Car Tech: Silicon Valley Takes on Detroit
SEE LAST PAGE OF THIS REPORT Paul Sagawa / Artur Pylak
FOR IMPORTANT DISCLOSURES 203.901.1633 /.1634
psagawa@ / email@example.com
October 26, 2015
Car Tech: Silicon Valley Takes on Detroit
GOOG and AAPL have been investing in automotive technology hoping to disrupt nearly $2T in annual car sales, $4T in car related expenses and $1T in local transportation and delivery. According to reports, AAPL intends to follow TSLA’s lead with its own plug-in electric vehicle (EV) by decade end, targeting innovation in the user experience and leveraging its brand/design assets. While the market for EVs is growing quickly, they remain a tiny portion of the overall market and must overcome the obstacles of uncertain industry standards, high costs, limited vehicle range, and a sparse support infrastructure. Established car makers are also investing heavily in EVs, and bring substantial advantages, particularly if the transition continues at a gradual pace. In this context, TSLA’s long term play may be as a battery supplier while AAPL may be better off with an established partner rather than as a stand-alone car manufacturer. In contrast, GOOG’s self-driving car play targets transportation as an on-demand service, following the Uber model. By eliminating drivers in fleets of personal transport/delivery vehicles, service costs will drop sharply while safety and predictability improve. AMZN, with its plans for autonomous delivery drones, is following on a parallel track. The concept faces stiff regulatory and consumer education obstacles but would be extremely disruptive to the existing model of personal vehicle ownership and the associated support infrastructure. Ultimately, we believe this paradigm shift foretells opportunities that will be far bigger and more lucrative than the more evolutionary transition to EVs.
- Transportation is a massive market. At nearly $2T in yearly sales, cars are one of the few consumer product markets bigger than mobile phones. That, the emphasis on design, the growing importance of electronics and the disruption of electric vehicles make it inevitable that AAPL would be interested. The emerging market for on demand transportation services initially targets the $1T world market for taxi, livery and local delivery, but ultimately should threaten personal vehicle ownership, and its $6T in annual costs, as well, once fleets of self-driving vehicles overcome regulatory bureaucracy and demonstrate their substantial cost, safety and practical advantages.
- EVs inevitable but uptake slowed by cheap oil. EV’s have made obvious progress but remain just 0.3% of annual new car sales, with cost of ownership still a considerable premium to gas powered cars and range limitations still an impediment. Cost and performance improvements in battery technology have been linear, while low gas prices keep the relative operating cost bogie farther ahead. This modest pace of change favors incumbents who have made considerable investment in EV tech and can gradually transition their product lines.
- Self-driving tech progressing rapidly but regulation is a major obstacle. Automakers have been introducing self-driving capabilities to augment their traditional vehicles, adding automatic parking, adaptive cruise control, and other functions as options. While many reports have focused on the batteries of sensors, the real innovation behind self-driving is a deep learning AI able to interpret sensor readings and react accordingly. Despite growing consensus that these cars are safer and more efficient than driver operated vehicles, regulators remain cautious in approving them for general use. However, there are countries, with particular congestion, pollution and/or infrastructure concerns, that may be early adopters, and their experience could hasten adoption elsewhere.
- Self-driving vehicles will be huge accelerant for transport-as-a-service. While an auto-pilot option for personal vehicles could prove popular, freeing commuters from drudgery and would-be drunk drivers from catastrophe, we believe that on-demand transport-as-a-service fleets will be the breakthrough application. The disruptive business model established by Uber threatens car ownership as well as traditional for-hire vehicles and local delivery when costs and reliability see step function improvement from autonomous operations. Autonomous vehicles are also a natural for the long-haul trucking market, improving efficiency and eliminating costly and dangerous driver errors.
- AAPL makes high risk move to lever brand and design skills. Cars are obviously a tempting market for AAPL, which could lever its peerless brand, design prowess, hardware/software integration skills and leadership in user interface technologies to give a go at its own high-end EV. Still going it alone is a risky proposition, given economies of scale, a steep learning curve, tight industry margins and their considerable gaps in many areas crucial to car making. Just as the widely rumored AAPL made TV set never materialized, we believe an all AAPL built car is unlikely. We look for a partnership instead.
