Earnings Call Transcript - MBLY Q4 2023

Operator, Operator

Greetings, and welcome to the Mobileye Q4 ‘23 earnings call. At this time, all participants are in a listen-only mode. A brief question-and-answer session will follow the formal presentation. As a reminder, this conference is being recorded. It is now my pleasure to introduce your host, Dan Galves. Thank you, Mr. Galves. You may begin.

Dan Galves, Host

Thanks, Kat. Hello, everyone, and welcome to Mobileye’s fourth quarter 2023 earnings conference call for the period ending December 30, 2023. Please note that today's discussion contains forward-looking statements based on the business environment as we currently see it. Such statements involve risks and uncertainties. Please refer to the accompanying press release that includes additional information on the specific factors that could cause actual results to differ materially. Additionally, on this call, we will refer to both GAAP and non-GAAP figures. A reconciliation of GAAP to non-GAAP financial measures is provided in our posted earnings release. Joining us on the call today are Professor Amnon Shashua, Mobileye’s CEO and President, and Moran Shemesh, Mobileye’s CFO. Also joining today for the Q&A session is Nimrod Nehushtan, Mobileye’s Executive Vice President of Strategy and Business Development. Thanks, and now I'll turn the call over to Amnon.

Amnon Shashua, CEO

Thanks, Dan. Hello, everyone, and thanks for joining our earnings call. Starting with our results in Q4, they were in line with the press release we provided on January 4th and with the prior guide, so no surprises here. At a high level, Q4 was a strong quarter in terms of revenue growth of 13% and adjusted operating income growth of 14%. We are pleased with the sequential growth in SuperVision volumes and we expect to see continued excellent growth in that product in 2024. I’d also call attention to our operating expenses, which were significantly lower than what we expected in 2023, much of which relates to transitory issues, but some of which captures efficiencies that should benefit our cost structure over the long term. Looking ahead, the guidance we provided today is unchanged from the outlook we provided in early January. The inventory correction that is impacting the first half of the year has been well publicized. While we did not learn of this build-up until late in the year, we believe we have our arms around this issue and the clear out plan. We have implemented additional processes to monitor shipments versus demand more closely. And we believe we have good visibility into how to put this behind us and get back to normalized revenue in the back half of 2024. Moran will provide some additional cover. Switching gears, as we close out 2023 and come out of CES, it's a good time to remind you of our high-level strategy and assess the progress made. Our strategy is very simple. Our products were in about 40% of auto production in 2023 and we can continue to grow that in the coming years as a bigger percentage of cars are equipped with some level of driving assist technology. But the more important growth driver is average revenue per vehicle, driven by our advanced portfolio products. SuperVision, Chauffeur, and Drive would generate much higher average system prices than our core ADAS products. Events and progress in 2023 gave us more confidence than ever that a very large market for these advanced products is developing, and our technology and business model makes us best positioned to enable and win in that market. On the industry segment itself, we see three clear distinct value propositions that we expect will drive consumer demand and turn into a very large automotive TAM. Number one is a meaningful improvement in safety, related to the surround cameras that are a must on the next generation of eyes-on, hands-free Level 2 plus systems like our SuperVision platform. It is underappreciated that in addition to the convenience of hands-free driving, the 360-degree perception can support a step change in safety. Current single camera systems don't support the many evasive maneuvers that can limit accidents, such as merging into an open lane to avoid a rear-end collision or avoiding vehicles running red lights. Number two is the higher productivity for the car owner. Eyes-off systems like Chauffeur can offer valuable time back to the car owner. If the operational design domain is only 80% to 90% of the time, this is seen as very high value by automakers. Number three is turning vehicles into highly utilized