Tesla Is Still Overvalued, Don’t Bet On Its Robotaxi

Summary:

• Tesla, Inc. is still overvalued, despite losing more than 30% of its value this year.
• CEO Elon Musk’s tweet that Tesla will introduce a robotaxi business on August 8th may be supporting this overvaluation.
• In my opinion, investors should realize that a robotaxi business is not feasible for Tesla, primarily because it lacks the necessary autonomous driving platform for this market.
• In addition, Tesla faces increasing competition, including from Tesla’s used vehicles.
• Eventually, the valuation gap between Tesla and others like Ford and General Motors should narrow.

Tesla, Inc. (NASDAQ:TSLA) stock, a significant investment for many, has experienced a substantial decline of over 30% since the start of the year. Despite this, it remains overvalued at a forward P/E ratio of 56X. The company’s recent announcement of a disappointing quarter and the decision to lay off 10% of its employees further adds to the concern.

Much of Tesla’s overvaluation seems to stem from the anticipation of a groundbreaking robotaxi business. However, this belief is not without its risks. Tesla CEO Elon Musk’s tweet on April 5th at 4:49 PM, announcing the introduction of Tesla’s robotaxi on August 8th, has bolstered this valuation based on the promise of improved Full Self-Drive (“FSD”) technology at a time when Tesla is grappling with a slowing electric vehicle (“EV”) market and heightened competition.

According to an article in Investors Business Daily, or IBD, Tesla will focus on a lower-cost robotaxi instead of a Model 2.

It began when Reuters reported Friday the EV giant has canceled its long promised next-generation $25,000 vehicle, choosing to focus on developing its self-driving robotaxi platform. Reuters, citing three anonymous sources and internal messages, wrote that Tesla no longer plans to make its low-cost entry-level vehicle, the Model 2, and will continue developing its robotaxi on the “same small-vehicle platform.”

A source told Reuters that Chief Executive Elon Musk is going “all in” on robotaxi. Tesla stock sank on the report, hitting its 2024 low of 160.51 intraday Friday. However, TSLA began to pare losses after Musk took to X, formerly Twitter, alleging that Reuters “is lying (again).”

This is Tesla’s big strategic bet, as Electrek reported:

However, we reported yesterday that we were able to confirm that CEO Elon Musk postponed Tesla’s cheaper model in favor of Robotaxi.

That is a huge bet as Tesla now absolutely needs to solve self-driving in order to keep growing at any significant pace, as it would not have a car program geared toward growth for years otherwise.

The autonomous ride-hailing market (robotaxi) has enormous potential. However, as I’ll explain, I believe it’s not feasible for Tesla, and investors should only be hoping for something substantial in that August 8th announcement. Otherwise, investors may support a high valuation by counting on Tesla to grow its revenue by increasing its market share, and I’ll also explain why that is unlikely.

To understand the potential opportunities for autonomous vehicles, you need to know the different types of autonomous driving technology platforms, as each addresses a different market.

Four Different Autonomous Vehicle Technology Platforms

In my book Autonomous Vehicles: Opportunities, Strategies, and Disruptions, I define four different types of autonomous driving technology platforms. Each platform is unique and designed for a specific market. There are four different autonomous-driving technology platforms:

Semi-Autonomous Driving

Semi-autonomous driving enables the driver to turn over control to the vehicle for specific tasks, while requiring the driver to maintain awareness by periodically touching the steering wheel or monitoring the driver’s eyes on the road. Semi-autonomous driving technology has evolved from ADAS systems with automatic lane-centering, adaptive cruise control, and automatic braking.

This technology platform primarily uses multiple cameras and radar sensors that provide input into an onboard computer to control the vehicle. It is mainly used for highway driving but also works on many other road systems. Semi-autonomous vehicles are limited to driving forward only because they generally cannot make turns.

This is an SAE Level 2 classification, where Tesla’s autopilot and the current FSD are classified. Tesla FSD has the potential to be more advanced but will still require a driver to be present, as will be explained.

Semi-autonomous driving technology is already available in millions of vehicles today and is used routinely to drive millions of miles daily on highways autonomously. I’ve used it for more than 10,000 miles over the last four years with a Tesla and Mercedes.