- TSLA endgame is batteries not cars. TSLA is an object lesson for AAPL, having made profits in just one quarter over its 13 years and selling just 31.6K units in 2014. In contrast, the world’s largest premium auto brand, BMW sold 2 million cars in 2014, including an estimated 17.8K plug-in EVs in just their first year of wide availability. As the EV market plays out gradually, incumbents have time to play catch up, while levering their assets, limiting the competitive disruption. By contributing its IP toward open standards and investing in its Gigafactory, TSLA reveals that its plan may not be dominating EVs but rather, their batteries, which will have sizeable residential and industrial applications as well.
- Uber prepping for future move to autonomous fleet. Unimaginative analysts may scoff at the $50B implied valuation for a mere “taxi and limousine” replacement, but Uber has set its sights on a global disruption of transportation writ large. As the cost of personal transportation and local delivery on demand falls, Uber competes not just with yellow cabs, but with car ownership and store visits as well, expanding the future addressable market many fold. Seeing the potential for driverless cars to both sharply reduce costs and raise customer convenience, Uber has begun to invest heavily in self-driving technology through a partnership with robotics hotbed Carnegie Mellon University.
- GOOG positioning to use self-driving tech to challenge Uber. GOOG began its work on autonomous cars in 2009, proving prescient as incumbent car makers and disruptors like Uber and TSLA have jumped on the bandwagon. Self-driving systems rely on highly accurate maps, elaborate sensor arrays, and adaptive on-board systems informed by massive databases of driving experience. With 1.5 million miles logged, GOOG leverages the most data parsed by the world’s most sophisticated data analytic and learning computing platform, supported by the most accurate maps. We believe that these are substantial advantages, even relative to the other very capable organizations that have been working on the problem. We note that although AMZN’s drone delivery initiative does not involve cars, it is levering similar expertise against a fundamentally similar idea.
SSR TMT Heatmap
Mo-Valley – Tech Takes On the Auto Industry
First came TSLA in 2003, out to prove that electric vehicles (EV) could end the century long domination of gasoline powered cars. Then came, GOOG and its widely lampooned self-driving car “moonshot” in 2009. Shortly after, Uber’s on-demand transportation service hit the world’s consciousness. Last year, AMZN shocked 60 Minutes viewers with its delivery by autonomous drone reveal. Finally, earlier this year, reports leaked that AAPL had been hard at work in secret on an EV project of its own. Clearly, the tech world sees opportunities in transportation, a multi-trillion dollar set of targets comprising automobile sales and life cycle costs, local delivery services and even long haul trucking. The keys to unlocking these opportunities are plug-in electric vehicles (EV) and autonomous navigation.
Progress on EVs is slow and steady – battery tech follows a linear improvement trajectory, and issues with industry standards and service infrastructure require cooperative solutions – making this more of a sustaining innovation rather than a disruptive one within the Christensen framework. TSLA, 13 years in and a pioneer in most aspects of EV design, still finds itself generally unprofitable with 2014 unit sales of just 31.6K cars and faces enlightened incumbents, like BMW, which can lever the scale and experience behind its 2 million annual unit sales toward its fledgling EV business and its 17.8K cars sold. Indeed, TSLA’s endgame may be as a dominant vendor of batteries – it has contributed all of its IP toward open standards and committed billions of dollars toward building its Gigafactory manufacturing plant in Nevada. Batteries produced there, and at future Gigafactories by TSLA’s plans, are meant not just for TSLA cars but for all standards compliant EVs, as well as for residential and industrial applications.
Into this walks AAPL and its 1,800 employee skunk works operation. AAPL’s track record is too good to count them out, and its device electronics and systems software expertise are surely powerful and unique assets. Still, it is also far down the experience and scale curves on many of the elements of auto design and manufacturing, with an ambitious rumored target date of 2019. Just as much rumored the AAPL TV set turned out to be something less audacious than first envisioned, we see a partnership with a like-minded incumbent like BMW or Mercedes to be much more likely than an AAPL car built from the ground up.
