Contents Introduction 3 Existing Players 4 The General 6 Driving Chips 9 Market Size 10 Who Stands to Lose 14 2
Introduction Like it or not, driverless cars are coming. The reality is that mass adoption is a long way away, but assisted driving is already here and 2017 looks to be a pivotal year for the industry. Most people know the top names in the industry already, but in order to be fully prepared for what this could mean to your portfolio a deep dive is necessary. This report aims to take a look at what exactly an autonomous car looks like, who will be selling them, what the important components are how the auto industry will be impacted. Beyond the auto industry there are a few other potential winners and losers from other sectors of the economy. Finally, we will take a look at some of the best investment ideas from this nascent space. Just What Is A Driverless Car Simply put, an autonomous car can detect its surroundings using a variety of tools such as radar, lidar, GPS, odometry and computer vision. These tools allow for a central computer to interpret the sensory information and make decisions on how to proceed in the car. There are plenty of images for cars with huge sensors sitting on the roof. There are almost as many videos of cars rolling down the road with no drivers. More recently, we see a driver sitting behind a wheel, but no hands on the wheel as it steers the car. 3
Existing Big Players A report on driverless cars should lead with the company that is the most innovative in the space. That company is Tesla (TSLA) and it is the lead not because it is an electric car, but because its cars already on the road have the auto pilot feature. This is not to be confused with the notion of a driverless car, auto pilot is more like assisted driving. Still, this is an important feature that gives TSLA a huge advantage over all other autonomous car companies. The idea of having auto pilot available on a large number of TSLA s already on the road is analogous with the idea of how the Google machine came to dominate search. One of the main reasons that Google ascended to its dominant position in search was that it got better with each individual use. So the more searches, the better the Google search results became. The same is likely to be true in the world of autonomous cars. 4
With more than 164,000 cars in service, Tesla (TLSA) already has a huge advantage over all other autonomous car companies. The early data collected can be increasingly fine-tuned and prove to be the biggest hurdle other competitors will face. This factor alone puts TSLA very far ahead of the competition. Speaking Of Google One of the earliest entrants in the market of autonomous cars was Google (GOOGL). There are numerous stories of Google trying to impress potential hires and they would send a car to get them at the airport. It wouldn t be just any car, it was a driverless car. This was years ago and the notion of a driverless car was still something right out of the Jetsons. Since then, Google has done a lot or work on the subject, or have they? The company has spun out its self-driving car unit Waymo. The company is focused on technology, not just on the cars. This implies that there will not be a car coming from Google but instead its technology will be there to partner with the automakers. Given the volumes of data that Google has from its existing efforts, partnering with the tech giant will look very attractive to many automakers. The problem is, there could be several roadblocks in the form of other partners that won t want to give Google a chance to replicate / re-engineer their technology. More on that later. Waymo and the self-driving car concerns from Google started back in 2009 and they have since seen 2 million self-driven miles. They also claim to have 1 billion simulated miles driven just in 2016. Clearly they are working on the issue, but there is a lingering question of who will they partner with. One big auto maker has recently made a big push into the space, while 5
another has not done much. Google has partnered with Chrysler for a ride sharing service deployment using the Pacifica van as the vehicle of choice. The service looks to be up and running by the end of 2017. At first glance this looks to be more of competing service for Uber and Lyft than for Tesla or the other autonomous car companies. The General General Motors Co., the nation s largest auto maker, acquired Cruise Automation a developer of self-driving technology for approximately $1 billion. The company also invested $500 million in Lyft Inc., a ride-hailing service that competes with Uber Technologies. General Motors is testing autonomous vehicles on Michigan s roads, after the State of Michigan signed into law a series of bills that permit the testing of driverless vehicles within the state. This development follows Google s testing in Austin TX and its Mountain View CA home. The company has very recently announced that the Bolt will not only be an electric vehicle with a 238 mile range, it will also have a base price of $37,495. Of course those details would not have been interesting at all if it weren t for the fact that it will also be a testbed for autonomous driving. Ford Mark Fields, Ford s CEO announced that the company plans to offer fully self-driving vehicles by 2021. The vehicles, which will come without steering wheel and pedals, will be targeted to fleets which provide autonomous mobility services. Fields expects that it will take several years longer until Ford will sell 6
