Despite the US operators’ big claims about “world-leading” 5G build-outs and investments, in fact they are being relatively circumspect about their capex spending (except when it comes to the C-band spectrum auction, of course). Compared to China’s massive expenditure on rolling out a gigantic network very quickly, the USA’s spending looks cautious, and is not set to jump significantly until 2022 onwards, though there is an uptick in 2021 compared to 2020.
Even from 2022, the operators are hoping to keep capex growth under control, even as they deploy 5G Standalone at scale, by leveraging the efficiencies of more open, cloud-based networks and by working with third parties, such as public cloud providers and tower operators, to spread or reduce some of the upfront costs.
The tower companies often provide the best insights into operators’ deployment patterns, especially in the USA, where the majority of sites are owned by neutral hosts rather than the telcos themselves. Even before the big three operators held their investor conferences last week, some of the tower giants were seeking to manage expectations among their own shareholders.
“This year is not going to be the robust year that some of my colleagues are predicting,” Alex Gellman, CEO of towerco Vertical Bridge, told a conference called Metro Connect recently, while Brendan Cavanagh, CFO of the USA’s third largest towerco, SBA, told the firm’s earning call: “Our outlook for 2021 does assume a pace of activity that is increasing throughout the year.” He said the assumption was that build-out would be “a little bit slower at the beginning of the year, and it picks up in the latter half of the year.”
Overall, the operators aim to spend just slightly more in 2021 than in 2020, which represented a dip from 2019. According to GlobalData analysis, the nine publicly listed wireless and wireline operators in the USA spent $67bn between them in 2020, down by $3bn or just over 4% on 2019.
This decline was mainly attributable to AT&T’s cost cutting, and was more generally driven by the pandemic and associated recession, as well as technology cycles. The big two operators, in particular, had invested heavily in 2018 and 2019 to have sites and fiber in place for 5G and then to build out their first networks. However, they have only just acquired significant amounts of the midband spectrum that is so central to the 5G business case, so a new wave of deployment will start later in 2021.
Both operators are keen to move quickly to open, software-driven architectures that will reduce overall cost (see separate item), and so may delay some build-outs until those platforms are fully commercial and tested, which could see some pent-up spending from 2022, but also a lower 5G investment than would have been the case using traditional architectures.
Their need to adopt the most cost-effective platforms possible, and to delay any spending that is not essential to competitive performance, will be accentuated by the huge sums they spent in the C-band auction.
“Big spectrum purchasing years tend to reduce spending over the short term on network infrastructure,” Jim Hyde, CEO of network construction company ExteNet Systems, told Metro Connect.
Those C-band sums have evened the spectrum position of the big two against T-Mobile considerably (though not completely), but because TMO acquired its 110 MHz of 2.5 GHz spectrum as part of its Sprint purchase, it is not incurring the capex burden that its main rivals have taken since the auction. TMO looks well-positioned to deploy a competitive network, with the flexibility of its early 5G core implementation, without adding significantly to the Sprint/TMO capex total of 2020 (which even when considering wireless alone, is lower than AT&T’s or Verizon’s).
But TMO has never been a high-rolling operator. It has traditionally been highly targeted in its capex, and skilled at investing where new business has already materialized, rather than speculatively focusing on the shiniest new technology (by contrast with Sprint, which was a technology experimentalist in its heyday, a strategy that rarely helped its profitability). TMO, for instance, built 4G out later than its rivals, but aligned it well with demand, and made up for lack of technical wizardry with excellent marketing. In 5G, it has uncharacteristically taken a lead with the Standalone deployment around the 5G core, but does not appear to be adopting radical approaches to the cloud architecture or vendor supply chain, unlike its rivals with their cloud-native, open network strategies (see separate item).
Financial analysts at New Street Research wrote in a recent report about towerco prospects: “Our call has been a little premature: The ramp in activity that we expected from T-Mobile has materialized slower than we expected, resulting not only in slower-than-expected growth in 2020, but also in 2021.”
Wall Street analyst firm MoffettNathanson has described the USA’s 5G spending as a “giant U shape” from the start of 2019 to the end of 2022, with early 2021 being the low point. “While the idea of a ‘U’ is fairly consensus by now, the cadence of the recovery and ultimate level it reaches has been less clear,” they wrote in a recent investor note following SBA’s quarterly results, which disappointed shareholders. “SBA’s initial 2021 outlook suggests a relatively long, modestly improving curve at the bottom of the ‘U,’ with expected growth coming in shy of expectations.”