resources. This corresponds to our fully autonomous Drive product. We'll be able to offer self-driving systems for lower than the annual cost of a driver. This unlocks an ability for our customers to generate revenue at a much lower operating cost per mile and with no need to pay or find drivers. We believe these value propositions align perfectly with our advanced product portfolio. And there was much evidence in 2023 to support our view that those products are the highest performing, most scalable, and lowest cost available options in the market. At a high level, all the industry trends were in our direction. The pace of innovation really picked up in China and the pressure on OEM capital efficiency rose. These both are pushing global OEMs to focus more on programmatic issues like time to market, cost, and performance, exactly where Mobileye has advantages. At the same time, we launched the ZEEKR SuperVision software to high praise, which was a significant proof point. Finally, we recently brought to our customers a collaboration framework called DXP that enables the automaker to control the driving experience of a SuperVision or Chauffeur-based system. Finding a sweet spot that enables the OEMs to control the look and feel of the system but rely on our core technologies for all the objective and safety-critical aspects is already paying dividends with customers. On a more specific basis, we announced the value of our 2023 design wins at CES two weeks ago. Future projected revenue was $7 billion for the second year in a row. This compares to our 2023 revenue of $2 billion. The implied ASP of these agreements was $122 in 2023 compared to $105 in 2022. This compares to the ASP for 2021 design wins of $65 and the ASP of our actual revenue in 2023 of $53. The volume associated with the design wins in the last two years is $60 million plus compared to mid-$30 million today. Beyond the design wins, 2023 was an important year for execution, customer acquisition, and expansion of OEMs in the opportunity set. Our SuperVision system is now on more than 190,000 vehicles. We delivered the full highway software in August and it's proving to be a highly capable system. We have expanded the design domain to 22 cities from only two back in September. We believe that proving ourselves in what is the most challenging environment for Mobileye, given data restrictions, proves our global scale and that's a unique selling point that is underappreciated. We were awarded SuperVision design wins with Porsche, FAW, Mahindra, and a major western OEM over the course of 2023. The number of models included in all our design wins is now projected at 30 models compared to nine models at the beginning of 2023. Our portfolio strategy where SuperVision serves as a bridge to Chauffeur is being proven out as Polestar, FAW, and a multi-brand major western OEM, all awarded production programs on the Chauffeur platform during 2023. We diversified the business significantly during 2023 from mostly Chinese OEMs and mostly electric vehicles to a diverse set of OEMs, price points, and powertrain types. The most important catalyst was the landmark design win to bring our entire product set to a major western OEM. It's the first global OEM to align behind our complete portfolio, especially mirroring their future intelligent driving product development plan to our portfolio. It more than doubles the number of vehicle models in the pipeline and spans across all markets and powertrain types. The endorsement of this automaker will be of high value in terms of closing additional deals. We were successful in moving many OEMs into our business development funnel as the high-level trends I described earlier increased the sense of urgency in the marketplace and the confidence in our solutions. We now have design wins or are in advanced discussions with 11 OEMs representing 37% of industry production compared to three OEMs representing 9% of industry production as of the start of 2023. In summary, we know of no other competitor in the ADAS-AV space with a similar breadth of design wins that has actual navigate-on-pilot systems in production and has production programs for eyes-off systems with multiple automakers. Overall, as we have shared previously, we do not expect 2024 financial results to be where we want them to be, given the inventory correction. But we expect to leverage all the groundwork laid in 2022 and 2023 to take a leap forward in terms of visibility toward the next leg of our growth story. I'll now turn the call over to Moran.

