The widespread commercialization of autonomous vehicles, also called self-driving cars, relies heavily on the expansion of wireless communications capacity, according to research by Morgan Stanley (MS), as reported by Barron's. The reason is that these vehicles must have reliable ultra high speed connections to the internet to exploit their full functionality, as is the case with other smart devices that comprise the so-called Internet of Things (IoT). Morgan Stanley writes that, once the market for autonomous vehicles really takes off, this would represent "a sea change for Telco services like the arrival of the smartphone (recall that mobile data accounted for just 5 – 7% of service revenues in 2006)."

Potentially big winners include, per Morgan Stanley: wireless telecom providers China Mobile Ltd. (CHL), Vodafone Group PLC (VOD) and Verizon Communications Inc. (VZ); wireless communications tower operators American Tower Corp. (AMT) and Crown Castle International Corp. (CCI); networking specialist Cisco Systems Inc. (CSCO); and digital telecom supplier Qualcomm Inc. (QCOM). Semiconductors made by Nvidia Corp. (NVDA) are used by various makers of autonomous vehicles, per Investor's Business Daily.

Rationale for Picks

Morgan Stanley assembled a list of 25 stocks that it calls the "Connected Autos 25." This list was prepared taking a long-term perspective, and thus the 12-month price targets for these stocks were irrelevant, as well as whether these stocks were rated as underweight or equal weight for the near term. The report, entitled "Autonomous Cars & Telcos: We're gonna need a bigger (5G) pipe," was dated May 7, 2018, and lists equity analyst Emmet B. Kelly as the lead contributor. (For more, see also: Self-Driving Cars to Benefit These Stocks: Morgan Stanley.)

China Mobile is attractive to Morgan Stanley on the basis of low valuations in the Chinese telecom sector due to concerns about the level of capital expenditures (capex) needed for an upgrade to advanced 5G service. Vodafone is the biggest wireless operator in Europe, with the biggest capex budget. Verizon is expected to be the 5G leader in the U.S., and "offers fleet management and connected vehicle solutions globally." The report estimates that the collective market cap of global telecom providers will jump from about $1.5 trillion to roughly $3.7 trillion in 2050 under the bull case, which is described in the next section.

"Cisco's strength is enabling connectivity of more devices onto the network. Through its acquisition of Jasper Technologies, it has a software company that eases one of the largest pain points for carriers / auto manufacturers." Qualcomm's radio technology makes it "an enabler of 5G." It also has "modern chips" and other technology for sale and licensing into the auto market.

The successful buildout of a 5G wireless network requires the ability to reach users through so-called edge technologies. This is where companies such as American Tower and Crown Castle come into play. They build and operate the transmission points from which wireless users connect to the network.

'iPhone on Wheels'

In Morgan Stanley's base case, which they subtitle "iPhone on Wheels," they project that telecom providers worldwide will derive over $200 billion of annual revenues from autonomous vehicles in 2050. To produce this forecast, they started with an average monthly mobile data charge of $25 for consumers today, assumed a compound annual growth rate (CAGR) of about 3%, and assumed 300 million autonomous vehicles on the road globally by 2050. This represents a revenue increase of about 15% from today's totals for these companies. 

In their bull case, the annual revenue figure grows to at least $1.4 trillion. For this estimate, they took a different tack. They assumed 30 trillion miles driven annually worldwide (an average of 100,000 miles across 300 million vehicles) at the highest levels of autonomy (levels 4 and 5; see next section), telecom revenue of about 1 cent per mile, and a CAGR of about 5%. This equates to a revenue boost of more than 50% from today's figures for global telecom companies.

The report compares the average iPhone today to a concept of the average autonomous vehicle in 2050. In an average year the latter will produce, per these estimates, 3,300 times more data, between 50 and 500 times more wireless usage, and up to 300 times more wireless data transmission. Since 5G networks are expected to deliver download speeds that are at least 30 times faster than under 4G, they represent a critical underpinning for the autonomous vehicle market, Morgan Stanley argues.

Autonomy Levels Explained

Level 3 technology can guide the car on limited-access highways, but it requires a human driver to be ready to take control at any moment. This is the most advanced automation currently available to consumers. Level 4 cars are designed to operate autonomously on limited-access highways, but a human driver must take control elsewhere. Level 5 vehicles will be the only true driverless cars, designed to operate anywhere without human intervention, as described by Car and Driver.  (For more, see also: How Big Automakers Will Speed Past Tesla.)

Challenges Ahead

Building out a 5G network worldwide will be a very expensive proposition, however, estimated by Morgan Stanley at a cumulative $1 trillion through 2050. Among other issues that remain, they say, are network coverage and capacity, as well as data privacy and security concerns.

Crashes involving autonomous vehicles produced by electric car maker Tesla Inc. (TSLA) or operated by ride hailing service Uber Technologies Inc. have produced a public relations setback for the technology. Shares of chip supplier Nvidia have been sensitive to these negative developments.

As pointed out by Scientific American, much of the data gathered so far on the performance and safety of autonomous vehicles reflects operations in ideal weather in western U.S. states. Moreover, that article continues, this data was collected mostly "on unidirectional multi-lane highways, where the most important tasks are staying in the car's own lane and not getting too close to the car ahead. Automated cars are rather good at those tasks—but then again, so are humans."