autonomous vehicles to the public. As previously mentioned, there is one major auto maker that just doesn t seem to have done a lot in this space. Ford is clearly that company as they have had very few announcements around the space. They have declared an autonomous fleet will be in production in just 5 years, but details are few and Uber Uber is a ride hailing service that has disrupted the taxi ecosystem. Since its inception, the company has been faced with all kinds of legal hurdles. Rules that governed taxi s and limousines were avoided as the peer to peer ride sharing service saw explosive growth in which people used their own cars and gas. In its current form, Uber is dependent on the human driver who has a car of their own to drive. A driverless ecosystem is not an advantage or disadvantage for Uber as the company is developing its own technology. This could allow Uber to license that technology to other auto manufacturers, which could be a significant Uber CEO, Travis Kalanick, has indicated in a tweet that he expects Uber s fleet to be driverless by 2030. Early predictions for a driverless Uber platform to be so inexpensive and ubiquitous that car ownership will be severely depleted. This is more likely to be the case in urban centers than in rural areas, but still a consideration for the automakers. The Direct Play Mobileye (MBLY) of all the names that have been discussed, the purest play in this space is the company that is already working with dozens of automakers to make future models. Some of 7
those models will have driverless options that will be provided by Mobileye. As a licensor of its technology, MBLY sits in the coveted spot of providing services to numerous automakers and thereby allowing it to increase its driven miles by orders of magnitude when full release comes. The concept of driven miles as discussed previously, is concerned with the idea of improving the autonomous car experience. In much the same way that Google was able to improve its search engine with each search, each mile driven by a car on the Mobileye platform will improve its service. Although early revenues for MBLY have been less than exciting, the amount of deployed vehicles using their technology is still just a fraction of the cars in on the market. As the market opens up, that should change for MBLY, but to what degree is hard to estimate. Earlier, the idea of Google partnering with automakers was discussed, which is exactly what MBLY has already done. There could be potential problems as previous Google partners have seen their technology marginalized or outright taken over by the search giant. How those partnerships play out is anyone s guess. MBLY has a large amount of information on its website including some high level of details on how the autonomous car actually works. If you are really motivated you can spend a lot of 8
hours going into the Deep Reinforcement Learning for Driving Policy, but you much be able to handle this sort of discussion: Driving Chips There are two major names in the microchip sector that are making deep inroads in the future of autonomous driving. These plays are indirect as they are only a small part of the ecosystem around autonomous cars. Still they are worth mentioning, as the chips are an integral part of the brains of the autonomous car. Nvidia (NVDA) was singled out by Elon Musk when he noted that Tesla was already producing self driving cars. The charismatic CEO said that a Titan chip from Nvidia would be powering the software that the company is developing in house. This was a huge boost for the chip maker as it validated their product to Wall Street. It is one thing to be in the know as far as the industry goes, but getting the nod from an industry leader 9
/ significant innovator is a major step. The other major chip name, Intel (INTC) also bears mentioning in this report. The chip giant has doubled its efforts to be a major player in the internet of things as well as being a presence in autonomous cars. It makes sense that the worlds largest producer of microchips is positioned itself to be a player in this nascent industry, but we won t know how deep INTC has penetrated it until some of the numbers begin to roll in. Market Size Boston Consulting Group has estimated that the driverless car market will balloon to $42B by 2025. Another consulting firm, Grand View Research says that there will be roughly 140K autonomous car on the road by 2024. That view, when broken down, calls for roughly 20K autonomous cars being sold each year. When compared to the recent monthly average of more 10
than 17M regular cars and trucks each month, that number looks ridiculously low. Given the rate of change in technology as well as the potential for these cars to drop in price over the next few years, we could see more those estimates look conservative. As much as the autonomous car market is the subject here, we have to take a look at the entire auto market to get a clearer picture of where we stand. Auto sales were hit hard during the great recession in 2009. Sales bottomed out at 9.1M per month in March of 2009, and it took 6 years to get back to the levels seen before the recession. Late in 2016 we heard the idea of a peak auto being discussed by the pundits. This was basically saying that we have seen the high water mark in terms of total new car sales in the final three months of 2015 in which 18.2M cars were sold in each month. Of course soon after that we made a new high level of 18.3M in November of 2016. December saw a slight decrease to 17.9M, in what has recently been a stronger month for auto sales. Just the discussion of the millions per month has to put some perspective on the idea of only 140K autonomous cars being solid at some future date. Even if 5% of cars available are solid with autonomous capabilities we are only two months from reaching that level. This tells us that the current expectations for autonomous cars may be very low indeed. 11