Moran Shemesh, CFO

Thank you, Amnon, and thanks for joining the call, everyone. Before I begin, please be aware that all my comments on profitability will refer to non-GAAP measurements. The primary exclusion in Mobileye’s non-GAAP numbers is amortization of intangible assets, which is mainly related to Intel's acquisition of Mobileye in 2017. We also exclude stock-based compensation. Starting with Q4 results, we had another very good quarter with revenue up 13% year-over-year, adjusted operating income up 14%, and an adjusted operating margin at 39%. SuperVision volumes were 38,000 units in Q4, up from 29,000 in Q3. The 67,000 units we did in the second half were significantly higher than the 35,000 in the first half. Operating expenses were again meaningfully below expectations, about $30 million this quarter. There were two main areas, each about the same magnitude. Favorable expenses were lower due to favorable ethics and due to some reimbursement for employees on military reserve duty. The other factor was higher than expected engineering reimbursement for pre-design win activities with certain OEMs. Over the course of 2023, our operating margin rose from 27% to 39% on sequentially higher revenue and consistent operating expenses. Obviously, this is backward-looking, but it should give investors some sense of the operating leverage possible once more meaningful SuperVision and Chauffeur volumes start to drive revenue significantly higher. On a cash flow basis, we generated almost $400 million of operating cash flow in fiscal year 2023. And it's important to note that we invested around $200 million in rebuilding the safety buffer of EyeQ chips on our own balance sheet. We expect to maintain a consistent level of balance sheet inventory in 2024. Capital expenditure was just below $100 million for the year, in line with our prior comments. Looking ahead, you are all aware that as part of the process of setting order schedules for Q1 and the remainder of 2024, we learned that there are 6 million to 7 million units of excess inventory of EyeQ chips at our customers. We understand that much of this excess inventory reflects decisions by Tier 1 customers to build inventory in the basic ADAS category due to supply chain constraints and a desire to avoid part shortages in 2021 and 2022, as well as lower than expected production in certain OEMs during 2023. The inventory situation is related to the base ADAS business only, as SuperVision inventory is at normal levels. As we noted in our January 4th press release and 8-K, we expect Q1 revenue to be down approximately 50% to around $230 million. We expect EyeQ volume to be around 3.4 million units in Q1. We expect SuperVision in the low 30,000 unit range, reflecting normal seasonality in China. Given the unusually low EyeQ volume, SuperVision will be a larger portion of revenues in Q1, which will result in a gross margin in the mid-60s range. Extracting our operating expenses, which will likely be a bit higher than the recent $200 million run rate, the outlook for Q1 adjusted operating income is for a loss of $65 million to $80 million. But we see revenue and volume snapping back fairly quickly and believe we have very good visibility on this. Due to the nature of our business, all of this inventory is for specific OEMs and production of specific vehicle platforms. The process to clear the inventory is simply to stop shipping chips for specific vehicles and have our customers use the existing inventory to satisfy demand. There is no uncertainty regarding who the customers are, there is no alternative product that can be used, and there is no discounting or other economic action needed to clear the inventory. As we compare our prospective shipment with vehicle production schedules, we believe approximately 5 million units can be cleared in Q1 and the vast majority of the remainder in Q2. In terms of our full year guidance, it is unchanged from the preliminary outlook we provided on January 4th, and our visibility has improved over the last several weeks. From a volume perspective, we are assuming 31 million to 32 million EyeQ shipments and 175,000 to 195,000 SuperVision shipments in 2024. We expect the cadence of EyeQs assuming the midpoint of the guidance to be around 3.4 million in Q1, an increase of at least 100% in Q2 versus Q1, and then the balance of unit shipments in the second half of the year. We believe that this cadence, based on our analysis and discussions with customers, should result in a vast majority of excess inventory being cleared by the middle of 2024. We expect average system price to increase in 2024 as compared to 2023 due to an increase of SuperVision as a percentage of total revenue. In terms of gross margin, we look at it on a product by product basis. On the ADAS side, we expect a slight downturn in growth margin this year for two reasons. One, as you know, the cost of EyeQ chips from our supplier went up at the beginning of 2023. We passed that along to our customers. However, there were a decent number of units in 2023 where we generated revenue at 2023 prices but used cheaper chips in 2022 costs. That's a minor headwind this year. We are also assuming some continued normalization of production mix after a very rich mix during the supply chain crisis. We expect these two headwinds to be partially offset by higher cloud-enhanced ADAS volume and REM recurring revenue. Regarding SuperVision, the optimized domain controller is now in production. This comes at a meaningfully lower cost, and we are sharing a portion of that with our customers. ASP will be down a bit compared to last year, but we expect gross margin to be up significantly to low 40% as of Q2 2024 compared to low to mid 30% in 2023. In addition, we would expect some level of software licensing revenue to begin making an impact in Q4 of this year once the ZEEKR free trials are over. Any consumer that chooses to pay for the SuperVision-based feature after the free trial will drive incremental revenue and profit for Mobileye. With respect to operating expenses, we are assuming a 20% increase over the final 2023 number on an adjusted basis, excluding amortization of the intangible asset and stock-based compensation. The OpEx bears a bit more discussion. Our forecast for 2024 is unchanged from what we projected several months ago and not too far above our original forecast for 2023. Much of the lower cost in 2023 related to more transitory things like foreign exchange and delayed moving to our new campus, good news on some engineering reimbursement, and reimbursement of certain payroll costs for employees on military reserves. But some of it is structural. Certain adjustments to the way we collaborate with OEMs, including ASP, means that SuperVision and Chauffeur programs can scale more efficiently than we originally envisioned. The refinement of our mobility-as-a-service strategy to focus on supplying the self-driving system leads to structurally lower costs, but we don't believe in a reduction in the opportunity. Bottom line is that we do believe our operating expenses in the near and long term should be structurally lower than we expected as of a year ago. We continue to believe that OpEx percentage growth in 2025 and beyond should be significantly lower than in 2024. Lastly, in terms of tax rates, we are assuming a non-GAAP effective tax rate of between 15% and 17% for 2024 compared to 11% in 2023. Thank you and we will now take your questions.

Dan Galves, Host

Thank you, Moran. Kat, if you could compile the Q&A queue, please, analysts if you could limit your questions to one main question and one follow-up. Thank you.

Operator, Operator

Thank you. We will now begin the question-and-answer session. Our first question comes from Mark Delaney from Goldman Sachs. Please go ahead.

Mark Delaney, Analyst

Yes, good morning and thanks very much for taking the question. I was hoping to start with an update on the engagements with the OEMs regarding some of your more advanced solutions like SuperVision and Chauffeur. I think as of the 3Q call, you said you were either booked or in advanced discussions with 10 OEMs. I think, Amnon, you said today that's now at 11, and then I believe there's another four OEMs where there were more preliminary discussions underway. Can you give any more color on how those discussions are going, your general sense of progress? And were any OEMs maybe deciding to go another direction, or do you still feel like you're well positioned with that set of customers?