Like EVs, most incumbent car makers are also investing in self-driving tech, intending it as an auto-pilot option for otherwise traditionally operated vehicles. GOOG’s plans are more expansive, looking to challenge Uber for the future of local transportation with fleets of autonomous vehicles, leveraging years and millions of miles of driving data processed by the company’s world leading deep learning and data analytics capabilities. AMZN has a similar idea, focusing on autonomous drones to attack the delivery aspect of the opportunity. Uber has its own skin in this game, having recently funded a substantial research initiative into self-driving vehicles with Carnegie Mellon University. While autonomous fleets face considerable obstacles – regulatory approvals and public acceptance chief amongst them – the technology could be ready for primetime within a few years. One approval by a green leaning jurisdiction could provide proof of concept that then begins a domino effect across others. Such a scenario would be truly disruptive to private car ownership and local delivery, yielding a step function drop in costs and a corresponding jump in efficiency.
Transportation is a huge market. Last year, some 90 million cars and light trucks were sold worldwide, generating nearly $2T in revenues for the automotive industry (Exhibit 1-2). Other non-gasoline costs of car ownership – parking, insurance, maintenance, etc. – take the annual total to more than $6T. Taxi, livery and local delivery services are an additional $1T. Global trucking is another $2T. By comparison, annual worldwide mobile phone sales are less than $400B, with wireless service revenues clocking in at less than $1T. Moreover, neither mobile phones, nor wireless services are expected to grow over the next 5 years. No wonder Silicon Valley is so interested in building cars.
The sense is that the transportation industry is ripe for disruption. After more than a century of domination by gasoline fueled engines, the electric motor, with its extraordinary low-end torque, simple mechanics, and zero emissions, has finally become a viable alternative. This, in turn, removes a major barrier of entry for a technology industry far more familiar with batteries than with internal combustion engine thermodynamics. With a foot in the door, electronics and software wizardry could be levered to rethink the user experience for drivers and passengers, an arena where tech stalwarts like Apple and Google have justifiable confidence. Or at least so goes the elevator pitch.
Worldwide sales of plug in EVs rose 53% YoY to roughly 320,000 units in 2014 (Exhibit 3). Still, penetration into the 90M/year global car market is miniscule, with high price premiums, range limitations, and an insufficient public charging infrastructure remaining obstacles to a broader paradigm shift. Hybrid vehicles, which use supplemental gasoline engines to minimize battery requirements and extend vehicle range, have been available for roughly 15 years but still account for less than 3% of total industry sales (Exhibit 4). Green issues be damned, it would seem that plug in EVs will remain a niche product until prices, performance and convenience can reach near parity with gasoline powered cars.
Exh 1: Global Automobile Unit Sales, 2000-14
Exh 2: Top 10 Global Automaker Revenues, 2010-2014
Exh 3: Global EV Sales, 2010-14
Exh 4: US Automobile and Hybrid Sales, 2000-14
It Keeps Going, and Going, and Going …
We believe that these obstacles will erode slowly. Typically, batteries make up 40-50% of the cost of an electric car, and battery technology has historically improved on a gradual linear trajectory rather than on a Moore’s Law driven accelerating path. Periodically, articles touting new power storage technologies with step function potential improvements over the now standard Lithium Ion formulations, but there are many, many hurdles to cross before breakthroughs in a university laboratory environment reach mass commercialization. For example, hydrogen fuel cells were first incorporated into an experimental automobile design in 1991, developed as prototypes by major auto makers over the past decade, and introduced by Toyota and Hyundai as limited production products a year ago. Once widely considered the likely power source of the future for automobiles, the technology is still plagued by very high costs, the need for a hydrogen distribution infrastructure and major safety concerns. Increasingly, scientists have become skeptical that these problems can be overcome in time for fuel cells to compete against Lithium Ion. The same long slog with similar pitfalls seems the likely fate for new alternative battery technologies ranging from solid state to metal-air.
