As a final, eleventh-hour barrier to the merger of T-Mobile USA and Sprint was removed, with an improved deal for TMO’s parent Deutsche Telekom, Dish Network should be free to start deploying its 5G network. The company, whose roll-out depends on assets which will be divested as a condition of the Sprint/TMO deal, was standing by its ambitious low-capex, rapid deployment targets.
The amended deal was struck following demands for renegotiation from DT, based on the falling market value of Sprint since the original merger was agreed. Last week, it seemed that the fate of the deal once again hung in the balance even after seeing off a legal challenge from a group of US states. But an amended agreement was made quickly, allowing DT to own about 43% of the New T-Mobile shares, as opposed to the 42% it previously would have held. SoftBank will hold 24%, instead of the 27% it would have gained under the previous terms, while the remaining 33% will be held by public shareholders.
That puts the spotlight firmly on Dish. After years of speculation about whether Dish, and its chairman Charlie Ergen, really wanted to build a national cellular network, or were just buying up spectrum for trading purposes, the proof will now be in the pudding. Dish could still build a network and then seek to sell it to another operator, such as a cable consortium. To achieve an economical national build-out and competitive cost base, it will almost certainly have to form an infrastructure partnership, perhaps with the New T-Mobile itself. So it will not be a conventional operator, nor will it have a conventional network – it has already issued several RFPs for various elements of its platform, citing modern architectures like Rakuten’s as its role model.
First, it needs to get its hands on the assets, such as Sprint’s Boost Mobile prepaid customer base and some divested sites and spectrum. And it needs to firm up a business model that will enable it to be fully competitive with the established MNOs, compensating for the decline in its pay-TV revenues while also, potentially, improving ARPU in its traditional video base by adding new wireless services. But it will also be aiming for new customers, using Boost as a starting point.
Dish had already started building an NB-IoT network in some of its existing patchwork of spectrum, which is worth about $22bn and includes former mobile satellite licences acquired from bankrupt satellite service providers, and licences bought in various auctions including the monster AWS-3 sale. However, this network largely seemed to be there to satisfy FCC roll-out obligations, while the new plans, bolstered by the Sprint/TMO assets, are far more ambitious.
They will now be able to go ahead as soon as April – Mike Sievert, president of T-Mobile USA, who will be CEO of the New T-Mobile, has stated that April 1 will be the target date to form the new company. Dish is now committed to deploy a 5G network covering 70% of the US population by June 2023.
It has previously said it expects to spend $10bn on that 5G network. On its fourth quarter earnings call last week, however, it said it expects to spend only $250m to $500m on wireless capex in 2020, which suggests it is angling for a far lower capex total than previously indicated, harnessing a combination of cloud-native architectures and shared infrastructure.
Analysts have been consistently sceptical about Dish’s ability to deploy at very low cost, given the huge market it needs to cover and relatively high equipment and labor prices in the USA. Craig Moffett of MoffettNathanson said in a recent client note that he expects build-out costs to be considerably higher than Dish claims, since virtualization will only impact a small portion of the total expenditure. “Equipment is only 20% or so of the cost of building a network,” he wrote. “The rest – labor, tower leases, powering, zoning and siting, and backhaul – don’t benefit from virtualization.”
Moffett noted that AT&T and Verizon have each spent about $200bn on building their networks over the past decade and have virtualized significant portions, but still spend more on maintenance than Dish expects to spend on building a brand new network. Of course there are efficiencies for companies that have no legacy equipment to integrate or upgrade, but he questioned Dish’s pledge to build 50,000 macrocells, since that would imply a cost per site of $250,000.
“Where in the budget are the literally hundreds of thousands of boosters and small cells for airports and stadiums and shopping malls and other public buildings that are required for a competitive network experience?” Moffett asked.
Another analyst firm, New Street Research, was more positive, as it focused on Dish’s potential to reduce the cost of delivering mobile data compared to rivals. It expects an all-5G network to achieve 30% to 35% lower cost-per-gigabyte than the incumbents, which “should enable Dish to take meaningful share in wireless,” New Street wrote. It also speculated that operating partners like cloud or cable companies could “serve as sources of low cost financing”, as well as infrastructure sharing.