Amnon Shashua, CEO

Well, SV and Chauffeur are complex systems. So early adopters need a lengthy due diligence process. For example, before winning SuperVision on Porsche, we underwent thousands of miles of public road rides in Europe and in the US with a due diligence that took almost a year. But now we are facing the effects of what I would call the law of innovation diffusion, which means that as more OEMs buy into SuperVision and Chauffeur, the shorter their due diligence phase becomes. So after winning the big western OEM, I believe that due diligence phase is getting much shorter. We foresee a number of design wins, both in western markets and in China during 2024. And I believe that the announcement of those design wins would be in the second half of the year.

Mark Delaney, Analyst

That's helpful. One of the things I was hoping for an update on was the DXP platform. It was a big part of your speech and presentation at CES this year. I imagine you met with a number of current and potential customers at CES. Maybe you can share more around how impactful DXP is and the receptivity of auto OEMs to DXP? And maybe touch a little bit on how DXP is different from EyeQ Kit? Thank you.

Amnon Shashua, CEO

I believe that DXP effectively addresses how OEMs can take control of the driving experience in a more efficient way. Previously, before DXP, OEMs approached us with a request to use two EyeQ 6 chips along with an additional powerful microprocessor costing hundreds of dollars to develop their driving policy code on that microprocessor. Others expressed concerns that writing their code on our EyeQ 6 chip would lead to conflicts between our code and theirs over resource management, prompting them to prefer a separate chip. This approach raised system costs, making economic scalability crucial. With DXP, there’s no need for an additional microprocessor, and they can write their code on the MCU using high-level programming while leveraging our infrastructure for their driving policy. This results in two key benefits: it lowers the material costs of the system since no extra chip is required, and it enables Mobileye to scale faster, as the code on the EyeQ 6 chip remains largely consistent across all platforms, with variations only on the MCU. Perhaps Nimrod would like to add more.

Nimrod Nehushtan, Executive VP of Strategy and Business Development

If I may add, we had the opportunity during CES to present this concept to multiple OEMs in dozens of meetings and the reception was very compelling in the sense that although OEMs now have more focus on pragmatic considerations like cost, performance, and time to market, this does not come at the expense of owning the user experience and being able to influence and craft their own kind of user experience for their customer base. What’s really missing in the industry is to find the sweet spot in between the best performance cost in time-to-market solution and full flexibility in crafting a unique user experience. This is where DXP comes in, and it is perceived as a driving operating system by OEMs, which is simplifying the task for OEMs who are now interested in offering new driving experiences in this new generation of driver assist and autonomous driving products.

Dan Galves, Host

Thank you. Next question, please.

Operator, Operator

Our next question comes from Emmanuel Rosner from Deutsche Bank. Please proceed.

Emmanuel Rosner, Analyst

Thank you very much. My first question is around the chip destocking situation that you flagged a few weeks ago. Can you maybe just go back over how you became aware of it? How do you get confidence around the magnitude of the issue and the timing of it being resolved in line with what you reiterated today, please?

Amnon Shashua, CEO

I think, Moran?

Moran Shemesh, CFO

Yeah, so as I mentioned in the script, the inventory buildup issue started actually three years ago, during the COVID period when global production went down dramatically and the entire industry was all about the desire to secure production and to go after every chip. That was also the atmosphere for us with the suppliers and from the sense of urgency that we got from our customers. That of course, we believe led to some stocking and billing activities. Additionally, related to your question, in 2022 and 2023, the ordering process changed, so we needed to make full year commitments to our chief supplier. So we asked our customers to do the same and make full year commitments for this year in both 2022 and 2023, which led to less ability from their side to adjust purchases to demand as they did in the pre-COVID period. So adjusting again the quantities was impossible in 2023. In 2023, we know the supply chain crisis was largely over, but it was still unclear when global production would get back to normal to pre-COVID levels, which might be the reason that the Tier 1s kept holding this inventory throughout the year. They were also, again, obligated with commitments, but they held it through the year, and we weren't informed of such quantities. Over the course of 2023, we believe that our core customers underperformed in terms of production, given that our top customers grew 4% versus the overall market growth of 9%. Towards the end of the year, when the commitments were ending, and now in 2024, we got back to normal in terms of orders, so we have quarterly orders and customers can adjust to the quarterly needs. We think that's the reason it came up towards the end of 2023 and not at some other time during this period.