Tesla’s Elon Musk is betting on the ongoing primacy of Lithium Ion with his pocketbook, committing $5B in partnership with Panasonic to build his “gigafactory” to produce batteries based on that technology and asserting his strong doubts as to the viability of alternatives still in the laboratory phase to pass muster within the decade. Assuming Musk is correct, battery costs will continue to trend down slowly and remain a significant cost burden on EVs for the foreseeable future, more so, if the industry cannot resolve its internecine battles over technical standards. If each car maker continues to insist on their own proprietary battery solution, all will continue to bear heavy R&D expenses and it will be very difficult for any of them to reach real scale efficiency (Exhibit 5). To that end, Tesla has pledged all of its patented battery innovations toward a royalty free open standard, with hopes that a broad and early agreement can accelerate improvement in EV costs and promote investment in a standard charging and battery swap infrastructure. The company believes that it can cut the cost of its batteries by 30% from the current $200-300/kWh with its gigafactory, but achieving that target likely requires increasing current auto sales by more than 10 fold (Exhibit 6).
Exh 5: Tesla R&D Spend, 2010-14
Obviously, hitting the target would be much easier if Tesla were supplying competitors’ vehicles as well as its own, but support from incumbent suppliers has been inconstant. Daimler bought almost 10% of Tesla in 2010, only to exit its investment by 2014. Toyota announced a partnership with Tesla in the same year, and the companies co-developed an all-electric RAV4, but the relationship unraveled and the vehicle ended production in 2014 (Exhibit 7). The most recent talk is that Tesla has been in discussion with BMW, itself a leader in plug in EVs, toward joint standards on batteries. Of course, these established car makers have much less at stake than Tesla – they have little incentive to hasten the transition to EVs as long as no one is too far out ahead and may be reticent to give Tesla too much of a boost.
Exh 6: Projected cost per KwH of Automotive Lithium-ion Batteries, 2010-2020
Exh 7: Tesla Partnership Summary
Not Disruptive, but Sustaining
With designs unlikely to near cost parity with gasoline powered vehicles for many years due to the intractable expense of batteries, with gasoline prices remaining low vs. recent history, and with a significant convenience burden for consumers (narrow range, home charging infrastructure requirements, inadequate public charging solutions, etc.), the conditions for mass adoption of EVs are not ideal. Moreover, incumbent automakers, with substantial manufacturing and distribution advantages along with considerable ongoing investment in EV technology, have the luxury of playing their hands cautiously. New entrants to the EV market are not surprising anyone and must shoulder serious scale and experience diseconomies. Widely accepted industry standard batteries and a broadly deployed public charging infrastructure are not going to appear by magic. To use Clayton Christensen’s well known framework, EV technology is a sustaining innovation that can serve to extend the dominance of market leaders rather than a disruptive innovation by which a scrappy start up can erode the incumbents from below.
Published forecasts for future plug in EV sales are all over the map, ranging from less than 25% to more than 50% annual growth (Exhibit 8). Given the relatively small penetration of EVs today, we are comfortable with the more pedestrian forecasts, but the aggressive ones imply a major tipping point that we do not believe is forthcoming, at least within the next decade (Exhibit 9).
Exh 8: EV Forecasts, 2010-2020
Exh 9: Global EV Sales, 2010-14
The Uber Revolution
In June 2014, NYU Stern School of Business finance Professor Aswath Damodaran posted a provocative piece of analysis on the Nate Silver’s well-known 538 blog detailing his case for why Uber was worth less than half the $17B that had been assigned to it by private financing rounds. His case revolved around the key assumptions: 1) Uber’s addressable market was the historically defined $10b/year market for taxis and livery vehicles; and 2) Uber’s potential market share was arbitrarily capped bat 10%. (http://aswathdamodaran.blogspot.com/2014/06/a-disruptive-cab-ride-to-riches-uber.html) A month later, Uber board member and Series A investor in Uber, venture capitalist Bill Gurley posted a detailed response (LINK) politely rejecting Damodaran’s assumptions, and asserting that by his reckoning, Uber’s fair valuation could be 25 times higher or more than the Professor had calculated. We are inclined to agree with Gurley and note that subsequent financings have increased the projected valuation to more than $50B.
