Making sense out of Intel’s recent Mobileye acquisition
Will MoTel become the new WinTel?
In early 1997 people started circulating a three-line joke email. It was about the computer industry and how it measured itself against the Big Three automakers. The comparison went this way: “If automotive technology had kept pace with Silicon Valley, motorists would be able to buy a V-32 engine that goes 10,000 miles/h (…) at a sticker price of less than $50.” To which Detroit would respond “Ok. But who would want a car that crashes twice a day?”
The gag was sticky. Eventually “Silicon Valley” was replaced with “Bill Gates” and “Detroit” which “GM” and it achieved urban myth status. The only thing true in this story though is the implied technological stagnancy of the automobile industry for 2 decades.
This it seems, is about to change. Automobile are entering a second age. Intel and Mobileye are positioning themselves to be the prime beneficiary of the value created by this second age. How this is supposed to happen? We’ll start with Intel.
Some background on the recipe that made Intel successful:
1. Only the paranoid survive:
Be scared, be very scared advised the late Andy Grove. “Business success contains the seeds of its own destruction. Success breeds complacency. Complacency breeds failure. Only the paranoid survive.”
Such was the bedrock of Intel’s culture during its Andy Grove era. A byproduct of that culture was the eventual partnership between Intel and Windows.
Since the days of MS DOS and the x86 platform Windows has shared with Intel the benefits of its vary nature. Nothing scales like software and nothing creates a moat for that software better then network effects.
In large parts thanks to Windows, Intel achieved unprecedented scale and thanks to its network effect it achieved unprecedented sustainable profitability. For nearly three decades WinTel would rule the PC industry.
2. The PC industry was a smiling curve:
At the PC industry-level, Intel and Windows profitability translated into what Ben Thompson calls a smiling curve (coined itself by Stan Shih of Acer).
Smiling curves are industry-level value chains where most of the total value generated by the industry is harvested at both ends of the chain with little value left for the players in the middle. When profitability for each industry players is plotted in a single line graph across the value chain, the resulting line appears to be smiling.
From the 80’s until recently, WinTel was harvesting most of the value on the left-end of the PC industry.
3. But the PC industry recently became a “meh” curve
In 2007 Apple introduced the IPhone and Box launched Dropbox.
Over a couple of years connectivity took over digitalization and smartphones became more desirable then PC.
As for Dropbox and the Cloud, it transferred the processing power needed to support digitalization from personal computers to remote servers. Powerfull microprocessors at the PC level where no longer needed.
Over just a couple of years, the PC industry became another Clayton Christensen case study.
It was time for Intel to become paranoid once again and to find its new smiling curved industry. Arrives the second automobile age.
The second automobile age
- The automobile industry could be Intel’s new smiling curved industry:
Here again, Stratechery has produced some amazing coverage on the subject. It lists 3 changes that the industry is currently undergoing.
- From combustion engine to partial or full electric engine
- From human operated to partial or fully automated
- From individually owned assets to a fleet operated asset
Ben Thompson plotted the expected split in profitability generated by all three changes the same way he did with industry players in the PC industry. And voilà… it resulted into a new smiling curve that if executed properly, could become Intel’s new playground. But how likely to happen. Will the cards fall the same way it did for the PC industry or is it only wishful thinking?
2. The automobile industry smiling curve key assumptions:
This curve makes two important assumptions on:
- Software overlay
- Data ownership
The curve assumes that both self-driving components and ride sharing networks will bundle their offer with software and data to achieve scale and moats in a way similar to what happened in the PC industry. This conveniently leaving car manufactures out of the equation. Although I doubt that car manufacturers will allow for their value to be diluted so easily. After all, car manufactures are the “landlords” of the automobile “real estate” and consequently have first claim on software overlay and data ownership.
In fact, Apple did just that in the smartphone industry. It used iOS as a “glue” to bundle hardware (Foxconn) to customers (in this case end-users). Doing so it avoided the mistake that IBM did with the PC industry and established a precedent for the automobile industry.
The automobile industry could be Intel’s new smiling curved industry but things might not be as straightforward this second time around. Intel needed a strong partner. Enters Mobileye.
Who is Mobileye and what is it worth?
- An industry leader in Driver Assistance Systems
The road towards fully automated vehicles is a continuous process whose foundations were laid with sensor-based Advanced Driver Assistance Systems (ADAS) and for which the next milestone is likely to be Road Experience Management (REM).
Mobileye has been shipping ADAS sensors for 18 years now. It currently has 80% of the ADAS sensor market — through supplying contracts with 13 of the 15 largest Original Equipment Manufacturers. At first glance, Mobileye seems to have both the technology and the contractual relations to ride the upcoming REM wave.
2. Back of the envelope valuation (model)
(Fast-forward to 4.The acquisition to skip technicals)
- TAM: ADAS is under way to reach 100% market penetration globally. Analysts covering the subjects tend to agree on full penetration by 2020 in North America and Europe — regulators are pushing for it. As the fundamental concepts have been already worked out, ADAS should enter a phase of very rapid change, innovation and feature expansion that would translate into a ~30% CAGR. Beyond 2020, as the market matures and there is only China left to fully penetrate, market growth should slowly decelerate and converge towards global GDP growth of ~2%.