Emmanuel Rosner, Analyst

Understood. That is helpful. And then just a quick one on SuperVision. I guess as part of this update a few weeks ago, you also sort of tweaked down expectations for SuperVision units in 2024. I think part of it was maybe an exercise in de-risking around timing launches. Could you provide us with a similar type of de-risking around what the trajectory looks like beyond 2024? Obviously, launches have generally sort of happened a little bit later than expected, at CES about a year ago. So just want to make sure that investor expectations are properly calibrated for what's beyond 2024 in the ramp-up initially when you only have sort of a few initial customers.

Amnon Shashua, CEO

At the beginning of 2023, we had a small number of car models and it was difficult to make accurate forecasts. Today, we have 30 car models and we can improve our forecasts. For 2023, in production, we had two car models in China to base our forecast, and all that we had to rely on was OEM numbers which turned out to be optimistic. Now we know that most of the production of SuperVision is coming out in 2026. By 2025, we'll have between nine to 11 models of SuperVision, including five models from the Geely Group, two from ZEEKR, one from Smart, one from Polestar, and one from Volvo, and then between four to six car models from FAW. So that's the production in 2024 and 2025. Then in 2026, we have Porsche, and we have the big western OEM with 17 car models, and we have Mahindra.

Dan Galves, Host

And just one follow up from me, this is Dan. In terms of calibration, for the leading edge analysts that have calibrated their models and estimates to the tracking document that we provided at CES, as well as their updated analysis, these estimates look reasonable and achievable to us. Thank you. Next question.

Operator, Operator

Our next question comes from Dan Levy from Barclays. Please proceed.

Dan Levy, Analyst

Hi. Good afternoon to you. Thank you for taking questions. I wanted to start with just tying into your comments about the engagements with 11 OEMs. I think broadly there is this narrative out there that, like you said, automakers do want to own technology, but the challenge is it’s been very difficult to scale. There’s a notable example of a North American automaker that is pulling back on some of their more advanced ADAS plans given some struggles there. So to what extent are you seeing more engagement with automakers that despite their desire to own the technology really are coming to the realization that they have no choice but to come to you because you are the easiest and fastest way to scale? To what extent are you seeing automakers come around to you?

Nimrod Nehushtan, Executive VP of Strategy and Business Development

Yeah, I'll take it. We are seeing a domino effect in one sense. It is becoming a realization in the industry that the next few years will be very heavily influenced by OEMs' ability to offer products that provide hands-free driving, eyes-off driving, and this is going to become an increasingly important feature for consumers. This is what gives rise to urgency among OEMs to create the shortest path they can for a high-quality product. A few years ago, OEMs believed they had time to invest and they could take a longer path to get there while still owning the technology stack. Now the clock is ticking for them and they are looking for the best performance at the shortest time to market so that they can compete. We already see this dynamic happening in China and it’s growing outside of China today. We do see more traction from OEMs who were in the past more bullish on owning the technology stack and have come to realize that they need to at least find a parallel path inside their company to de-risk the activity toward the next generation of ADAS products.

Amnon Shashua, CEO

I can say that it's really a flywheel effect. A few years ago, the only reference to such a system was Tesla's Autopilot and FSD. Now you have Chinese automakers with similar systems, and some of those models are being exported to the West, featuring mobilized systems. They are now aware that Porsche and other major Western OEMs are also introducing these systems. This creates a flywheel effect. As an OEM, you must respond to increasing competition in intelligent driving. This situation provides more incentives to collaborate with Mobileye on advanced products, either instead of developing in-house or alongside that development.

Dan Levy, Analyst

Thank you. As a follow-up, could you provide us with an update on the competitive landscape in China? I understand that some of your earlier engagements with SuperVision are based in China, but we know that China currently has the fastest development cycle. There are also data barriers in the region. Can you discuss share trends or win rates in China, and to what extent your competitiveness in China compares to your standing in the West?

Amnon Shashua, CEO

It's a complicated landscape. With the western OEMs, we have about 90% share in eight out of the 10 biggest OEMs. This is true not only today, but given all of the design wins to date, this is true for the foreseeable future. Given all the current design wins, our market share is growing. With Chinese OEMs, there are several where we have above 90% market share, including some local OEMs like Chery. There are also some local OEMs where we have around 30%, and others where we do not have a relationship yet. For example, with BYD, we have 30% market share, and with Changan, we have zero. Chinese OEMs are growing faster than our market share growth. BYD and Changan are expanding rapidly. In terms of competition, there's a low-end and high-end. At the low-end ADAS, we have some suppliers competing with very low-cost solutions. At the high-end, we have the in-house development of OEMs. Competing systems suffer from a large performance gap in basic features like autonomous emergency braking, for example. In order to catch up, they would need more expensive systems that would reduce their competitive offering. Moreover, we are adding REM to low-end ADAS to provide a cloud-enhanced system, and Chery just announced that in two months, car models with Mobileye Cloud Enhanced solutions will be launched. This will add more pressure on our competitors by creating a moving target for single camera ADAS. For high-end systems, those are developed in-house by some OEMs, but significant proof points are still needed.