The Benefits A number of reports are out championing the benefits of the wide scale use of autonomous cars. The most stunning projection will be a 90% decrease in the number of car related fatalities. This benefit alone could be a savings of more than $200B (mostly due to healthcare savings) in the US alone. It is also estimated that an autonomous fleet of cars on the road could save 10 million lives per decade. Other benefits of a driverless fleet is that there would be less traffic. Anyone that has taken an Uber Pool ride can attest to the idea of the computer maximizing the ride efficiency. That means not only will the car be directed in the most efficient route, it will also pick up the maximum number of passengers. Along with efficient routes and full cars, the connected cars will offer another benefit. The idea of less traffic also means cars can move more quickly on the roads. Congestion will be decrease which will open up lane space, but the real increase in commuting speeds will be achieved as a result of computers efficiently spacing cars and causing less braking and accelerations that tend to cause braking and congestion. 12
That is a long way of saying smoother rides with fewer man made interruptions. Finally, global warming stands to take a big hit when the world has more automated cars than human driven vehicles. Studies show that the reduced traffic (congestion and number of cars) will lead to fewer miles driven. The miles driven will also be done so more efficiently and thus less gasoline will be consumed. Some reports suggest as much as a 20% decrease in fuel could result from autonomous cars. The Dangers When we talk about a potential future event, there are a lot of unknowns in the equation. This can lead to plenty of irrational conclusions, but often times the concerns are valid. In the case of autonomous cars, the primary concern is around the decision making process for how a car either avoids one accident which will cause another. This situation is best described in an event where are an autonomous car has to decide which car to hit if one has to be. Do the riders of the target car come into play? Another hurdle, beyond existing regulations, could be the acceptance by human drivers. This idea is not only meant to address the human drivers that will never give up control of their car, but those that simply cannot afford it either. It is one thing to talk about who can afford the top of the line Tesla, but another to suggest that the public will accept an additional $10K or more for the self-driving option. The biggest worry that seems all too likely is that the cars will be hacked. This sort of thing could lead to all sorts of criminal actions, the least of which would be kidnapping. If a hacker has you trapped in a car, no one will be coming home making 13
a robbery that much easier to achieve. More nefarious ideas could include bodily harm of enemies but with those dangers already existing, it is just about making them easier to occur from nearly anywhere in the world. Licensor vs Licensee We discussed a lot of participants in the autonomous car, but one of the best ways to look at the company s involved is to determine where they stand. The early stages of the industry seems to have two segments, either they are making the technology and licensing it out (TSLA, MBLY, GOOGL, UBER) and there are those that pay for that technology (GM, F, most automakers). It is too early to know which position will win out, but it seems to be clear that the longest runway will be with the companies that have a little of both. Right now, that looks to only be Tesla. Previously, Google was on a path that led to a mixture of both the car and the software to provide the service. However, they ended their program of building the car with the technology built in and opted for the more neutral stance of partnering with automakers. Who Stands To Lose Depending on who or what you believe, several reports suggest that the biggest losers will be the automakers themselves. A fleet of self-driving cars that could offer super low prices for very efficient rides could make the idea of owning a car become a bad idea. This would only likely be true in the densely populated urban areas, where as wait times in rural areas would still require a farmer to have his / her own transportation. 14
It would be hard to say that there is one big loser from less accidents, but the auto repair shops stand to lose business in a world with autonomous cars. This idea extends to the suppliers of paint for fender benders / auto parts for more major incidents. Add to that the idea of increased efficiency and the discount part suppliers take a hit as do the providers of tires. Another loser could be the insurers, as the number of accidents drops significantly. That is really the first worry that an insurer would have, as the question of who is to blame for the accident really gets into the splitting of hairs. Is it the car marker, the technology that was licensed or maybe the chip that powered the brains of the car? Stealth Winner There could be one company that has the solution that we all gravitate to. It was somewhat hard to believe so many years ago that most people would migrate to a $600 phone that had a few more bells and buzzers on it. When we realized how easy it was to use and how powerful it was the stock was really off to the races. Since then, tablets and watches have followed and products from this company are ubiquitous. Apple (AAPL) has been hinting at the idea of entering the autonomous car market, but recall that this company is super secretive and fond of huge announcements when they do enter a market. Should AAPL throw their hat in the ring, they would be an immediate favorite to dominate the industry. This is due to the fact that they have sold millions and millions of phones, tablets and watches. These items all serve to further lock in a customer base that is fanatically about their products. 15
The only thing we have so far are rumors and artist renditions of what an Apple car might look like. For the faithful, these images are what dreams are made of. The rest of us android users still have to admit that we would love to kick the tires of what could be a very cool car. 16