- Market share and Operating margin: Empirical data shows that industry leaders in mature industry rarely exceed a 40% market share — that is in traditional industry. The increase in ADAS TAM and in competition — gradually eroding Mobileye’s capability-based competitive advantage, the risk of in sourcing and higher R&D requirements to support REM development is likely to shrink Mobileye market share to just that level and operating income to 25%. Pressure on market share and Average Selling Prices has in fact already started with Tesla severing its ties with Mobileye, Toyota recent partnership announcement with Microsoft on some overlapping features and the rise of Nvidia.
- Discount rate: we’ve applied a 9% initial discount rate converging towards 7.65% — the retail automobile average discount rate.
- REM and the road towards fully autonomous: it’s hard to evaluate the time that it will take to reach L4 and L5 (fully autonomous). Uber disclosed earlier this month that its self-driving cars could travel an average of .67 miles without human interventions. LIDAR is a very expensive technology and quite unaesthetic on top of that. We aren’t quite there yet. Consensus among industry specialists is to get to L4 and L5 by 2030 but even then, it wouldn’t be enough to materially affect Mobileye’s numbers.
3. Stock price distribution
Running a quick simulation on the back of these numbers yields a share price of $41. It is in line with Mobileye’s share price pre Intel acquisition announcement. Sometimes, the market just gets it right.
4. The acquisition:
Intel just acquired Mobileye for 80x its EBITDA. There is no doubt after the Mobileye acquisition that Intel is after the automobile industry. Committing the equivalent of ~100% of your cash reserves is a big statement that lives little room for doubt.
We mentioned earlier that the second automobile age could very well be Intel’s new smiling curved industry provided that cards fell a certain way. A closer look at how Intel and Mobileye have allocated tasks under the recent BMW agreement provides us with a great case study on how they intend to pull it off.
Intel in the second automobile age, the MoTel era:
- Intel and Mobileye partner with BMW:
There are three main aspects to the BMW agreements.
I’ve highlighted what I think is key to consider here: driving policy (the algorithm in charge of autonomous driving) and middleware. Something that needs to be tied back to software overlay and data key assumptions. A claim to both is critical in order for Intel to shape the second age automobile industry into a smiling curve.
2. Software overlay
In “Taking a shot at Valuing SNAP” I alluded to how the cloud and widely available programing tools have commoditized software. I’ll take this one step further in my next publication about “MuleSoft and the rise of middleware”. Today it’s not as much about software features — since they are being commoditized — then it is about software integration. Hence allow me to replace the software overlay concept with the notion of integration/middleware.
3. Data
Data in autonomy is what drives safety. More data means an increase in safety and safety is likely to be one of the main selling points in what will be among otherwise quite undifferentiated fleets of cars. Tesla understands it very well and that’s why data ownership was a point of contention between them and Mobileye. It is likely that this dispute over data ownership is what led to the separation between Tesla and Mobileye in 2016.
4. Et voilà:
The BMW case study shows us that Intel is set on pursuing just that: it will take responsibility for middleware while Mobileye will take care of driving policy — the driving algorithm that leverages data. Both elements are being packaged on an open platform reminiscent of the WinTel platform. There are indeed trying to replicate the PC smiling curve.
5. Acquisition rational:
But why acquire Mobileye then and not simply establish a partnership?
I can see three reasons for why this happened:
- We mentioned in the valuation part that L4 and L5 aren’t likely to happen any time soon and are likely to require heavy R&D investments. With eroding market shares and shrinking average selling prices, Mobileye might not have the financial resources to pull that R&D properly. Intel’s main businesses provide it with ample recurring cash flows that could definitely be put to good use here.
- The automobile industry is a legacy industry. Scale, dedication and respectability play a huge part in deciding who works with whom. Intel is a legacy company in its own industry. It has scale and respectability, acquiring Mobileye is a statement of dedication to the automobile industry. That’s something that is likely to play in Intel’s favor when crafting new agreements with large car manufacturers.
- Intel was late to the second automobile age party. By acquiring Mobileye not only does it secure critical know-how while pre-empting Nvidia — a strong contender to the crown.
I believe that Intel sees the second automobile age as an opportunity to double down on its smiling curve strategy. The way value was being distributed across the PC industry could very well be replicated in the automobile industry provided that Intel is able to secure some software integration and some data access — everything under a MoTel plateform that has become a new industry standard.
I valued Mobileye at $41 based on an expected 40% market share. However, if Intel was able to pull it off — to build economics similar to what it did with Microsoft in the PC industry — it could very well achieve a similar market share of 90%. This would lend us an $84 per share valuation for Mobileye and a second lease on life for Intel, not a bad deal after all.
Thanks for reading — feel free to reach out with any thoughts by replying here or on my twitter 👋