Dan Levy, Analyst

Thank you.

Dan Galves, Host

Thank you, Dan. Next question, please.

Operator, Operator

Our next question comes from Itay Michaeli from Citi. Please proceed.

Itay Michaeli, Analyst

Great. Thanks, everybody. Good afternoon. Just two quick questions for me on the pipeline. First, hoping you can kind of dimension what portion of the pipeline is both looking at SuperVision and Chauffeur combined, just given the momentum you've seen with Chauffeur recently? And second question, since the big announcement at CES, have you seen an increase in activity and conversations, including maybe with the second wave OEMs you described in the last earnings call?

Amnon Shashua, CEO

Yes, Nimrod will take it.

Nimrod Nehushtan, Executive VP of Strategy and Business Development

Regarding the first question, it depends. In some of the engagements we have, we have parallel engagements for both SuperVision and Chauffeur with the same OEM. This is in cases where the OEMs are interested in a few car models for which they may want to have different offerings for different market segments and price levels of cars. On the other hand, there are some OEMs who might be more interested in specific product offerings and maybe to start from Chauffeur or to focus on hands-off for their cars due to various considerations. We can say we have a mixed bag of engagements and generally see a good mixture of both Chauffeur and SuperVision in our pipeline engagement. I think regarding your second question, one of the interesting outcomes of CES is that we have had, just over a couple of weeks after CES, I think three engagements within our pipeline that specifically want to focus on evaluating DXP hands-on, to actually start working technically on DXP, to start experimenting with it, and to see how they can influence the driving experience. This is just after a couple of weeks since we announced it for the first time. So it is a very promising start to this promotion process that we are executing.

Dan Galves, Host

Thanks, Itay. Next question, please.

Operator, Operator

Our next question comes from Joe Spak from UBS. Please proceed.

Joe Spak, Analyst

Thank you, everyone. You've mentioned the progress on the SuperVision program wins. Could you provide some insight into how customers view implementation for these programs or their take rates? Are they confident they can monetize these features? Amnon, you noted that for more advanced features like Chauffeur that save time, do we need such features to significantly increase adoption levels and customers' willingness to pay?

Amnon Shashua, CEO

In the West, we have a reference point regarding the cost of Tesla's Full Self-Driving. This highlights the importance of having a very economical system to give automakers flexibility in their pricing. Mobileye's system, which includes sensors, costs less than $2,000. This allows original equipment manufacturers a lot of flexibility in pricing, especially considering that Tesla's FSD costs $12,000 for end customers. Most of our partnerships view the system as a standard feature, so it’s integrated into every vehicle rather than based on a take-rate calculation. This is the current situation.

Joe Spak, Analyst

Thank you. Regarding the inventory issue you mentioned, could you elaborate on the procedures you're implementing to improve sell-through tracking? Also, you noted that over the past couple of years, you've made a full-year commitment. Has that led you to reduce those full-year commitments as part of your new procedures?

Moran Shemesh, CFO

Yeah. So I'll take it. First of all, I must say that we never experienced this situation before as our customers have done through the 10-plus years in a very good job in ordering demand. Given the recent history and the inventory issue, we are taking action on several fronts to add capabilities to monitor shipments versus demand. So, the first thing is that the order process is now back to normal. Commitment for 12 months is no longer relevant. The commitment is only a quarter ahead. We do receive 12-month forecasts, but that’s just a forecast. It’s not a commitment. It's the same thing with our suppliers. The industry and I’ve been saying, the 12-month commitment has not been relevant in 2024, and we don’t expect it to be relevant also in the future as it was specific to COVID. We’ll also try to receive some input from our customers on inventory levels, but it’s not something we have visibility on. It’s important to mention. In terms of actions we’re taking internally, we have established a regular process to match shipments to detailed vehicle production, looking both backwards and forwards. We are putting more focus on market-based forecasts that incorporate the adoption rate and OEM share trends. We’ve always used and updated forecasts and compared to the market, but we will now put more weight on those as inputs from customer-provided forecasts. We will also consider working with external vendors to build some statistical model forecasts to incorporate macro-level data and add additional headcount to support this. Those are the steps we are taking.

Joe Spak, Analyst

Okay. Thank you very much.

Moran Shemesh, CFO

Another clarification that Dan mentioned, it wasn't clear in my script. So the full guidance for 2024 hasn't changed from January 4th. For EyeQ, we are anticipating 31 million to 33 million units of EyeQ shipment. A clarification since the line was down a bit. Thanks.

