Implicit Questions about Payback from 5G and FTTH

Though there still is no firm consensus about the number of firms that a home broadband market can support, the math suggests that in markets with one or more smaller specialist providers, as many as five firms could sustain themselves. When competition across a full market is necessary, it is possible that, in some markets, that number drops to about three.

Globally, observers note that take rates for the new FTTH services often run in the 20 percent of homes passed up to 40 percent of homes passed range, with the threshold for a payback generally being somewhere around 35 percent on a blended basis.

In other markets there are two ubiquitous providers and several small specialists who serve only parts of a city. It is too early to know how the long-term market will shape up, as most of those new contestants are just entering the market now, or have operated for only a handful of years.

Still, those possibilities far exceed older thinking about access markets, when it was assumed only a single provider could sustain itself. In some markets, we already see that two access providers, competing on a whole-city basis, can sustain themselves.

Many other large questions, such as the role of wholesale mechanisms, open networks, the role of mobile versus fixed networks and the mix of revenue from consumer and business customers, all are open questions at the moment. Nor can we rule out some eventual move, in some markets, back to monopoly regulation of fixed network access services.

A related question is the ultimate shape of mobile service provider markets as well. At the moment, many observers believe markets with four leading mobile operators are unsustainable. Those observers believe markets will stabilize only when the number of suppliers is reduced to three (or perhaps fewer, in some cases).

Much of the debate over "fair share" mechanisms in the European Union countries seems to revolve about support for mobile networks rather than fixed networks.

But observers might also have questions about the sustainability of private equity investments into fiber-to-home networks as well, and for similar reasons: payback on potential investments.

According to consultants at Bain and Company, private equity investors have poured about $32 billion into building fiber-to-home networks between 2019 and 2022. It is not clear whether that sum also includes the cost of acquiring copper access assets that are then rebuilt using FTTH, though.

Payback can be obtained at lower levels when construction costs are low, such as in rural areas and when aerial construction is possible, rather than underground. Payback periods also are affected by other issues such as the availability of wholesale transit or access fiber, ability to pull new cables through widespread duct infrastructure, government subsidies and so forth. 

The point is that profitability often hinges on local conditions, including the level of expected competition from other internet service providers. 

source: Bain and Co. 

Other studies and transactions might confirm the risk involved. In many cases, before an FTTH network actually is built, a firm will buy an existing telco service provider business. That can add $1,000 per home to the payback model for an eventual FTTH build. 

Metric	Description	Example
Construction costs	The cost of building and deploying an FTTH network, including the cost of materials, labor, and permits.	$1000 to $1200 per location passed; perhaps $2,000 - $4,000 per household if existing copper assets are purchased.
Asset purchase costs	The costs associated with purchasing the necessary equipment and infrastructure for an FTTH network, such as fiber optic cables, routers, and switches.	$500 - $1,000 per household
Take rates	The percentage of households in a given area that subscribe to an FTTH service.	40% - 60%
Demographics	The characteristics of the population in a given area, such as age, income, and education level, can impact the demand for FTTH services.	Higher-income and more educated households are more likely to subscribe to FTTH services.
Housing density	The number of housing units per unit of land can impact the cost of deploying an FTTH network and the number of potential subscribers.	Higher housing density can make it more cost-effective to deploy an FTTH network.

Some recent transactions, especially those made by private equity, have involved purchase of existing telco assets, with the intention of building FTTH to change the revenue model. But that also can effectively double hurdle rates for earning a financial return, all other things remaining equal. 

In other words, instead of building a greenfield FTTH network for about $1100 per passing, the acquisition cost, plus FTTH cost, can effectively double the payback hurdle. 

Transaction	Date	Buyer	Seller	Asset Description	Number of Homes	Cost per Home
Apollo Global Management to acquire Lumen's copper access assets	2022-07-25	Apollo Global Management	Lumen Technologies	Copper access assets in 20 states	10 million	$600
KKR to acquire Consolidated Communications	2022-02-01	KKR	Consolidated Communications	Copper access assets in 21 states	1.3 million	$900
EQT to acquire Frontier Communications	2020-10-09	EQT	Frontier Communications	Copper access assets in 25 states	2.4 million	$1,000

Will Access Network Disaggregation Increasingly Take the Form of Joint Ventures?