Sufficiently Autonomous Driving

Sufficiently autonomous driving enables a vehicle to drive by itself without any driver present but only on predefined routes and locations. It is limited to specific geo-fenced areas defined by the company managing a fleet of vehicles for that geo-fenced area. The areas covered are expanding rapidly.

The technology platform for sufficiently autonomous driving uses Lidar in addition to cameras and incorporates high-definition maps. Together, these position the autonomous vehicle precisely on the HD map, providing critical information such as traffic signs, restrictions, and turning radius. It is the safest way to implement autonomous driving, at least initially.

Sufficiently autonomous driving is not suited for privately owned vehicles because of the limited driving areas. This technology platform is focused on autonomous ride-hailing services (robotaxis), autonomous delivery, and autonomous long-haul trucking. These are the three most significant market opportunities for autonomous vehicles. Limiting autonomous vehicles to a geo-fenced area is acceptable in these markets. Autonomous ride-hailing services and autonomous delivery limit customers to an area, but continue expanding that area. Autonomous long-haul trucking drives back and forth on highways on their geo-fenced routes from and to special depots located just off the highway.

Fleets based in specific metropolitan areas and managed by fleet management staff will serve all these markets. This is a new business model, as will be explained later.

Substantially Autonomous Driving

Semi-autonomous vehicles are on a path to evolve to a point where they can operate more independently. They will be able to stop at traffic lights and signs and make left- and right-hand turns independently. Possibly, more than 90% of the driving could be done autonomously with substantially autonomous driving. However, the vehicle cannot drive autonomously everywhere, and a driver will still be required, even if they don’t need to take control immediately. This is classified as Level 3 in the SAE classification scheme.

Tesla’s FSD has the potential to become substantially autonomous. It can stop at traffic lights and stop signs and turn corners. But, as I will explain later, it will still need a driver present.

Here is the critical difference to understand: Sufficiently-autonomous vehicles are limited to constrained areas but can function without a driver. Substantially autonomous vehicles require a driver on board. Substantially autonomous driving will be a feature in personal vehicles, not in the three markets mentioned.

Fully-Autonomous Driving

Fully autonomous driving equates to SAE Level 5 and enables driving anywhere at any time without a driver in the vehicle. Currently, there is no viable fully autonomous driving technology, and it is not likely for a very long time.

This is what Tesla, Inc. (TSLA) FSD aspires to be, but it is doubtful. Being able to autonomously drive a vehicle anywhere at any time on dirt roads, alleys, complex, park in between cars in a driveway, detours, on boat docks, on fields for even traffic circless, etc., is impossible for Tesla or anyone. It may become possible with a complete smart infrastructure and high-definition maps of every place, but this will take over a decade. Indeed, it will require different technology from FSD.

A Tesla Robotaxi Is Not Feasible

Investors betting on the Tesla Network and a Tesla robotaxi are in for a rude awakening. This is not feasible for several reasons. The first is technical: Tesla does not have the right autonomous driving technology platform. It requires unconstrained, fully autonomous driving, which isn’t working. On top of this, Tesla doesn’t have the necessary business model and hasn’t addressed the daunting task of obtaining regulatory approvals throughout the country for an unconstrained service.

Tesla doesn’t have the right ARS/Robotaxi technology platform

Although Tesla wants its camera-based FSD platform to be fully autonomous, it’s not. It’s an advanced semi-autonomous and potentially substantially autonomous system, but it always requires a driver in the vehicle. Here is a simple example of the difference between a camera-based FSD and an HD-map-based platform. Using Navigate on Autopilot on a highway, the Tesla knows that it needs to turn off an upcoming exit, but it doesn’t know precisely when the exit is coming, so it needs to wait until the exit line appears on the highway. Then, it must slow down abruptly from 75 MPH to turn off the highway. With an HD-map-based system, the AV knows precisely how far it is from the exit and can start slowing before reaching it. This is the way a human driver acts.