Gurley is particularly thorough in addressing Damodaran’s definition of Uber’s addressable market. Not only does Uber address the existing car-for-hire market with a superior product, it expands the market by enabling entirely new use cases based on its lower price, greater convenience, wider reach, and superior experience. Gurley also drops the bomb that Uber is even becoming a credible alternative to car ownership itself, ball-parking the total annual worldwide cost of car ownership at $6,000 per car times an installed base of $1B cars, or $6T (Exhibit 10). If Uber and services like it were able to take even a tiny fraction of that market, Damodaran’s assumption of a $10B addressable market could easily be a magnitude or two off. Gurley didn’t even bother to add in the potential of adding local delivery to the mix, a recently added service for Uber and a further $1T in addressable market.
Exh 10: Uber’s Addressable Market
Now With Robot Drivers!
Alphabet (Google) began its foray into self-driving cars in 2009, a “moon shot” investment that was widely lampooned at the time. Since then, its laser range finder equipped test vehicles have logged over a million miles, collecting data that has fed the deep learning AI that acts as the brains of the system. It is this system that understands that human drivers sometimes accelerate to beat a yellow light, that pedestrians sometimes cross in the middle of a block, or that a ball bouncing into the street is sometimes followed by a child. Over the course of those years and miles, Alphabet’s self-driving cars have been involved in 11 accidents, every one of which was caused by a human driver, most of whom smacked into the back of an Alphabet car at a stoplight.
Today, investments in self-driving technology are in vogue. Many major auto companies, Tesla included, have begun to introduce elements of autonomous operation – from self-parking to freeway autopilot functionality – as options on their higher end vehicles. Of course, these systems, designed to work within fairly limited scenarios, are not intended to replace a human driver, but rather to offer a convenience to make the driver’s task easier. This, like electric operation, is a sustaining, rather than disruptive innovation.
However, push button parking is not Alphabet’s vision. Imagine an Uber-like fleet of completely autonomous cars, available at a moment’s notice. With 360 degree vision, exact knowledge of the road ahead, constant vigilance and hundreds of driver-equivalent years of accumulated driving experience, these cars are dramatically safer than driving yourself or hopping a taxi. Coordinated by an AI-based system, the cars are much more likely to be in the right place at the right time, minimizing waiting time and maximizing utilization. Without human drivers, who we estimate represent roughly 50-60% of Uber’s total costs excluding estimated operating costs borne by the drivers, prices could be dramatically lower, particularly factoring the further effects of better utilization (Exhibit 11). There would also be social benefits, beginning with the safety improvements, but also including reduced congestion, less real estate required for parking, and less wear and tear on infrastructure. The cars in this scenario would almost certainly be plug in EVs, able to charge en masse at depots during fallow hours, and assuring the fleet a green pedigree. With time, governments might liberalize speed limits for autonomous vehicles, given their superior safety capacity and their potential ability to co-ordinate close driving at speed on highways.
Exh 11: Uber Contribution Margins
This would be Uber on steroids – a huge step up in service performance with a huge step down in cost. Old habits die hard, and predictions that a majority of the world’s car owners might cede their driving autonomy are likely farfetched, but we believe the real number would likely increase Bill Gurley’s estimate of the market addressable by transport-on-demand services like Uber by yet another magnitude. No wonder Uber has amped its own investment in self-driving technology by poaching an academic team from robotics stalwart Carnegie Mellon University.
We are confident that on-demand transportation provided by fleets of autonomous electric vehicles will be commonplace worldwide within the next 20 years, generating trillions of dollars in industry revenues and displacing a significant portion of the world’s privately owned cars. However, the pathway to that eventuality is complicated and uncertain, with major hurdles to pass, some technical, some regulatory and some psychological. The technical hurdles are fairly straightforward – Alphabet anticipates having resolved its remaining issues for completely autonomous vehicles before the end of the decade, a perspective echoed by several other companies investing in the arena. Governments will take some convincing – driverless transportation-on-demand will threaten jobs and vested political interests, and regulators will take their time in evaluating the safety, congestion and public infrastructure impacts (Exhibit 12). Meanwhile, even if a car with no driver is demonstrably safer than a traditional taxi service, consumer perception, biased in the opposite direction, will take time to adjust. However, given the public benefits of reduced congestion, pollution and necessary infrastructure, it would seem likely that some government, some place, will take the leap sooner rather than later, and once the concept proves in in that one place, inevitably, others will follow.