There is one element beyond “consolidation” or “asset shuffling” that is not yet a significant development on this chart of major U.S. telco acquisitions since 1985: deconstruction, deconsolidation or disaggregation of functions. 

The chart, for obvious reasons of showing only formal asset ownership of brands, does not show internal or structural changes across the industry to separate application creation from asset ownership; with revenue flows following. 

source: Wall Street Journal, Seeking Alpha 

The chart does not show shifts in business strategy beyond “gaining scale” and does not show the acquisition of any mobile service provider assets, either. 

source: Quexor Group 

Beyond shifts of asset ownership, connectivity providers have moved on a variety of fronts to disaggregate functions, sometimes retaining asset ownership; sometimes divesting assets and retaining functions in some other way. The adoption of TCP/IP as the “next generation” architecture, for example, necessarily entails separating connectivity functions into layers. 

Fundamentally, that means applications can be created and consumed by customers or users without app owners having formal business relationships with access network providers. 

In other cases, mobile service providers have opted to sell off their cell tower networks, in favor of leasing arrangements. More recently, telcos have shifted their computing networks from internal and owned platforms to use of cloud computing suppliers. 

And while virtualization of network functions, or separated control and data planes, do not intrinsically require ownership disaggregation, it always enables function disaggregation. 

Category	Moves	Examples
Sale of cell towers	Major telecom companies have sold off their cell towers to independent tower companies. This allows the telecom companies to focus on their core network functions, while the tower companies can focus on managing and maintaining the towers.	In 2019, AT&T sold its 8,200 cell towers to Crown Castle for $8.1 billion. In 2020, Verizon sold its 10,000 cell towers to American Tower for $5.1 billion. In 2021, T-Mobile sold its 4,000 cell towers to SBA Communications for $3.4 billion.
Virtualization of network functions	Telecom companies are moving away from traditional, hardware-based network functions and towards virtualized network functions (VNFs). VNFs are software-based network functions that can be run on generic hardware. This allows telecom companies to be more agile and to scale their networks more easily.	In 2017, AT&T announced that it would be moving its network functions to a virtualized architecture. In 2018, Verizon announced that it would be moving its 5G network to a virtualized architecture. In 2019, T-Mobile announced that it would be moving its network functions to a virtualized architecture.
Use of wholesale	Telecom companies are increasingly using wholesale networks to provide services to their customers. Wholesale networks are owned and operated by independent companies, and they sell capacity to telecom companies on a wholesale basis. This allows telecom companies to offer their customers a wider range of services without having to invest in their own network infrastructure.	In 2017, AT&T announced that it would be using the FirstNet wholesale network to provide 5G services to first responders. In 2018, Verizon announced that it would be using the CBRS spectrum to provide wholesale services to its customers. In 2019, T-Mobile announced that it would be using the DISH 5G network to provide wholesale services to its customers.
Adoption of TCP/IP	Telecom companies are increasingly adopting TCP/IP as the underlying protocol for their networks. TCP/IP is a standard protocol that is used for data communication over the internet. This allows telecom companies to interoperate with other networks and to offer their customers a wider range of services.	In 2017, AT&T announced that it would be migrating its network to a TCP/IP-based architecture. In 2018, Verizon announced that it would be migrating its 5G network to a TCP/IP-based architecture. In 2019, T-Mobile announced that it would be migrating its network to a TCP/IP-based architecture.
Architectures using data plane and control plane	Telecom companies are increasingly adopting architectures that use separate data planes and control planes. This allows the data plane to be optimized for performance, while the control plane can be optimized for flexibility.	In 2017, AT&T announced that it would be using a data plane and control plane architecture for its 5G network. In 2018, Verizon announced that it would be using a data plane and control plane architecture for its 5G network. In 2019, T-Mobile announced that it would be using a data plane and control plane architecture for its 5G network.