The US operators’ capex highlights:
- T-Mobile planned to spend between $5.9bn and $6.2bn last year, excluding the spending of Sprint, which it acquired in mid-2020. This year, it aims to spend between $11.7bn and $12bn, and as this excludes spectrum purchases, most of the sum will be available for 5G build-out and the process of finishing integration of the Sprint network and spectrum. The company said most of the Sprint-related capex would be completed in 2021-2022, and so overall capex will fall from 2023.
- In 2020, AT&T spent about $19.7bn in capex, which was down on its 2019 figure, and in 2021 it plans to spend about $21bn. Of course, unlike TMO, AT&T (and to a lesser extent Verizon) have wireline networks to support as well as mobile, along with other platforms such as TV and media. The $23.4bn AT&T spent on C-band spectrum is not included in its capex total but is likely to constrain its 5G expansion efforts, especially as it is also trying to reduce its debt mountain. AT&T expects to spend between $6bn and $8bn specifically on building out 5G in its new C-band spectrum, to reach 70-75m people in 2022.
- Verizon amassed capex of $18.2bn in 2020 and in 2021 will spend between $17.5bn and $18.5bn excluding its huge C-band spectrum fees. “Our 2021 capital program includes capital to fund advanced networks and services, including expanding our core networks, adding capacity and density to our 4G LTE network in order to stay ahead of our customers’ increasing data demands and deploying our 5G network, transforming our structure to deploy the Intelligent Edge Network while reducing the cost to deliver services to our customers,” said Verizon in its 10K filing. The operator said it would cover 250m people by 2024 with its C-band build-out, spending an extra $10bn over three years in the process.
- Dish has promised to build a national 5G network for just $10bn by using its greenfield status to adopt open, software-based network architectures from day one. Many analysts have doubted this sum, but the company is keeping its wireless capex projections low at $390m. It has not confirmed how much it spent specifically on wireless last year, but it was between $250m and $500m. It is obliged to spend about $340m in 2021 for minimum payments related to tower contracts and other 5G deployment commitments. Of course, Dish may end up spending far more than it is publicly projecting. In its 10K filing, it wrote: “We anticipate expenditures for our 5G network deployment to increase substantially throughout 2021 as we ramp up the build-out phase.”
Verizon acquires Montana-based Triangle Mobile:
Large telcos snapping up small rural and regional operators used to be a common feature of the USA’s M&A landscape, and both AT&T and Verizon went through a phase of extending their reach and customer base in this way. There are still many small fixed and mobile operators left, as the membership list of the Competitive Carriers’ Association (CCA) demonstrates, though their business models are being challenged by the charges to roam into national operators’ networks, especially for 5G, and in some cases, by the requirement to swap out Huawei equipment.
We may, then, expect a new wave of acquisitions to fill out the footprints of the major operators. Verizon may have set a new ball rolling with its purchase, announced last week, of tiny Montana-based Triangle Mobile, which will bring about 3,000 new customers to Verizon’s total. Verizon will not buy the smaller firm’s landline or fixed wireless systems, nor its midband spectrum assets in AWS-1 and AWS-3 bands.
Verizon will assimilate Triangle’s LTE network in low band spectrum, with a view to improving its overall rural coverage, but needs to rip out a Huawei-supplied CDMA network.
“While Triangle’s LTE network is already built to Verizon’s specifications, Triangle’s CDMA service includes equipment manufactured by Huawei. Verizon Wireless does not use Huawei equipment in any of its network infrastructure and, among other things, will replace all Huawei equipment before integrating Triangle’s CDMA network into Verizon Wireless’s network,” Verizon wrote in a filing with the FCC. “Until Verizon technicians have removed and replaced all Huawei equipment, Triangle’s management will continue to operate the CDMA network.”
Triangle is a member of Verizon’s LTE in Rural America (LRA) program, which includes about 18 operators serving a total of about 2.5m people in 13 states.
Triangle follows Bluegrass Cellular in Kentucky, Chat Mobility in Iowa and Blue Wireless in New York and Pennsylvania into the Verizon fold as the operator seeks to improve its spectrum breadth and depth and fill holes in its coverage map.