Exh 12: Autonomous Vehicle Legislative Action (as of September 2015)
Signed, Sealed, Delivered
Human transportation is not the only major application for autonomous vehicles. Local delivery is currently a roughly $1T global market, but by eliminating half or more of the costs, autonomous delivery services could attract more merchants to offer delivery on more products to a wider set of customers within broader geographic boundaries. Driverless cars and trucks will be joined by flying drones in speeding packages to their destinations more quickly, more safely and with lower costs.
There are major challenges. Procedures for completing deliveries must be established – without a driver, there is no one carry the package to the door, ring the bell and get a signature confirming receipt. Security must be assured – unmanned deliveries will be tempting to thieves and vandals. Laws and regulations will need to be written or revised – how will these cars and drones share the roads, skies and airwaves? The first application may be levering drones to make deliveries in places with inadequate roads – a project to develop drone depots in Uganda is already underway, while Alphabet has invested in developing a system for rapidly deploying drone delivery in response to natural disasters and armed conflict. Ironically, these humanitarian focused projects could be a substantial proof of concept for future commercial services.
China has already permitted Alibaba to test market drone delivery of tea to customers in Beijing. Meanwhile the FAA is at work trying to craft the regulations under which it may allow Amazon, Google and others to conduct similar testing in the U.S., and has allowed a test for the delivery of medical supplies to a remotely located clinic in western Virginia. International rivalry amongst China, the U.S., Korea, Japan, and the E.U. could clear red tape for drone delivery even more quickly than approvals for self-driving cars.
We also note that autonomous vehicles could also seriously disrupt long haul trucking, where driver fatigue is a serious problem with profound implications for public safety and operating efficiency. Autopilot operation that allows truck operators to rest and attend to other tasks without stopping or risking accident would be immensely valuable, even if drivers were still assigned to take responsibility for the cargo.
The Apple Car
The preponderance of evidence shows us that Apple is working on cars. Dozens of executives and engineers have been poached from automakers and their suppliers, including Mercedes Benz R&D exec Johann Jungwirth and Chrysler Senior VP and head of service and quality operations Doug Betts. The Wall Street Journal reports that “Project Titan” has been raised to the level of committed project with hundreds of dedicated workers at a secret location close to Apple headquarters. The company has contacted automotive testing facilities to negotiate terms, and has had meetings with automotive contract manufacturers. Just this past week, CEO Tim Cook publically stated that “the automotive industry is ripe for disruption”. Where there is this much smoke, there is usually fire.
Still, it doesn’t mean that Apple will, in fact, deliver a car of its own design and construction to the market. Similar rumors a few years back that Apple intended to introduce a branded television set proved to be misplaced. Apple may be kicking the tires on the idea of building a car, but could decide to end the project before bringing it to market. Apple could be developing technologies that will find their way into vehicles primarily designed and built by others. While the sheer size of the market and the sleepy pace of change is undoubtedly intriguing, particularly given the maturity of the premium smartphone market (http://www.sector-sovereign.com/2015/09/17557/), the economies of scale and experience, the historically tight profitability of the business, and panoply of new skills that building a car would require should certainly raise hard questions for Apple as it considers its options. However, let’s assume that Apple is hell bent on building a car and intends to introduce it in the rumored 2019 timeframe. What will it have to overcome and what could success look like?
Apple has already introduced Car Play, a schema for tying iOS devices into the electronic systems of a car. It would seem natural for it to extend its platform to manage these systems and control most of the human interaction with the car. Innovations could involve heads up displays, rethinking notifications, gauges and controls, and enhancing navigation and infotainment – areas where incumbents have offered only incremental improvements at a measured pace. Apple’s mastery of hardware/software integration, and processor design would be a boon to the electronic systems that monitor and manage most aspects of the modern automobile. Apple also has obvious expertise in Lithium Ion batteries, albeit on a dramatically smaller scale
Still, that leaves a lot of the car where Apple has no applicable experience or skills to go on, including mission critical items like electric motors, drive trains, suspensions, steering and breaking systems, aerodynamics, climate control, safety systems/crashworthy design, etc. While the eye rolling from automotive industry executives about an Apple car is reminiscent of the scoffing from mobile phone makers at the introduction of the iPhone, it will be difficult and expensive to develop sufficient expertise in these areas quickly enough to deliver a world class vehicle to market on that 2019 timeframe.
