Transportation as a service, more commonly called TAAS, with autonomous vehicles will be a ride-hailing business competing with the likes of Uber and Lyft. GM already owns about 9% of Lyft. We think the market was disappointed that GM said it will be ready to deploy autonomous vehicles in a commercial service at scale in dense urban environments in 2019. In recent months, management had said the service would be ready in quarters, not years. We expect the first cities to get the service to be San Francisco and New York, with Phoenix also possible, since GM already tests autonomous vehicles there.
We think GM is right to focus more on complex driving in San Francisco with its Cruise subsidiary rather than rack up miles mostly in suburban and closed-course driving as Alphabet’s / Waymo has done. The vehicles can learn the most in a dense city, as GM’s data shows its autonomous vehicle fleet of 180 Chevrolet Bolts encounters special situations such as emergency vehicles up to 46 times more often in San Francisco than in suburban Phoenix.
GM will expand testing to New York City in 2018. All GM autonomous vehicles will be pure electric vehicles (battery electric vehicles) as GM believes the autonomous vehicle hardware can be more easily integrated in a BEV than in a hybrid or internal combustion vehicle. GM will have two new BEV crossovers by 2020 and will have at least 20 new BEVs launched by 2023. A new battery platform in 2021 will help drive GM’s BEV costs down more than 30%. This reduction, plus eliminating a driver and GM gradually lowering its LIDAR costs to $300 from $20,000 presently, gives management confidence to say it can profitably make BEVs but also have a profitable TAAS business.
TAAS today accounts for only about 0.1% of vehicle miles in the United States, but as more autonomous vehicles enter the market, that percentage will only rise over time to a multi-trillion-dollar market. We agree with management that autonomous vehicles are the biggest thing since the Internet, but we think they may even be bigger because mobility affects everyone on the planet. GM sees its early autonomous vehicle service costing customers $1.50/mile, with GM’s costs just under $1.50/mile. By 2025, GM sees the price falling to well under $1/mile, as will the cost, based on the aforementioned reductions, and these vehicles will be used far more than a privately owned vehicle. We think annual miles of at least 70,000 per vehicle is possible.
The key to the cadence is safety and getting vehicles to take customers where they want to go. These two issues are where GM is wisely spending its time, since it thinks other issues, such as getting someone there cheap, fast, and entertained en route, are easier to address than the first two matters. Management repeatedly stressed that GM’s TAAS can have an economic moat. GM can scale fast when ready, since it’s the only automaker making autonomous vehicles in an assembly plant today; GM claims to be the only company with a fully integrated autonomous vehicle solution thanks to 15 capabilities such as high-definition mapping and routing, proprietary autonomous vehicle sensors, OnStar, and vehicle connectivity and data collection.
GM may be right, at least early in the autonomous vehicle game, but we are not sure it or any other player can build a moat in ride hailing; there will be too many players in the space. We think it is possible that a company like GM can initially offer better routing or a safer ride. Longer term, however, we see ride hailing as a highly commodified service with every player able to provide a fast, safe ride on the most optimum route at an affordable price, likely eventually well under $1/mile and even under $0.50/mile. The interior of the vehicle, which may radically change, may be the only differentiator besides services provided on board. Services can be a variety of things, such as virtual reality, e-commerce, media, or digital medical consultations. These services, the interior, and the ability to request a private vehicle will be a way for ride-hailing companies to vary pricing. GM’s telematics capabilities through OnStar may have a bigger role to play in this world, but time will tell. Management talked about eventually having a GM autonomous vehicle network with an extensive Internet of Things platform. GM intends to have a global TAAS business, including in Europe, despite selling Opel and Vauxhall to PSA in August.
GM talked about a first-mover advantage, but Uber and Lyft are already in the U.S. ride-hailing space, and there are many companies globally such as Didi Chuxing in China, Ola in India, Grab in Southeast Asia, Careem in the Middle East, Yandex in Russia, and Gett in Europe and other markets. We expect Tesla will be joining the autonomous vehicle ride-hailing game soon, and we suspect many other automakers will too; and Apple and Waymo are also possibilities. Despite a crowded space, we think GM is right to aggressively pursue TAAS business so as to avoid becoming a provider of low-value-added hardware (the vehicle) and nothing more. GM already knows how to make a vehicle, so there’s no reason it should let Silicon Valley startups take all the ride-hailing pie. Unlike Tesla or Uber, GM actually makes money and generates cash, so it can use those profits to build out more autonomous BEVs while also dominating the highly profitable full-size SUV segment with about two-thirds segment share in the U.S. The company also makes about $2 billion aftertax annually in China, the world’s largest auto market and GM’s largest market with nearly 4 million vehicles sold per year. It will be interesting to see if over time GM keeps going alone in ride hailing or seeks an alliance with a company such as Lyft to broaden its customer base. As a solo effort, GM in our opinion may market TAAS offerings via its recently created Maven brand rather than an existing brand such as Chevrolet.
There are plenty of reasons to be optimistic about GM’s prospects longer term as the company focuses on its most profitable areas, such as Cadillac, light trucks, GM Financial, and China. Adding a TAAS business with what we see as massive option value makes the GM story more interesting to us, and we remind investors that GM is not shy about buying back what we see as quite undervalued stock. In 2017 alone, GM will return about $7 billion to shareholders with about $5 billion of that from buybacks. GM is $9 billion into a $14 billion buyback authorization as of Sept. 30, and the dividend yield is currently about 3.5% versus about 1.8% for the S&P 500.