Tesla has and continues to improve FSD software, but these improvements can’t fix all the limitations of a camera-based system. I’ve driven a Tesla with FSD beta and encountered the following types of problems, which are limitations in the camera-based technology:

1. Turning left on an undivided highway with three approaching lanes, the cameras only identified two oncoming lanes because no vehicles were in the third lane. There were no lane markings on the undivided highway to guide a turn, so my Tesla turned into the third lane of oncoming traffic. An HD-map-based autonomous driving platform would have defined the three lanes in the map and specified the appropriate left-turn radius.
2. A Tesla’s cameras can identify stop signs as long as they are not blocked or out of the camera view; when they are, it cannot see them. When I drove with FSD in a seaside town, there was an intersection where the stop sign was placed to the far right, which was difficult to see. My Tesla couldn’t see the stop sign and repeatedly tried to travel directly into the intersection without stopping. Again, an HD-map-based system would have the stop sign on the map.
3. Right-turn-on-red signs are not a standard format and can sometimes be complicated. One example is “Right Turn on Red except for Mon-Fri from 8-11 AM.” The Tesla cannot read and understand it.
4. Multi-lane traffic circles with no lane markings are almost impossible for any AV. HD-map-based systems will most likely avoid them and designate a longer route around them. Tesla FSD doesn’t have these maps and can’t do that, which could create chaos in the traffic circle.
5. Some winding roads have posted speed limits that are too fast to maneuver safely through some of the curves. For example, one road has a 45 MPH speed limit, but a human driver realizes that the car needs to slow down a little to go through the curve safely. My Tesla didn’t know to slow down and stayed at 45 MPH through the curve until it went off the road. Here again, an HD map would indicate the need to slow down at that point on the map.

These examples illustrate the limitations of a camera-based autonomous driving system, and they cannot likely be overcome with software improvements and miles-driven experience. This limitation becomes more critical because Tesla doesn’t have any road limitations.

Tesla’s FSD is unrestricted

Autonomous Ride-Hailing Services (Robotaxis) are restricted to specific geo-fenced areas and routes. Their HD maps prevent them from driving outside an approved geo-fenced area, avoiding roads they cannot drive on. This is an essential element in making them viable. Initially, these areas and routes are limited, but still commercially viable. Over time, the areas, and routes will be continually expanded.

By definition, without HD maps or something similar, Tesla’s FSD cannot be restricted to safe areas. There are intersections with unmarked lanes that AVs can’t maneuver, tight roads and alleys, construction detours, traffic circles, etc., that AVs need to avoid, at least for now. Tesla FSD vehicles will attempt to drive these and cause problems. They will get stuck in alleyways and tight roads, stop at dangerous locations to try to pick up a passenger, and, most critically, cause accidents.

Unrestricted, Tesla expects its robotaxi vehicles to drive autonomously in every state and municipality without any approval. This is impossible.

Elon Musk has repeatedly portrayed Tesla vehicles on the Tesla Network as individually owned vehicles that leave their owner’s garage and drive on their own to any location requested by someone paying for a trip in the vehicle. This is simply unrealistic. Autonomous ride-hailing must be done in fleets managed in defined metropolitan areas, which is a different business model.

ARS is a very different business model

Autonomous ride-hailing services (or robotaxis) are a very different business model from Tesla. It requires fleets of autonomous vehicles in each metropolitan area it serves, managed by a staff of people for that area. This fleet-based approach is critical for several reasons:

• Each fleet vehicle has specific HD maps for that metropolitan area, defining its operational domain, geo-fenced area, and routes. The fleet’s management team updates these maps regularly, changing routes based on construction and other obstacles.
• The fleet of vehicles is cleaned, powered, and tested at the fleet garage to ensure that they are appropriately operational.
• The local management team works with municipal authorities to obtain all required approvals and make changes as needed. These changes could include avoiding specific portions of roads like school zones during certain hours, being notified to avoid particular areas when an emergency is being handled, eliminating or changing specific drop-off/pick-up locations to avoid traffic issues, etc.
• The fleet management team sets pricing for trips in that municipality, determines the number of AVs needed for the fleet, and promotes the autonomous ride-hailing service locally.
• Most importantly, the local management team intervenes when necessary. They can communicate directly to passengers with questions, issues, or problems. They can take over the remote control of the vehicle if it is in a situation it cannot handle. They can immediately dispatch staff to a vehicle site when an issue or accident needs to be managed.

This business model is critical for ARS and not just optional. Tesla doesn’t have any fleet operations like this. Even if it somehow launched a new AV with the correct technology to provide ARS (robotaxi) services, it would take years to implement this infrastructure.