Moreover, the obstacles that EVs must overcome on the way to widespread adoption – high relative costs, range limitations, insufficient infrastructure – will hamper Apple as they have all EV makers, with Apple’s taste for ample margins suggesting an even greater price premium vs. gasoline powered cars. Even if the patina of the Apple brand and the presumed radical design innovations enable a substantial price gap over rival EVs, the economies of scale and experience will make it difficult to generate profits at all, much less at Apple margins. Note that Tesla earns gross margins of roughly 26% on average, contribution that is largely eaten up by ample R&D and SG&A spending, on annual vehicle unit sales of about 50,000 units and expected revenues of $5.6B. Apple’s corporate gross margins are nearly 40% and its operating margins are better than 30%, levels that are not plausible for a fledgling EV manufacturer (Exhibit 13).
Exh 13: Cost Structures of Major TMT and Auto Names
In the context of Apple’s $224B in annual sales, a business the size of Tesla would be just 2.5% of revenues with a minimal impact on earnings. With plug in electrics unlikely to make up more than a percent or so of the total auto market by 2020, and with the majority of those sales presumed to be “affordable” models with price tags of less than $40K, the opportunity for an Apple car business to be a lot bigger than Tesla is today could be a decade or more away.
We think that the prudent approach would be to partner with an established car maker with an advanced EV development program. Rumors surfaced earlier this year that Apple had approached BMW about such an arrangement, but that negotiations were unsuccessful. On the surface, the two companies would seem to be made for one another – it is easy to imagine a BMW i3 or i8 tricked out with Apple software for the driver and passengers – so perhaps there is room for a deal yet. Other candidates could be Mercedes Benz, Nissan, Honda, Hyundai or even Tesla. Then again, perhaps Tim Cook and company will shock the automotive world, just as Steve Jobs shocked the wireless industry back in 2006. Perhaps, but probably not.
The Alphabet Fleet
These days, everyone seems to be working on self-driving cars. Established automakers have already rolled out autonomous driving features, like automatic parking and adaptive cruise control that can speed up, slow down and even steer to keep a vehicle moving with a lane of traffic. Still, no company has progressed as far as Alphabet (Google). In June, it reported passing 1M miles of autonomous driving, adding roughly 10K miles per week. This data set is unique, and a substantial advantage vs. the more recent and less extensive efforts of the auto industry or fellow upstarts, like Uber. Google Maps, with considerably greater detail, coverage and accuracy than rival map services, and Waze, with crowd sourced real time traffic conditions, add to this trove of unique data assets. With all of this information and experience, Alphabet also has the deep learning AI leadership to create a truly superior autonomous driving solution. Someone else can build the car to spec, Alphabet will define the brain that lets it drive by itself.
Beyond the car itself, Alphabet can tap growing fleet logistics expertise gained from its fledgling Google Express local delivery business and further lever its extraordinary data processing infrastructure. Alphabet also has considerable experience in dealing with governments – a skill that will be critical to finding a test market or two to prove in the concept. This is probably the biggest hurdle – there will be vigorous opposition from drivers, organized labor, taxi and livery services, skeptical public advocates, luddites and such (Exhibit 14). Google’s battles with the EU over antitrust and privacy have been well publicized, but it has also had a big hand in regulations, like the FCC’s net neutrality rules, and legislation, like the defeat of SOPA copyright protections. Importantly, it also has experience working at the municipal level, having negotiated favorable deals to deploy residential fiber networks in a growing number of communities. This experience will be invaluable when the first brave cities begin to approve fleets of autonomous vehicles for transportation-on-demand.