Likewise, any shift to use of wholesale mechanisms is a form of disaggregation from formerly-vertically-integrated asset ownership. 

To be sure, there are practical reasons for undertaking these moves. At one level, no grand shift of strategy is required, and each single move can be seen as an incremental change to improve operating economics. 

Move	Reason
Sale of cell towers	Allows telecom companies to focus on their core network functions, such as switching and routing, and to outsource the management and maintenance of their cell towers to third-party companies. This can help telecom companies reduce costs and improve their flexibility.
Shifting to wholesale	Allows telecom companies to offer their customers a wider range of services without having to invest in their own network infrastructure. This can help telecom companies reach new customers and compete with larger rivals.
Separating network and business functions	Allows telecom companies to become more agile and to respond more quickly to changes in the market. This can help telecom companies stay ahead of the competition and offer their customers the best possible services.
Taking other steps to separate asset ownership	Allows telecom companies to interoperate with other networks and to offer their customers a wider range of services. This can help telecom companies stay ahead of the competition and meet the growing demand for connectivity.

But all the steps, taken together, in a context where revenue growth remains sluggish, capital investment requirements arguably are rising and competition is growing, might signal continued pressures to disaggregate. 

Indeed, the new involvement of private equity investors in digital infrastructure asset ownership might be part of the shift. To be sure, most such investments involve acquiring both operating and physical assets. To that extent the asset shifts are simply part of the background of asset disposals or acquisitions on an incremental level, and not a “grand strategy.”

But many business plans envision both retail and wholesale operations. And the same is true for the growing number of municipal networks, independent facilities-based internet service provider operations as well. Over time, more fiber-to-home networks are being added that allow wholesale access to other ISPs. 

So far, only cable operator mobile service operations have been based extensively on wholesale mechanisms. And cable operators are moving to shift substantial reliance to their own assets, including both spectrum assets and small cell networks, for example. 

The larger point is that business model drivers might, over time, increase the value or necessity of further disaggregation in the direction of a more-layered organization of the business, especially for non-dominant service providers. 

One clear example, however, is the growing use of joint ventures to build fiber-to-home infrastructure. For some observers, a growing role for app providers such as Google might seem quite noteworthy. 

Year	Location	Partners	Details
2020	United States	Google and Frontier Communications	Google will invest $1 billion in Frontier to help the company build a fiber-optic network to 1 million homes in 25 states.
2020	United Kingdom	Virgin Media O2 and Vodafone	The two companies announced a joint venture to build a fiber-optic network to 1 million homes in the U.K.
2021	United States	Google and AT&T	Google will invest $2 billion in AT&T to help the company build a fiber-optic network to 3 million homes in 10 states.
2022	United Kingdom	CityFibre and Macquarie Infrastructure and Real Assets	The two companies announced a joint venture to build a fiber-optic network to 5 million homes in the U.K.
2023	United Kingdom	Google and Openreach	Google will invest £1 billion in Openreach to help the company build a fiber-optic network to 2 million homes in the U.K.

So far, Google alone has made FTTH investments in firms including Zayo Group, SiFi Networks, Ting, Webpass, Frontier Communications, AT&T and Openreach, in addition to operating its fully-owned Google Fiber business. 

ear	Location	Joint Venture Partners
2020	United States	Zayo Group and Frontier Communications
2021	United Kingdom	CityFibre and Vodafone
2021	Australia	Telstra and TPG Telecom
2021	Canada	Bell Canada and Rogers Communications
2022	Brazil	TIM Brasil and Telefonica
2022	India	Reliance Jio and Google
2022	Japan	NTT Docomo and KDDI
2022	South Korea	SK Telecom and KT
2022	Taiwan	Chunghwa Telecom and Far EasTone

Perhaps one would speculate that facilities disaggregation for FTTH networks will often manifest itself in the form of joint ventures. That is an incremental step that only “shares” asset ownership rather than dispensing with it. 

-----------------