Dan Galves, Host

Thank you, Moran. Next question, please.

Operator, Operator

Our next question comes from Vijay Rakesh from Mizuho Securities. Please proceed.

Vijay Rakesh, Analyst

I have a quick question about the company landscape. Mobileye is performing well in the Western hemisphere, especially with CARIAD no longer in the picture and GM Cruise facing challenges. However, in China, there are concerns about NVIDIA potentially gaining market share. What are your thoughts on this? Is it related to price performance or are OEMs looking to diversify? Could you share your observations on this situation?

Amnon Shashua, CEO

I believe that in China, the in-house development systems still have considerable proof points to address. They need to tackle geographic scalability, performance scalability, and economic scalability. Economically, these systems are significantly more costly than ours. For instance, at NIO, they utilize four Orin chips, which are owned by NVIDIA and are quite expensive. Even with four chips, they have not achieved the performance levels comparable to SuperVision. On the economic front, in-house systems require additional sensors to achieve acceptable performance levels. Mobileye has distinct advantages in geographic performance; moving to eyes-off capabilities represents a significant advancement in performance and emphasizes safety. In eyes-off systems, perfection is essential, with no room for errors, which further strengthens Mobileye's position. The Chinese market is fascinating and dynamic, and I believe we are performing well there, with expectations for our market share to increase. While the market is expanding faster than our market share growth, I hope we will catch up soon.

Vijay Rakesh, Analyst

Got it. And just another question on SuperVision. Obviously, there is good ramp there, but as you look out through ‘25 and ‘26, can you talk about what kind of conservatism you're embedding either on the OEM unit side or on the number of OEMs, I guess, as you build that outlook on the SuperVision side?

Amnon Shashua, CEO

In 2024, there will be five car models from the Geely Group; two from ZEEKR, one from Smart, one from Polestar, and one from Volvo, and then starting at the end of 2024, beginning of 2025, we expect four to six FAW car models to be added. The big jump is in 2026 when we have Porsche, as well as the big western OEM with 17 car models, and Mahindra.

Dan Galves, Host

Just to follow up, Vijay, we want to be more conservative, right? This is why we provided a tracking document with the number of models and OEMs and launch dates for analysts to make their own estimates for these years. Analysts who have calibrated to this and adjusted their forecasts, we see those as reasonable and achievable. Thank you. Next question, please.

Operator, Operator

Our next question comes from Luke Junk from Baird. Please proceed.

Luke Junk, Analyst

Good afternoon. Thanks for taking the question. First question, you stated in the prepared remarks that EyeQ visibility has improved in recent weeks. Just hoping you can expand on what is better understood sitting here in late January.

Amnon Shashua, CEO

I think the visibility is better understood.

Moran Shemesh, CFO

Based on the shipment schedule for 2024 and the information that we have on specific inventory levels from our research and based on also on customers' input. We reviewed the detailed production forecast and, as we mentioned, we expect the majority of the excess inventory to be cleared by the end of Q1 with most of the rest clearing in Q2. Of course, actual production levels of per OEM customers will play a role. However, based on our projections, we believe the excess inventory will be fully cleared by year end. At the end of the day, we performed a true analysis of the ADAS self-treatment rate and lower production per OEM to understand that in order to meet this year's production, that’s what we need to provide, considering also the inventory issue. We have better visibility. We also, of course, mentioned Q1, we have also some visibility to Q2 that we said will be at least 100% higher than Q1. For the rest of the year, again, in line with production expectations based on our analysis and what we gathered from our customers, that's where we think we're going to land.

Dan Galves, Host

Thanks, Moran. I’ll just follow up with a couple of things because we've been obviously looking very closely. The visibility since January 4th has improved because we have commitments for Q1. We know generally what the volume is going to be in Q1. We have more visibility in terms of Q2 starting to adjust into commitments. So that's why we have the confidence to say that Q2 will be at least 100% higher than Q1. In terms of the back half, we have indications from the Tier 1s that match up with the market-based forecasts we're looking at as well. Overall, we feel good about the visibility toward the 31 million to 33 million units of EyeQ volume in 2024, and that level of production will result in the elimination of the excess inventory almost completely by the middle of the year and then potentially with a little bit left in the back half.

Luke Junk, Analyst

Understood. Thanks for that, Dan. And then my follow-up question, hoping you could comment on some of the key facets of the anticipated expense growth in 2024. In particular, there’s been certainly an increasing focus on the AI-related facets of driving policy development. I think it would just be clarifying to understand how Mobileye is investing incrementally in Generative AI tools this year? Thank you.

