Intel Mobileye announced Asian deals to bring L2+ to China and MaaS to South Korea.
L2+ has ~10x higher ASP, while Mobileye’s global MaaS footprint is the largest of all providers.
Mobileye generated less than $1 billion in 2019 revenue, but I estimate this could quintuple to $5 billion by 2023 based on L2+ adoption.
Based on Mobileye’s historic P/S ratio, this would make it a ~$135 billion business as a standalone company, implying a 50% stock upside to Intel’s shares from Mobileye alone.
I view Mobileye as an undervalued Intel asset, but one that could contribute meaningfully to Intel’s growth in the medium term and has significant potential to increase Intel’s shareholder value.
Intel’s Mobileye continues to fire on all cylinders. At CES, the company announced two new, important Asian deals: one to bring its REM crowdsourced RoadBook to China via SAIC, the other to launch its MaaS operations in South Korea in 2022, joining a global range of countries currently announced.
Both these deals meaningfully expand Mobileye’s near, medium and long term outlooks. Near term, I estimate that L2+ could help the company to quintuple its revenue in the next four years. Additionally, Mobileye’s MaaS leadership sets it on track to leadership in a $160 billion TAM market.
While Mobileye was Intel’s smallest reported business at just under $1 billion revenue in 2019, I believe Mobileye the above indicates that Mobileye will contribute materially to Intel’s earnings and revenue growth in the near-term. I view it an undervalued Intel asset that will contribute significantly to shareholder value in the coming years, as a simple valuation model suggests.
L2+ has Significant Earnings Potential
L2+ is an unofficial level that Mobileye has invented that goes beyond Level 2 (ADAS), but below the initial self-driving Level 3 requirements. It is powered by its latest 2 TOPS EyeQ4 chip and offers some additional capabilities. Most noteworthy, it is based on the company’s REM RoadBook, its crowdsourced 10cm accuracy map (at 10kB/km) that is set to offer complete U.S. and Europe coverage within the next year as millions of vehicles collect data.
At CES, Mobileye announced a partnership with SAIC to bring both L2+ and REM to China. This opens a large new market for the company’s L2+ endeavors.
Intel’s crowdsourced RoadBook is a significant competitive advantage in L2+ (and eventually autonomous driving): as Mobileye had previously announced, it has an 11/12 model share of current L2+ systems and a further 100% nomination track record on another 13 systems. This not only proves Mobileye’s ADAS leadership, it will also significantly contribute to Mobileye’s revenue growth. The company shared the following slide at CES:
Moreover, the slide states that L2+ systems have an impressive 3-15x higher ASP (Mobileye CEO Amnon Shashua rounded it to ~10x in this presentation). As a public company, Mobileye reported a $45 ASP for its EyeQ chips. This puts L2+ at anywhere from $135 to $675, giving it a tremendous increase in the dollar value it can capture per car.
Meanwhile, regular L2 adoption is also expected to continue increasing significantly over the next few years as ADAS adoption is expected to increase to 60%, up from 30%. Mobileye shipped 17.4 million EyeQ chips in 2019, continuing to grow volume at a 40% CAGR, so a significant opportunity remains.
Taking this information together, although Mobileye’s current 40% CAGR in EyeQ shipments might not be sustainable, this will be more than made up for by the ASP increase in L2+ systems as they ramp to higher volume within the next one or two years. For instance, if Mobileye ships two million units in 2022 (a small amount compared to legacy ADAS currently) at $400, up from a negligible amount last year, then that could result in an $800 million increase in revenue in just two years from this new category, almost double its revenue last year.
Given that analysts are currently not expecting Intel to grow revenue by much in the next two years (about $2.7 billion), then an $800 million contribution from just one segment of Intel’s smallest reported business puts those analyst projections in context.
Mobileye’s MaaS leadership
Intel also extended its global Mobility-as-a-Service (MaaS) footprint at CES. It announced an agreement with Daegu City, one of South Korea’s largest metropolitan areas. This agreement extends its previously announced intentions to enter global MaaS operations in Israel, France (and potentially Europe), China and the U.S. This gives Mobileye the largest footprint of all MaaS providers.
Mobileye expects that only a few MaaS operators will succeed and is taking a slow and steady approach. Moreover, Mobileye estimates that the autonomous MaaS TAM will reach $160 billion by 2030. If the company’s current list of deals and progress is any indication, it has a good chance to gain significant market share of that TAM. It will start its MaaS operations in 2022, and in the U.S. in 2023.
I am not sure how fast Mobileye could be able to scale up its robotaxi service, despite this large theoretical footprint, but the company seems to be taking a diligent, measured approach via key partnerships and a phased roll out. However, Mobileye’s success in robotaxis, or lack thereof, should not alter the near-term picture.
Mobileye currently has a very lucrative and growing ADAS business, but the company is also investing in additional opportunities. I think people are easily overlooking the extent and magnitude of Mobileye’s scope. Combined, these various revenue streams might quintuple the company’s revenue by 2023, as I’ll explain.