Tesla would require extensive approvals for a robotaxi service

Even if Tesla developed the necessary autonomous driving technology to provide and establish a new business model, it would need extensive approval from local and state governments. Because the Tesla FSD robotaxi network aspirations are unconstrained, this would mean approvals in every municipality and state in the country, which is virtually impossible.

Cruise and Waymo have shown how difficult and time-consuming it can be to get a few approvals. It would take Tesla several years to obtain the necessary approvals for even a few municipalities.

Tesla hasn’t initiated any contact with regulatory agencies yet, or that would be disclosed publicly. All companies testing autonomous vehicles in California must file annual reports of the number of vehicles, miles driven, and disengagements. Tesla did not file the required reports for 2023, possibly because it didn’t do autonomous driving there.

There are already significant competitors in the market for autonomous ride-hailing services.

If Tesla could enter this market, it would face existing competition. Waym0 has already established autonomous ride-hailing in San Francisco and the peninsular, Phoenix, and is now entering Los Angeles. GM Cruise was successful in a dozen markets before it suspended operation, but it is now getting back more carefully. Uber and Lyft are doing some testing and will need to remain competitive with autonomous ride-hailing (robotaxi). As I’ve mentioned in previous articles, the economics of autonomous ride-hailing will cannibalize ride-sharing with much lower costs.

In Addition, Tesla Will Lose Market Share In Its Current EV Business

If you discount any impact from a robotaxi in Tesla’s valuation, you still must consider that it will probably lose market share over the next few years. While some of this could be offset by the new Cybertruck and potential Redwood crossover SUV, there are several reasons to expect a market share decline.

Tesla’s biggest competitor will be — Tesla

Since its inception, Tesla has enjoyed the unique advantage of not having to compete against itself, with customers buying used Tesla cars instead of new ones. Until recently, you could only buy a new Tesla. But this is changing. There are now millions of Tesla owners, and increasingly, many of these will resell their cars and create a growing used car Tesla competitor. I did this after having a Model 3 for 3 1/2 years.

For example, this could happen some years in the future: Customers could buy 2 million Tesla cars, but 500,000 could be used, so Tesla would only sell 1.5 million. Maybe people will buy 2.5 million, but 1 million may be used. Investors need to be prepared for this dynamic to affect Tesla sales.

Related to this is the challenge of upgrading its previous models while simultaneously building new ones. Again, avoiding this was a luxury that Tesla enjoyed, but now it is getting precarious. For example, I would have bought a new, updated Model 3, but Tesla still needed to update it, so I bought something else with an updated design and system. This will increase Tesla’s required investment to both update previous versions and develop new ones. And if it doesn’t invest in updates, customers will have little reason to buy newer Tesla vehicles.

Other auto manufacturers are launching many new competitive EV models

There was a time when you had to buy a Tesla if you wanted an EV. Then, Tesla was a better choice among a handful of alternatives. However, there is now an increasing variety of EV alternatives. And there are some significant reasons why people will buy another EV over a Tesla:

• Many mainstream customers (after the early adopters) prefer a particular car brand they have had for years. It’s just more comfortable for them and what they have always done. Increasingly, they can buy an EV version of that model.
• Similarly, many mainstream customers prefer to buy from the dealer and salesperson they have trusted for years. They know the people, and the service is local. For them, ordering a $60,000 car online and not knowing whom to call for service is intimidating.
• There is a much more comprehensive range of other EVs available. The number of alternative models across many manufacturers is many times more than what Tesla offers, and this number is increasing with new EV releases. People have varying preferences, and Tesla won’t be able to satisfy most EV customers.

While Tesla may continue to have the largest market share in the U.S. for a while, this share will decline due to the vast array offered by competitors.

Some competitors are catching up with Tesla’s FSD

Tesla’s FSD, even though it still requires a driver in the vehicle, could be an important competitive advantage. However, some competitors are catching up and gaining an advantage.

General Motors Company’s (GM) Super Cruise uses Lidar and cameras to position the vehicle on high-detail maps containing much more detail than navigation systems. These maps contain details on lane designations (even if unmarked), traffic signals and stop signs, turning radius, and speed requirements. Super Cruise restricts operation to the detailed mapped areas; only certain GM vehicles have this capability. Depending on the vehicle, 20,000 to 40,000 miles of roads are available. GM Super Cruise is an adaptation of sufficiently autonomous driving and could have advantages over Tesla in substantially autonomous driving.

