Long road ahead for investors in self-driving cars

Waymo’s robo-taxi
Waymo’s self-driving taxi will soon be a familiar sight

Electric vehicles (EVs) seem stuck in the slow lane. But the outlook for fully self-driving cars is compelling. Fasten your seatbelts: it’s going to be a bumpy ride, says Rupert Foster.

Electric vehicles (EVs) and autonomous – self-driving – vehicles (AVs) are rarely out of the news. But do these exciting long-term investment themes or mega-trends really justify the hype? We read a great deal about Tesla, while most of us have heard that Google is dabbling in self-driving cars. Sometimes electric-battery makers and software providers appear on investors’ radar screens. But there have been few genuine success stories for investors to get their teeth into. Will that ever change or are these sectors just futuristic fantasies?

A short-circuit in the EV story

Let’s start with EVs. Investors focus on two markets: hybrids (cars containing a petrol engine and a battery) and fully electric vehicles. The huge demand for Tesla’s Model 3 in 2016 – the firm received 500,000 orders – changed perceptions of the size of the potential market. The EV morphed from virtue-signalling purchase for the eco-sensitive liberal elite to a must-have toy for hipsters and suburbanites. The likes of Volkswagen and Nissan climbed on the bandwagon, launching their own versions. There was soon talk of mass-market adoption by 2020.

However, momentum has since stalled. Tesla has run into awful production problems and its Model 3 pre-orders will not be fulfilled for years. Goldman Sachs and JP Morgan now anticipate that pure EVs will comprise only 7% of the global car market by 2025.

The key variable, moreover, is demand – not in the West but in China. This is now the world’s largest car market, with annual sales near 30 million – the joint total of the US and Western Europe. China intends to become the global leader in EVs by 2025. It has an added incentive to encourage EVs because of the horrendous pollution in most large Chinese cities. JP Morgan estimates that by 2020, 59% of all full EVs in the world will be sold in China. Nonetheless, by 2025, they will still only comprise 12%-15% of the total Chinese market for new cars, while traditional internal combustion engines will have around 80%.

EVs in the slow lane

As for the West, 50% of all EVs sold in the US are sold in California. Its government has stipulated that 50% of all ride-sharing vehicles must have zero-emissions by 2025 and has invested heavily in building infrastructure to aid battery-charging. Meanwhile, some industry players worry that the cost of building the power industry infrastructure to provide for the charging needed for mass-market adoption of EVs will be prohibitive. Goldman Sachs estimates that $6trn will be needed to be spent globally on power generation and grids to allow for full worldwide mass market adoption.

However, this is unlikely to be a problem. The Californian model is hardly the norm and with the very gradual take-up of full EVs currently anticipated – the market is growing at 1% a year – the infrastructure should be able to keep up. The odds are that home charging will be EVs’ main source of electricity, as cars in the developed world spend 80% of their time at home turned off and only 7% actually on the roads; most journeys are under 40 kilometres.

The upshot for investors is that the pot of gold at the end of the rainbow is retreating ever further into the distance. There is unlikely to be a rapid acceleration in EV adoption unless the Chinese government redoubles its efforts. Tesla remains the key player and may well become the dominant brand in EVs, but this is likely to mean a 10%-20% market share. But to justify its share price, EV penetration would need to reach 30%‑35% by 2025. That appears nigh-on impossible. All its rivals will have cheap EVs on the market by 2020, capable of travelling 300 kilometres, when Tesla may finally have overcome its Model 3 production issues.

Call a “robo-taxi”

The other key EV beneficiaries should be the main battery companies and their suppliers. However, production and development issues have again hampered progress and allowed the Chinese to catch up with leading players such as LG Chem and Samsung SDI. The Chinese are likely to increase capital spending aggressively and crimp long-term margins in the industry – thus making it less attractive to long-term investors.

If EVs mark a gradual evolution in the car market, AVs are a revolution, as they remove the human element entirely. That should not only save many lives – 1,750 people died on British roads last year – but also help us use our time far more efficiently. The average American spends 100 minutes a day in their cars.