Amnon Shashua, CEO

Our OpEx growth is mostly devoted to our desire to execute all these programs in SuperVision. The Porsche, the western OEM, all of those are converging to 2026 to continue supporting ZEEKR, of course with the many OTAs that are going forward. Our move from a Tier 2 to Tier 1 with Porsche and the western OEM means we’re acting as a Tier 1 supplier. In other cases, we’re acting as a Tier 1.5, but these require more resources to support this well. As for using AI, this does not need growth. This is our normal activity with the existing manpower that we have. It is more about moving from a Tier 2 to Tier 1 and supporting the many car models we're producing in the next few years that require some growth.

Moran Shemesh, CFO

Of course, the headcount cost increases, a few tens of millions, relates to the higher number of employees and maybe more enhanced salary raises, as well as the savings that we had in 2023. We need to recall that in Q4 we had some reimbursement for military service. These are not things that necessarily will happen in 2024. The ILS effect, Israeli Shekel effect in 2023. Beyond that, we also have significant facilities growth of a few tens of millions. We're moving to the new campus with an appreciation of approximately $20 million higher. There will also be growth in the EyeQ platform, EyeQ 6 and EyeQ 7, our Radar product as we approach 2025, as well as in the Lidar domain, we expect some growth in 2024.

Dan Galves, Host

Thank you. Next question, please.

Operator, Operator

Our next question comes from Chris McNally from Evercore. Please proceed.

Chris McNally, Analyst

Thanks so much, Amnon. Thanks team. So maybe just some quick math, cleanup here on the ADAS hand. I think previously, Mobileye has discussed 50% penetration for the industry moving to about 75% by ‘25 and ‘26. This seems slightly pushed out now looking at ‘25 or ‘26, something like in the low to mid 60% penetration. So first, can we talk about industry adoption on base ADAS? And then second, around global market share, Amnon, you discussed lower share on maybe some of the domestics of China where share is maybe 50% or below. Is it fair to see your 65% to 70% historical share maybe move to this kind of 64%, 65% on that mix effect over the next couple of years? Any new high-level math that you could provide us on industry trends later in the decade would be really helpful.

Amnon Shashua, CEO

I think with the western OEMs, our market share is continuing to grow. Based on all the design wins from 2023 and 2022, we are growing our market share. As I said before, eight of the ten biggest OEMs have more than 90% market share. With China, the growth of the Chinese OEMs is faster than our market share growth. Our market share there is reducing naturally. As I mentioned before, Changan is growing very fast, and we have no relationship with them. BYD is growing very rapidly as well — we have only 30% market share with BYD. We continue to get design wins, both in low-end and high-end ADAS like SuperVision. However, how the market share in China will play out in the coming years is still an unstable position because the market is moving very quickly.

Nimrod Nehushtan, Executive VP of Strategy and Business Development

If I may add, I think that if you refer to a correction to the market share calculation based on the inventory levels, we think the inventory levels that we've disclosed were accumulated over a period of time longer than a year, likely closer to three years. If you compound what the effect would be for the annual volumes we've disclosed, it’s about a $2 million reduction per year on average, which accounts for maybe 1% or 2% of our market share calculation over the past. We don't believe that indicates a significant change in our market share forecast on a global level.

Chris McNally, Analyst

All makes sense on the explanation on market share. Maybe if we can go back to industry adoption, sort of when will we hit on a global basis, 75% penetration? Obviously, some of those western players, Amnon, have been pretty slow to make standard fit, obviously outside of Toyota. But any view on what is sort of a ballpark year we can think about industry ADAS penetration being around 75%?

Amnon Shashua, CEO

From external sources, this is the number projected to be achieved by the end of the decade: 75% market share of ADAS. We have no reason to believe this will change.

Dan Galves, Host

Even possibly higher. The western markets are pretty well penetrated, probably above 70%. But there's still some growth there. However, a market like India, which is maybe single-digit ADAS penetration, has seen extremely successful systems on the road in production in the last couple of years. Everyone is trying to catch up, and we have a very good position there. This will be a strong market for ADAS adoption as well as our share growth in the next couple of years.

Nimrod Nehushtan, Executive VP of Strategy and Business Development

We see growing pull from regulatory bodies in emerging markets to promote ADAS system adoption. This started just recently in Europe with regulations like GSR, which is not just a bonus feature, it's a mandate to sell cars. So these two driving forces are what we believe will push the industry towards higher adoption rates in the next few years.

Chris McNally, Analyst

Thank you.

Dan Galves, Host

Thanks, Chris.

Operator, Operator

This concludes our question-and-answer session. I would like to turn the floor back over to Dan Galves for closing comments.

Dan Galves, Host

Thank you, Kat, for managing the call. Thanks to the management team of Mobileye, and thanks to everyone for joining. We'll talk to you next quarter. Thank you.