First, Mobileye estimated at its investor meeting in November that it expects to generate ~$1.8 billion from its legacy ADAS business in 2023. This is likely a conservative estimate since as the company probably does not want to overestimate the design wins it will gather that far out.
Secondly, as explained above, L2+ is a brand new segment, but expected to grow at a >50% CAGR in the coming years. Based on four million cars at a $400 ASP, it could become a $1.6 billion business and double Mobileye’s ADAS business compared.
Moreover, I deem it likely that he company could also release some sort of L2++, as the company is aiming at what it calls Vision Zero: zero accidents. It intends to accomplish this by bringing its RSS safety policy for autonomous vehicles down to its ADAS business. Such an L2++ is likely to command even higher ASPs, so I will assume this results in a further $400 million upside revenue.
Thirdly, Mobileye is also monetizing its REM RoadBook via licensing for automotive and non-automotive use cases, and expects this to represent a $750 million TAM by 2024. Mobileye expects to generate one-tenth of this $750 million number from in 2020 from its partnership with the Brittish Ordance Survey.
Besides these, Mobileye is also developing its own radar and lidar sensors in what it calls its EyeC portfolio, and intends to sell these sensors to other players as well. Mobileye thinks this market has a $10 billion TAM by 2030.
The revenue from its sensor and robotaxi businesses is a wildcard since they currently do not contribute any revenue yet. But given Mobileye’s global scope, up to $600 million combined MaaS and sensor revenue in its second year of operations might not be unfeasible.
The above information provides good support for my ambitious $5 billion 2023 revenue model, representing an over 5x growth rate in just four years. This model is just for illustrative purposes to show that despite its small size, the company has the potential to keep growing at a rapid pace and thereby contribute materially to Intel’s top and bottom lines:
- $1.8B from L2 (Mobileye’s own estimate)
- $2.0B from L2+ (assuming >4 million cars at ~$400 ASP)
- $0.6B from REM licensing (based on Mobileye’s TAM estimate)
- $0.6B wilcard: MaaS and EyeC sensors
As described, most of Mobileye’s potential near-term growth will hinge on the adoption rate of L2+ and the ASP Mobileye will be able to charge for this. In any case, its current market share, RoadBook mapping progress and design win share should give a lot of confidence.
On a less serious note, this would represent 12 years of growing at at least the pace of Moore’s Law, from 2011’s $40 million until 2023.
If L2+ volume growth does not meet the 60%+ CAGR projections Mobileye put out, or if its ASP is below the ~$400 mark that Mobileye CEO Amnon Shashua suggested, then this result in much less growth than I estimated, although this could perhaps be mitigated if other business lines grow faster.
Qualcomm entered the ADAS and AV markets at CES with Snapdragon Ride, but initial commercialization is only expected in 2023. Its 2023 chip will offer 700 TOPS at 130W, compared to Mobileye’s 2023 EyeQ6 with 130 TOPS at 40W. Nvidia has also gained some traction already with its own Drive platform.
However, it takes more than just fast chips to gain traction in the automotive space. I believe Mobileye’s existing market leadership in ADAS gives it a significant headstart to any other competitor. Mobileye has a comprehensive strategy that spans L1-5 ADAS to full self-driving systems, sensors (cameras, lidar and radar), mapping (REM RoadBook), driving policy (RSS), hardware (EyeQ), software (both open and closed solutions) and MaaS strategy.
Importantly, unlike other competitors Mobileye as a whole is already highly profitable from ADAS despite its strong investments in full autonomy.
As described above, Mobileye’s current ADAS leadership looks set to continue firmly into L2+.
While Mobileye will still be just a sliver compared to Intel’s full size even in three or four years from now, partnerships such as the ones it announced at CES are laying the cornerstones of its future growth.
I mainly wanted to highlight the massive, likely underestimated, ASP upside that L2+ brings, at around ~10x higher than legacy ADAS, and thereby giving Mobileye (and consequently, Intel) enormous near-term revenue growth potential.
This puts Mobileye on track to rival and surpass the size of Intel’s FPGA, IoT and NAND groups over the next few years, as I illustrated with my $5 billion revenue model in 2023.
While $5 billion would still be less than 10% of Intel’s, it would be a fast growing $5 billion business, contributing the top and bottom lines and overall growth. There is also no indication that the company’s momentum will slow down after that timeframe, as it will still be the early innings of L2+ and initial ramp of full autonomous L4/L5.
Hence, I view Mobileye’s momentum as a key argument for any long thesis (or against any short). Given its small size, Mobileye still seems a massively overlooked asset in Intel’s portfolio and one that has yet to be priced into Intel’s share value.
Mobileye was trading at a roughly 27 P/S (price to sales) ratio prior to rumors of Intel’s acquisition. Assuming the company would sustain that ratio, which seems plausible given that the premise of autonomous driving will be ever more looming in 2023, Mobileye would be (or would have been) worth $135 billion by 2023 as a standalone company.
This represents a 50% premium to Intel’s current $260 billion market cap. Mobileye might hence become the most important contributor to Intel stock value in the medium term.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. 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.