tesla(ticker: TSLA) A.I. Day was supposed to shed light on the topic of autonomous driving. Instead, it opened a Pandora’s box of questions that need answering.
What investors really wanted to know was when drivers will be able to text and drive safely on highways without fear of getting a traffic ticket. That answer wasn’t forthcoming. Instead, investors were asked to digest the importance of computer- designed self-driving simulations, machine learning, and cloud-based A.I. training technology. The detail was arcane.
Tesla bulls, however, came away from the event blown away by Tesla’s A.I. prowess and the improvements to its autonomous-driving features. New Street Research analyst Pierre Ferragu wrote Monday that Tesla is 10 years ahead of the competition in “real-world” artificial intelligence. He believes Tesla’s autonomous- driving subscription sales can generate about $23,000 in gross profit per vehicle over a 10-year span. That is significantly more than the roughly $7,000 in gross profit Tesla makes from selling a car.
He rates Tesla stock at Buy and has a target of $900 for the stock price. Shares closed Monday at $706.30, up 3.8%.
Tesla bears, on the other hand, are convinced that Tesla’s claims about self- driving cars are overstated and that other companies are doing a better job developing self-driving technology.
The bears’ belief is partly rooted in two reports from the research firmGuidehouse, which publishes automated-driving “leaderboards.” Tesla ranked dead last in the 2020 and 2021 reports.
That might feel odd, given that Tesla CEO Elon Musk often talks about how advanced Tesla’s self-driving offerings are. But the Guidehouse rankings are partly based on vision systems and partnerships, two areas where Tesla doesn’t score well. The company doesn’t use laser-based radar, while other makers of self-driving cars do. And it doesn’t enter partnerships, doing everything in house.
It is possible that bulls and bears are both right and are talking past each other. Waymo, for instance, leads in the Guidehouse rankings. It has deployed fully autonomous robotaxis in Arizona. That is quite a feat.
Tesla’s systems aren’t good enough to be robotaxis. They qualify as level 2 autonomy, which means drivers must be engaged at all times. But the systems enabling the Waymo robotaxi can cost upwards of $100,000 per vehicle. That is too pricey for consumer vehicles.
In October 2020,Consumer Reports ranked the automated-driving systems available to, well, consumers. Overall, Tesla’s automated driver system ranked second to General Motors(GM). Tesla scored highest on capability and ease of use, but ranked poorly on keeping the driver engaged.
One reason for that low score might be that Tesla’s automated-driving systems aren’t “hands free” solutions. Drivers need to keep their hands on, or touching, the steering wheel to keep the system on. There are “hands free” systems that use interior cameras to make sure a driver’s eyes are on the road at all times.
Camera monitoring might be better than haptic steering-wheel feedback. That is a debate for car companies and regulators.
Practically speaking, hands free and hands required isn’t the basis for competitive differentiation. All the systems on offer today, hands free or not, require drivers to be engaged 100% of the time.
Autonomous driving studies and reports will proliferate in coming years. Each will have different criteria for judging what is best. Eventually, investors will probably have to test out autonomous-driving systems for themselves. That way they won’t fall victim to ratings criteria that aren’t explicitly defined.
Tesla stock is about flat so far in 2021. The S&P 500 and Dow Jones Industrial Average are up 19% and 15%, respectively.
Comments