Earlier this week auto semis maker Mobileye filed to go public. Again. Mobileye was public once, before Intel acquired them in 2017, and now that there is a new regime at Intel, they are spinning it off. Below, we have some initial thoughts on Mobileye’s prospects.
First, full credit to Mobileye, they have spent 20 years plugging away at this market and have made considerable progress. Auto semis, as we have noted often, are a difficult proposition. The customers are large and slow to make decisions. It takes years to get qualified as an auto supplier and years more for design wins to convert to actual revenue. Mobileye has put in the time and has a clear first-move advantage as a result.
We also appreciate Mobileye’s product strategy. Auto semis are appealing today because of the opportunity for assisted driving (ADAS) or Auto Processing Units (APUs) or as Mobileye calls them Electronic Control Units (ECU). Autonomy is a huge opportunity but remains years away and in the interim the action will center on ADAS systems. Mobileye is well positioned for both.
We also think the spin-off makes a lot of sense. Intel does not have a great track record with acquisitions and there was always the risk that Mobileye would stagnate the way Altera has inside the far larger Intel. Whether Intel will be able to recoup the $15 billion it paid for Mobileye remains to be seen. Mobileye had $1.4 billion in revenue in 2021, which would require a 10x EV/Revenue multiple a feat which seems unlikely in the current market. On the other hand, Intel will still have a stake and it seems highly possible that Mobileye will grow into the valuation, to say nothing of the fact that the company owes Intel $3.5 billion, to be paid back over time.
That all being said, we have a few questions or areas to keep an eye on.
First, this market is highly nascent. As much as all the participants like to throw around multi-billion dollar market estimates (TAM), Mobileye’s $1.4 billion revenue is likely the bulk of it today. More to the point the ultimate solution remains very much a work in progress. Mobileye lists dozens of competitors including most of its current customers. Everyone is experimenting right now and it will likely be several years before the industry reaches any kind of consensus of who is building what pieces of the system, and thus who captures what share of the value pool. Mobileye is first, but the whole thing could move in an entirely different direction.
This neatly ties into our second area to watch. During their Auto Analyst Day last week, Qualcomm repeated their mantra over and over again that they are “open”. They want to build chips that will fit into any design. By contrast, Mobileye supplies “solutions”, of which semis are just one part, but which also include sensors, modules and a whole range of other software and hardware components. This likely has played a big role in speeding their products to market, but many automakers still want to have a hand in designing their solutions. Qualcomm’s progress at the very least demonstrates their approach resonates with many customers. This is one of those critical questions upon which the industry has not yet settled. We could see this cutting either way – the auto makers may very well try to design their own solution but not be able to achieve that and end up forced to work with a more complete provider. Or not. We cannot fault Mobileye for their approach.
All of this does raise our third concern which is focus. Mobileye rightly recognizes that the ultimate industry structure is protean and they are taking steps to capture as much of that value pool as they can. So they are not only building auto modules, but are taking steps to actually participate in the business of building robo-taxis and providing autos-as-a-service. This is where we get a little nervous. It is never a good look to compete with your own customers, and while Mobileye is not exactly doing that, their model is close enough that we imagine some may see the company going all the way into production.
A bigger concern is the risk that the company is doing too much. They have a lot going on, and while Mobileye is not a start-up it is still not that large, with a lot of irons in the fire. We worry that company risks losing focus. Remember, in the past five years the company sold itself, went through a full-blown integration process, is now preparing for an IPO, and almost certainly will then go out and look for strategic investment. That is a massive cognitive overhead for any executive team. While we have no reason to think Mobileye’s executive team is not up to the task, it begs the question are they doing too much?
Overall, we think there is a lot to like about Mobileye – the right products, in the right market at the right time, with a strong track record of execution. Depending on when they list and where the valuation comes in we could see Mobileye being an attractive stock. On the other hand, its long term prospects remain a very open question.
Mobileeye’s “Blackbox” solution is being increasingly ditched by new generation of OEMs, especially in China. Their desire to own the whole ADAS stack will risk alienating them from future high end car OEMs who have in-house ADAS development capabilities. I think their TAM will shrink in future.
Good point. I didn’t get into the China EV market, but I agree with you the China EV OEMs are ‘all’ designing their own chips.
We’ve seen this approach in the much-less-complicated display industry. Plastic Logic developed the first organic thin film transistors but couldn’t get any traction with existing display manufacturers. So they went vertical – all the way to producing a tablet-like device.
Producing a device got them so far from their core expertise that their organic TFT value proposition was diluted. They exist today after more than 20 years of operation, but continue to seek market opportunity capable of more than subsistence-level revenues.
This reads like the beginning of a really bleak novel. Cormac McCarthy does technology.