Autonomous driving systems have already appeared in high-end cars to help with parallel parking and negotiate traffic jams. These are classified as Level 2 or 3 AVs. At Level 4, the vehicle can move around without a driver in certain conditions. Level 5 means that the machine is in charge all the time. The industry used to focus on sensor technology, but the crux now is artificial intelligence. The main players are all competing to get their trial vehicles on the road for as long as possible to learn how to drive perfectly.

The current leader is Google subsidiary Waymo, which has conducted more than five million miles of test-driving with its Level 4 and 5 AVs. This is streets ahead of its nearest rival, General Motors (GM). The key safety metric used by the industry is the number of “disengagements” per 1,000 miles driven – a disengagement being when the driver has to take back control of the vehicle. For the past three years, Waymo has had fewer than one disengagement per 1,000 miles.

This has enabled Waymo to announce that it will launch its first “robo-taxi” service in Phoenix, Arizona, before the end of this year; GM hopes to launch its service next year. Robo-taxis will be very popular AVs. Without a driver, the cost of any journey will be reduced by about 40%, allowing robo-taxi operators to undercut current taxis. They will save companies money on deliveries, too. Robo-taxi take-up could be extremely fast, depending on their safety record. The media will tend to hype up a death from a driverless car, even though AVs are less dangerous than traditional vehicles. However, at the current rate of progress, we could see mass-market adoption in the West in two to three years.

The AV car-ownership model

Robo-taxis are only the beginning. Once AVs are widely available, most people will rent, rather than buy them, as is now the case for so many products, particularly music and television programmes. People are used to making a monthly payment for their cars. In the AV world, it would appear likely that the AV companies (with the help of finance companies) will keep control of all the vehicles and provide all maintenance and insurance costs. Consumers will take out an AV subscription, which will allow them to use an AV whenever they need one – or whenever their subscription allows.

US consumers only drive their cars 7% of the time, so a subscription model would allow the vehicles to be used by other people or for other purposes the rest of the time. This implies a subscription cost significantly below the current cost of a monthly payment on a car. However, AV companies should still be able to make enough to cover the cost of offering insurance and maintenance, and earn a bit extra on these services, if they can keep their car usage levels above a certain point.

The virtuous circle of the network effect is key here. As more people sign up to an AV service, the company will be able to keep the price of a subscription reasonable, thus enticing more customers, and so on. Occupancy of around 50% should give the AV company high profitability. A well-established network is hard to dislodge (witness Amazon’s dominance), so gaining a foothold early is crucial. The winners will be able to improve their cars’ AI by drawing on the increased usage of their vehicles, which in turn will make their offering more appealing.

Who will benefit?

We are now only around three or four months away from the launch of robo-taxis. Level 5 AVs may not become a mass market feature until 2025, but Waymo and GM are among the likely beneficiaries in the next few years from networks and monitoring research, as companies build on robo-taxis.

Waymo is currently a small part of the valuation of Google as a whole. But, along with YouTube, it presents the best opportunity for Google to grow significantly while the search business slows. GM has worked hard on its GM Cruise business and it is likely to reap the dividends. However, the company has a large legacy vehicle business focused on US light trucks. Its steady demise will offset some benefits from GM’s AV success.

Then there’s Aptiv, an offshoot of GM’s old car-parts business, Delphi. Managers have cleverly crafted a business that has the best exposure of any large listed company in the US to the AV (and also to the EV). As a result, it has a premium price-earnings ratio of 21. It is growing at less than 10% a year, so the valuation is not cheap, but as the AV industry ramps up, its software, power distribution and sensing businesses look well placed. It has already recruited 6,000 engineers to drive technical development and sales.

Lastly, I would argue that Baidu, China’s version of Google (and Waymo), will see good upside in time from its monopoly of the Chinese AV industry. The Chinese government is extremely competitive in AVs and can brush aside domestic safety concerns easily. The only question with Baidu, however, is given the global geopolitical backdrop, how much will it be allowed to dabble in developed markets?