The author is a Reuters Breakingviews columnist. The opinions expressed are her own.
By Jennifer Johnson
LONDON, June 2 (Reuters Breakingviews) - Companies usually rejoice when customers want more of what they’re selling. In telecom land, though, meeting future demand typically requires massive upfront investments from the likes of Verizon VZ.N and Vodafone VOD.L, which haven’t always earned an attractive financial return. The sector’s CEOs may therefore be quietly happy that mobile-data growth seems to be falling – and apprehensive about the prospect of artificial intelligence eventually sending it spiralling again.
It’s been obvious for several years that fixed-line broadband demand has stopped rising as rapidly as it once did. According to research outfit Enders Analysis, internet traffic volume growth across most developed markets is now around 10% to 15% a year, compared with 30% or more before the pandemic. This deceleration reflects the fact that most rich-world households are by now already heavy users of streaming services like Netflix NFLX.O and Amazon.com’s AMZN.O Prime Video. In other words, the shift to internet-based TV was a one-off step-change, rather than an everlasting tailwind.
Something similar now seems to be happening with mobile data. Enders’ analysts reckon some European markets may have seen annual rates of increase in mobile-data demand as low as 5% in recent quarters, though there is significant variation between countries. Forecasts compiled by Analysys Mason, a consultancy, suggest the trend is playing out worldwide. It predicts that annual growth in global mobile-internet traffic will fall to 15% in the next few years, compared with 23% in 2023. That's a sharp comedown from 2018 levels, when volumes were roughly doubling each year according to the same consultancy.
It’s possible that many Westerners already have all the mobile data they need. There is only so much time in the day to watch Netflix or scroll social media sites like X. Practically everyone who wants a smartphone already has one.
Slower demand growth may, counterintuitively, be a relief for telecom bosses and investors. The key reason is that past binges on spectrum and equipment, to meet rising data requirements, have often delivered poor returns. According to a 2023 report by professional-services firm PwC, mobile network operators made a 2% to 3% return on assets over the preceding five years – a period which spanned 5G and 4G-related capital expenditures. That’s akin to a regulated utility asset, but without the comfort blanket of a monopolistic market and inflation-protected revenue.
Another way to look at the problem is the return on invested capital $(ROIC)$, usually calculated as post-tax operating profit divided by the cumulative cost of the assets a company uses to run its business. For major U.S. and European telcos, the average ROIC this year will be 7.3%, based on analyst estimates gathered by Visible Alpha. That’s roughly in line with the sector’s weighted average cost of capital, as calculated by New York University’s Aswath Damodaran, implying that as a group the companies are not meaningfully creating any shareholder value. This is reflected in valuations: the same companies on average trade at a measly 7 times 2025 EBITDA.
One hope is that peak data growth will change all that, by allowing the telcos to avoid a massive new round of value-destroying capex for the foreseeable future. That’s roughly what Wall Street brokers expect right now: physical-asset investments will stagnate for major telcos over the coming years, according to Visible Alpha figures. The positive flipside of lower investment is that free cash flow, and therefore returns, should rise.
The wildcard is AI. To date, most of the burden from large language models, from OpenAI and its ilk, has fallen on Big Tech-owned data centres, rather than consumer-facing 5G networks. That could all change if applications, like so-called AI agents, really take off. Imagine millions of Vodafone and T-Mobile TMUS.O customers constantly running ChatGPT or Alphabet’s GOOGL.O Gemini app in the background, performing administrative tasks that a personal assistant would usually handle while also condensing and summarising emails and reports.
Video, audio, and image-creation apps could be particularly data-intensive, necessitating more 5G and even 6G investments over time. It’s also possible that autonomous vehicles, which require sensors and cameras to send information to the car’s central computer instantaneously, could end up using 5G networks.
As is stands, that’s more of a risk than an inevitability for telcos. But it’s worth getting prepared for such an eventuality. Telecom and broadband bosses, particularly in Europe, have moaned about having to handle all the extra Netflix-induced traffic without the streamers contributing anything towards the upkeep of the networks. AI could, eventually, cause a re-run of that debate.
Follow @jenjohn_ on X
Major telecom group's meagre average return on invested capital https://www.reuters.com/graphics/BRV-BRV/jnpwlndqmvw/chart.png
Analysts are expecting flattish capex for major telco companies https://www.reuters.com/graphics/BRV-BRV/lgpdxyymlpo/chart.png
(Editing by Liam Proud; Production by Pranav Kiran)
((For previous columns by the author, Reuters customers can click on JOHNSON/Jennifer.Johnson@thomsonreuters.com))
免责声明:投资有风险,本文并非投资建议,以上内容不应被视为任何金融产品的购买或出售要约、建议或邀请,作者或其他用户的任何相关讨论、评论或帖子也不应被视为此类内容。本文仅供一般参考,不考虑您的个人投资目标、财务状况或需求。TTM对信息的准确性和完整性不承担任何责任或保证,投资者应自行研究并在投资前寻求专业建议。