Elon Musk is all over the news these days for reasons that have nothing to do with investing.
But before Musk decided to go all in on DOGE (the Department of Governmental Efficiency), he made a few comments on the last Tesla (TSLA – NASDAQ) call about self-driving cars that made me take notice.
Musk said that Tesla’s future, and in his opinion everyone’s future, lies in two innovations: autonomous vehicles and autonomous humanoid robots.
The robot angle, I’m not going to touch that one here. It could be HUGE (imagine robots that do your dishes, clean your house, even help prep dinner), but it is years away.
What I’m interested in are autonomous vehicles – in other words self-driving cars.
If we believe Musk, Tesla’s full full self driving (FSD) robo-taxi is coming to Austin, Texas in the summer of this year.
Of course, Musk promises don’t always pan out. On the call he even called himself “the boy who cried wolf several times”.
But he followed that up with a caveat: saying this time, “there is a damn wolf and… It can drive you. It’s a self-driving wolf.”
MAKING CENTS OF SELF-DRIVING
Taking a step back, FSD cars just makes sense.
A commute where I can read a book, the family napping on vacation to arrive rested at the destination, even just getting out at the front door while the car finds its own parking spot.
Now I know, I know, it also sounds scary. It is, because it’s a brand-new thing. But with AI improving at such a rapid (alarming?) rate, I think it won’t be long before cars are more safely driving themselves than we ever are.
That makes it a big opportunity for someone – and maybe not just Tesla.
Tesla is no doubt the leader here. Yet given how Tesla has morphed into a political proxy on the Donald Trump administration, I’m just not sure I want to go there.
Fortunately, there are other beneficiaries.
SELF-EDUCATING ON SELF-DRIVING
My introduction to self-driving cars came from a deep-dive into what became a portfolio holding, Arteris (AIP – NASDAQ).
Arteris develops semiconductor IP.
What they do is akin to an auto-OEM that only designs transmissions. The OEM doesn’t design the whole car, and Arteris doesn’t design the whole chip, but they are a key piece of the final product (in this case the semiconductor).
Arteris specializes in the interconnection interface between chips. One big end-market for this is to power advanced driver assist systems (ADAS) that we have today.
ADAS is the stepping stone to fully autonomous driving and you probably use it every day if you have a car that beeps when another car passes behind you in a parking lot, or if a car is close by on the freeway when you begin to change lanes.
Arteris’s biggest customer is Mobileye (MBLY – NASDAQ). Mobileye uses Arteris IP in their ADAS EyeQ6 and EyeQ7 chips.They will (soon!) use the IP to power their next-gen autonomous driving semiconductors.
Lately I started looking at Mobileye, and what I was surprised by was just how dominant they are. Mobileye has somewhere around 70% of the ADAS market, including 60% in China.
You might think that Mobileye would carry with it an eye-popping capitalization. After all, Tesla is a trillion-dollar company.
Yet Mobileye has a modest market cap of about $14B. It is roughly the same price Intel paid to take Mobileye over in 2017.
That market cap took a hit in 2024. Based on the chart, you’d never guess this company could be on the verge of something big.
What happened? The short story: Mobileye had some revenue declines. In 2024, full year revenue was $1.65B, which was down from $2.1B in 2023. Ouch.
The longer story is why. I see a couple reasons. Neither is about the long game.
SUPPLY CHAINS ARE NOT YOUR FRIEND
Mobileye identified they had an inventory problem in late-2023. They warned that their 2024 numbers were going to be impacted in early January. That coincided with the first big drop in the stock price.
The inventory issue was partly COVID, partly EV’s.
After COVID wreaked havoc on supply, auto-makers decided this wasn’t going to happen again. They began to over-order on key components.
One key component that REALLY held up car production was chip availability. As a result, many, many chips were ordered. Too many.
When auto demand slowed in 2024, the OEM’s had a problem and cut back on orders.
Exacerbating this was Mobileye’s dependence on EV’s and China.
Mobileye’s advanced hands-off solution is called SuperVision. After going into last year SuperVision was only on EV’s and largely in China.
It may surprise you to know that China has kept up with North America on developing assisted driving technology.
Remove Tesla from the picture and China is actually leading in next-gen “hands-off” ADAS.
Mobileye’s early wins were with Chinese OEMs like SAIC Motor Corporation, FAW Group, Geely Auto Group, and NIO.
Those wins meant early revenue. But being reliant on China is a blessing and a curse.
Mobileye was vulnerable to replacement as China learned the technology and followed the playbook of taking development in-house.
Which is what began to play out in 2023 and accelerated in 2024.
In fact, Mobileye’s recent 2025 guidance took a very conservative view of China, including some more moves to in-house production.
DON’T LOOK BACK, IT’S ALL RIGHT
That was 2024. It explains the stock price going from the $40’s to the teens.
What it misses is two-fold:
A. Mobileye has a 70% share in ADAS with their EyeQ technology stack
B. Mobileye is beginning to win next-gen deals for SuperVision, for Chauffeur (eyes-off) and for Drive (no driver)
Why is Mobileye winning? There isn’t a lot of competition.
Mobileye provides an end-to-end solution. They give you the software, the cameras, radar, and LiDAR and the hardware to run it.
Both Qualcomm (QCOM – NASDAQ) and Nvidia (NVDA – NASDAQ) have chip solutions for ADAS and higher-level autonomous driving.