It is a big world, and we believe that some government will find the idea of leading the world with this solution and providing its citizens with the benefits of Alphabet’s investments attractive. We expect the first systems to be up and running within 5-7 years. Other companies will look to follow Alphabet’s lead – Uber may be a few years behind the first Google self-driving cars, but it is highly motivated to get in the game and one can expect others, like Lyft to move that way too. Automakers, tied to the model of personal vehicle ownership, have little incentive to refocus their self-driving vehicle development toward the on demand transportation-as-a-service concept, much less to foster its adoption. Still, it seems likely that Larry Page and company will have the last laugh on the armchair skeptics who ridiculed the self-driving moonshot way back in 2009.
Exh 14: Trucking Expenses per Mile
On the Sunday after Thanksgiving in 2014, CBS’s 60 Minutes aired a segment on Amazon and its founder, CEO Jeff Bezos. At the end of the recorded piece, Bezos added a little sizzle by revealing the company’s top-secret work on autonomous drone octocopters that could deliver packages by air. Since then, rivals like Alibaba and Google, have revealed similar ambitions, with Alibaba already conducting live tests, delivering tea to customers in Beijing.
Once again, regulation appears to be the gating factor – skeptics have raised objections to the idea of fleets of drones clouding the skies, delivering all manner of dangerous payloads, smashing into people, property and each other, and otherwise disrupting daily life. The FAA, with responsibility for America’s airspace, has taken up the task of drafting rules for commercial drones, allowing an initial experiment in rural Virginia for the delivery of medicine to a remote clinic. Unlike the transportation-on-demand market, we expect first approvals for drone deliveries in relatively sparsely populated markets in the US, moving to more populated areas only after the concept proves safe and viable in the initial markets.
This likely suits Amazon, which has been known to play the long game, and works to the disadvantage of potential rivals like Google or, someday, Alibaba, who are far, far behind Amazon in the US package delivery market. Like Google, Amazon has enormous computing capabilities to apply toward the complex systems necessary to manage many thousands of drones simultaneously racing to complete deliveries with unpredictable weather and other unforeseen obstacles. Success for Amazon would enable even faster delivery intervals to much wider geographic territories at much lower costs. Yet another reason why traditional retailers may not be frightened enough about the future.
Winners and Losers
The development of all electric and self-driving vehicles is far along, and while the widespread adoption of these technologies is still 5-10 years away, winners and losers have already begun to reveal themselves (Exhibit 15). We expect EVs to grow gradually as a percentage of the overall vehicle market, limited by a substantial and slowly shrinking cost premium, by competing technical standards, and by massive necessary investment in infrastructure. Within this context, we believe that incumbent automakers hold many important cards. We believe that Tesla has much more future opportunity as a supplier of batteries – assuming that it can manage the industry standards setting process to its benefit – than it does as car maker. Apple, looking to lever a shift to EVs to enter with its own branded car, will have an even more difficult path to success despite its enviable brand and relevant electronics expertise. We believe that “Project Titan” would be more likely to drive shareholder value if Apple produces its car in partnership with an established manufacturer.
We see self-driving vehicles as a catalytic element in the future of on demand transportation-as-a-service. This concept is far more disruptive to the auto industry status quo than EVs, as it greatly changes the calculus behind personal car ownership. Fleets of autonomous cars, available at short notice to cheaply, and safely, take people to where they need to go, will supercharge the already fast growing Uber model of transportation. We believe Uber is likely to be a winner building on its already vibrant franchise, but expect Alphabet to emerge as the early leader based on its technology advantages. Once proven in early adopting markets, we believe this paradigm will spread rapidly into a huge future market. We also believe that drone delivery, led by Amazon and Alibaba, will disrupt the local delivery truck model, lowering the cost and time for deliveries even further.
Exh 15: Winners and Losers
©2015, SSR LLC, 1055 Washington Blvd, Stamford, CT 06901. All rights reserved. The information contained in this report has been obtained from sources believed to be reliable, and its accuracy and completeness is not guaranteed. No representation or warranty, express or implied, is made as to the fairness, accuracy, completeness or correctness of the information and opinions contained herein. The views and other information provided are subject to change without notice. This report is issued without regard to the specific investment objectives, financial situation or particular needs of any specific recipient and is not construed as a solicitation or an offer to buy or sell any securities or related financial instruments. Past performance is not necessarily a guide to future results.