Mercedes-Benz Group AG (OTCPK:MBGAF) Drive Pilot is a similar system that uses Lidar and detailed maps to position the vehicle. This is in SAE Level 3 classified EQS sedans approved in Nevada and California. It could also turn out to be a superior substantially autonomous system to Tesla FSD.

So, even if FSD is improved, sustainable competitive advantages may not continue. Mercedes and General Motors Company’s decision to adopt a sufficiently autonomous platform variation may prove superior.

Finally, there is the competition from China

Chinese EV competition for Tesla in China and eventually from Chinese car manufacturers in the U.S. cannot be ignored. I won’t attempt to delve into this competition or estimate the extent of competition in the Chinese market, or when they will make inroads into the U.S. market. However, it is essential to consider when valuing Tesla.

Valuation considerations

So, what is an appropriate valuation for Tesla? Is it a diversified technology company or an auto manufacturer? In my opinion, it was once an innovative technology company, one of the first to make electric vehicles viable and promising autonomous vehicles. Over time, EVs become more mainstream, and Tesla can’t deliver on its autonomous vehicle promise. So, it should be characterized as more of a car manufacturer. For that matter, Ford and GM could be classified less as just auto manufacturers, especially GM, which already has an autonomous ride-hailing business (being restarted), and Super Cruise potentially becoming better than FSD.

You don’t need to be precise because it is much lower than today.

Consider this comparison of Tesla, Ford, and GM financial metrics:

	TSLA	F	GM
P/E Ratio	58X	6.5X	4.7X
Mkt. Cap. (B)	$514	$49	$49
2023 Revenue (B)	$97.0	$166.0	$157.0
2023 Net Inc. (B)	$14.9	$4.4	$10.0

Click to enlarge

Do these different valuations really make any sense? Ford and GM are much bigger, yet Tesla is valued at 5X their combined valuation. Yes, Tesla is currently more profitable, but how long can that last? Profitability is also considered in the P/E multiple.

Eventually, these valuations should approach parity around a tighter range of P/E multiples. The average P/E ratio for the S&P 500 (SP500) is around 19X-20X forward earnings. In this comparison, Tesla is overvalued, and Ford and GM are undervalued. For the sake of discussion, Ford and GM have the potential to double or even triple their current P/E multiple and stock price and still be under the S&P 500 P/E multiple.

Tesla’s P/E multiple of 58X is difficult to justify. It would require a 3X increase in net income, which is most likely unrealistic without a huge robotaxi business, especially while market share is declining. Let’s say it still maintains a higher P/E multiple than others for whatever reason, maybe brand loyalty or fan investors. Perhaps it could be 30X, which would still be several times higher than GM or Ford. This would still reduce the valuation by half.

In conclusion, Tesla’s valuation as a rapidly growing technology company is likely based on the hope of a substantial robotaxi business, which is not feasible. Its valuation will likely decline over the remainder of the year into next year when this becomes more apparent.

Risks to This Thesis

Of course, there are some risks to my negative thesis on Tesla’s Valuation, including:

• Tesla’s Energy Generation and Storage Business, which Tesla reported at $6 billion in revenue in 2023 out of its total of $96 billion, could grow rapidly and justify its valuation. Perhaps the Tesla Optimus robot could become a viable product in a large market.
• Tesla has many fans among its shareholders who may want to hold on to their investment no matter what.
• Tesla could unveil a new robotaxi vehicle with the right technology platform, achieve rapid government approvals more rapidly than others, build a new business model, and overcome competitors already in the market.

---

Analyst’s Disclosure: I/we have a beneficial long position in the shares of GM, F either through stock ownership, options, or other derivatives. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Seeking Alpha’s Disclosure: Past performance is no guarantee of future results. No recommendation or advice is being given as to whether any investment is suitable for a particular investor. Any views or opinions expressed above may not reflect those of Seeking Alpha as a whole. Seeking Alpha is not a licensed securities dealer, broker or US investment adviser or investment bank. Our analysts are third party authors that include both professional investors and individual investors who may not be licensed or certified by any institute or regulatory body.

---