These are “open” systems. You buy the chips from Qualcomm or Nvidia but it is up to you to program them, tie them into cameras, radar and/or LiDAR.There has been a debate going on about whether automakers would go “open” or “closed”, which is what Mobileye’s full stack is. Would the big Western automakers follow China and take their designs in-house?
Mobileye allayed those fears on their Q4 call. When asked directly about it, CEO Amnon Shashua said, “the third bucket in terms of in-house development, we don’t see a trend there. We don’t see anything that is serious about in-house such advanced product”.
This isn’t surprising. Some automakers have tried and failed. One well-known example is Nissan, which tried to solve the L4, L5 (the highest levels of autonomous driving) problem and realized that they could not make the software work.
Mobileye’s competition for a closed system solution are only doing it for their own fleets: Tesla and Google (GOOG -NASDAQ) owned Waymo.
KEEPING UP WITH TESLA
And here is the second big complaint about Mobileye. The Tesla solution is better.
Each of Tesla, Waymo and Mobileye use very different approaches to accomplish FSD with varying reliance on cameras and radar and LiDAR.
Tesla is camera only. Musk believes that cameras alone can solve autonomy. In his usual, first principle, kind of statement, Elon said thathumans have eyes and a brain, why should cars need LiDAR?
Waymo, on the other hand, is LiDAR based. While Waymo has cameras and radar, their primary sensor for perception is LiDAR. LiDAR stands for light detection and ranging – it is a sensing technology that uses lasers to detect distances. It is excellent at depth perception but comes with a cost.
Mobileye is camera centric. The primary detection is via camera, but their approach includes two layers of redundancy – they layer imaging radar, which can create high-resolution, detailed 3D images of the surroundings and use LiDAR as a 3rd tier fail safe.
There are plenty of arguments to be made in favor of each. Tesla’s solution is the most elegant. It does it all with only 8 cameras, less than Mobileye’s camera/radar/LiDAR design.
To make it work Tesla relies heavily on AI learning which they call their neural network.
That would be hard for an automaker to replicate in-house. And it’s unlikely that Tesla is going to be selling this to Volkswagen or Ford, their competition.
Does Tesla matter? If Mobileye has a solution that works and is expandable to multiple OEM vendors, I suspect they’ll win their share of business.
CRUNCH TIME
Mobileye’s reliance on EV’s may be true for cars in production, but it isn’t true for cars in development. Nearly half of their advanced system deals are for internal combustion engines.
What Tesla’s shot across the bow has done is strike a fire under some of the larger OEMs.
The broader issue with Mobileye is that the traditional OEM’s – Ford, Volkswagon, Toyota and so on – have been slow to adopt the next generation of autonomous driving.
If Elon Musk is good at one thing it is charging ahead. The rest of the auto industry, not so much.
While Mobileye has had the product ready to go, the automakers simply haven’t been there.
That is changing.
On their Q4 call Shashua said, “the work of Tesla is really creating a sense of urgency with our OEMs. We still believe that 2027 is really the right timing for introduction of these systems”.
At their Investor Day in December Mobileye compared their OEM engagements at YE2023 to YE2024 and showed substantial progress.
Bank of America, which is anything but a Mobileye bull (they believe Tesla will be the big FSD winner over the next 15-20 years), admitted in a recent note that Mobileye
“appears close to securing a SuperVision/Chauffeur contract with a Japanese OEM and a Surround ADAS contract with a European OEM. It also appears close to a Surround ADAS contract with a US-based OEM.”
Mobileye has already announced deals with Volkswagen (which will start production in 2026) and say others will start producing self-driving capabilities until 2027-28.
STOP LOOKING OUT THE REAR-VIEW MIRROR
Let’s get back to what we own. Arteris has always been a waiting game with the payoff a few years out.
We acknowledged this when we bought the stock. But we also knew that the market looks ahead.
Sure enough, Arteris moved on more design wins and greater certainly about future revenue. The stock has appreciated by some 50%+ since our initial buy.
I’m thinking that same dynamic could play out with Mobileye.
The 2025 numbers are not going to be great. Though they low-balled guidance so blatantly that they could do a nice beat. Yet, new deals could set the stage for 2027-28 growth and that might be enough for the market.
On the Q4 call Mobileye guided weaker than the Street would have hoped. Yet, the stock went up 10%+ on the day.
That had nothing to do with 2025. It was because they said they are getting close on deals and are not worried about western automakers going in-house.
Their new products like Supervision carry price tags 10x higher than legacy ADAS stacks that they have been selling in the past.
If they start winning deals for these stacks, revenue projections for 2027+ will quickly rise.
Back in March of last year, CCO Dan Galves said that their bookings were translating into future annual revenue of $7B, versus the run rate of a little under $2B right now.
Bank of America, who again I want to stress is anything but a bull, is projecting even larger numbers if you look 5+ years out.
I’m not going to tell you Mobileye is “cheap”. The midpoint of their 2025 operating income guidance is $217M, while the market cap is about $13B. That’s a hefty P/E.
Still, when I look around at what the market is pricing “growth” stocks, especially those with unique semiconductor tech powering next-gen solutions, I can’t help but think that the long-term potential is not in the price.
Everyone is focused on Mobileye’s recent failures and short-term headwinds. Yet we are in a market that has been extremely forgiving to growth stocks as long as “the story” is intact (see: Tesla).
Well, times change. If Mobileye keeps its lead and wins the next-gen autonomous business, I think the stock has higher to go. And you don’t have to bring any politics into it.