Ergen has said Dish will need to raise capital to afford the $10bn price tag he has put on the 5G roll-out, but he will be looking for infrastructure partners to reduce the capex cost. “We’re probably not going to build towers and we’re probably not going to lay a bunch of fiber,” he has said in the past, nor does he expect Dish to build its own cloud and edge compute capability. To help fund its network, Dish is proposing a rights offering of $1bn.
New Street expects financing efforts will take place over the next 12-18 months, but only after initial stages of network planning, business planning and partnerships are complete.
“We suspect that the company will seek to finalize vendors and perfect its network plan before moving ahead, as the network is the most important underlying component for the value creation opportunity,” wrote the market watchers.
The RFPs already issued cover end-to-end network elements, software, deployment services, physical network components and transport. Tom Cullen, Dish’s EVP of corporate development, said recently that the firm had received “dozens and dozens” of responses to its three RFPs, but has not finalized any deals yet.
“Because we’ll be cloud-native in that network we can just do things faster, better cheaper,” Ergen said Dish’s third quarter earnings call. “We don’t have the legacy of 2G, 3G and 4G. The cost structure goes down across the board.” He added: “The vast majority of capex for the incumbents is not for 5G, it’s to maintain the legacy. We don’t have that cost going forward.”
In the first phase, Dish aims to keep the network asset-light, saying that, since it would not have large numbers of customers on day one, as a new entrant, small cells will be a second-phase priority. It will look to minimize equipment on macro towers – and therefore costs – by moving a substantial proportion of baseband processing to the cloud and stealing a march on rivals by creating a cloud-centric, highly virtualized platform that can behave in an agile manner. The company has strongly hinted that it will improve its costs further by using some non-traditional suppliers, again referencing the Rakuten deal.
Rakuten has said it can build its network with a capex bill that will be 40% lower than for a traditional, non-cloud network of the same scale. But analyst Roger Entner of Recon Analytics told FierceWireless that Dish’s estimates are still over-optimistic. “By doing a virtualized network based on open RAN, Ergen can realize massive 40-60% reductions, but not 100% reductions,” he said. “It’s not a free network. You’re not repealing the laws of gravity.”
Even Ergen, despite Dish’s bullish comments about low capex, seems prepared to go higher. “Any good business plan today can raise way, way more than $10bn,” he said. He even hinted that he might invest personal money, when asked that question by an analyst on the earnings call. He said he could not comment for legal reasons, but added: “I have liquidity.”
Late last year, Dish announced two new senior hires for its mobile team. Marc Rouanne, formerly head of networks at Nokia and chairman of the board at Alcatel-Lucent, joins as chief network officer; while Stephen Bye, formerly president of regional US operator C Spire, and CTO of Sprint, comes on-board as chief commercial officer. “Marc will architect the network and Stephen will commercialize it,” Erik Carlson, Dish’s president and CEO, said.
“A visionary and advocate for software-defined and natively automated networks, Marc will be the ideal partner and leader as we design and build the nation’s first cloud-native 5G broadband network,” said Ergen, in a statement. Rouanne was pivotal in ensuring that Nokia became software-driven at an earlier stage than its rivals, introducing architectures such as Liquid Net before the industry was widely talking about open, virtualized platforms.
Interestingly, Bye also headed up cableco Cox Communications’ development of its mobile platform and services in the early years of the century. Cox exited that business in 2011, after grand plans by the four leading cable operators of the time, to partner with Sprint and build a near-national broadband wireless network between them, fell apart. However, several of those cablecos are now re-entering the mobile market with better technical enablers and partnerships under their belt, and the likes of Comcast and Charter have acknowledged the valuable lessons learned from that earlier experiment.
With Dish trying to pull of the same trick of transforming from a video-centric business to a mobile and quad play model, Bye will bring important insights – Dish, after all, has better tools than Comcast, Charter and Altice, thanks to its extensive spectrum, but it lacks real world experience in cellular services, and its core service and brand are suffering more than those of the leading cablecos.
But Dish is certainly entering the market at a time of change and of rising competition (including from the cablecos). “The United States has the highest prices in the world for mobile services,” Ergen said. “It’s ripe for disruption.”
Dish may find itself helping to bring down those prices rapidly in the 5G era if it responds directly to New T-Mobile’s promised entry level tariff of just $15 a month. One of the assets Dish will buy in the event of that merger is Sprint’s Boost brand and customers, which has targeted the lower end and youth consumer market which has also been most susceptible to TMO’s disruptive Uncarrier pricing.