Of All Mobile Operator Assets, Perhaps Spectrum is the Enduring Driver of Value

"Assets" and "value" are not the same thing. The former can be shown on financial statements; the latter not so much. And an argument might be made that mobile operator asset value is becoming more driven by spectrum licenses and troves than other network-related physical assets.

By some measures, the capital investment in spectrum of the largest three U.S mobile operators ranges from 30 percent to nearly half of the committed capital investment of those firms. Also, as the growing involvement of infrastructure investment in telco access networks continually grows, we must ask new questions about the value of the actual access networks and assets.

Infrastructure investors are playing a bigger role in access network capital investment, often for the same reason they invest in airports, seaports and other forms of long-lived infrastructure. They see “moats” and stable, long-term demand with predictable cash flows. 

As did other investors of 30 to 40 years ago, connectivity infrastructure and the cash flows built on it are seen as relatively stable sources of free cash flow bolstered by their relative scarcity. It started with cell towers but increasingly is moving towards optical fiber access networks, small cell network providers and data center infrastructure.  

All of that raises new questions about where value lies in the connectivity business. To use the obvious analogy, money can be made operating a seaport or airport as money is to be made moving goods from manufacturers to end users and retail buyers. 

The best connectivity industry examples are wholesale access platforms, where one entity owns the infrastructure and retail service providers all use the one infrastructure. 

The business choice between facilities-based versus leased network access is an issue in other contexts as well.  At scale, the former tends to offer better economics than the latter. That is why the leading mobile service providers in almost-all markets own their own networks. 

At lower volumes, and especially “outside” the core geography, leased access  typically offers better economics. 

Hybrid models seem to be developing, though, where the access infrastructure is partly owned by a service provider and an infrastructure investor. The advantage is lower capital investment by the service provider, at the cost of shared revenues and profit. 

Monetization opportunities will often depend on the ability to sell infrastructure access to multiple buyers, where traditionally a network has been virtually exclusively for the use of the network owner or its wholesale customers. As data centers arguably do best when they are “carrier neutral,” access assets might in some cases also benefit from multi-customer business models. 

Indoor coverage by small cell networks provides an example. It will make sense for all mobile operators to take advantage of an indoor coverage network, for example. 

It is not yet clear just how far changes in the physical platform could evolve. It might be fair to say that if value tends to migrate elsewhere in the ecosystem or value stack, away from the “access” function, such moves create new possibilities. 

As networks increasingly are virtualized, it is conceivable that a new division of labor could develop in some markets, with asset owners providing physical facilities and other participants providing the service enablement or services. 

As we presently see with wholesale-only infrastructure, one entity might provide the access functions, while all retail service or app providers pay to use that infrastructure. Though this might always make more sense for fixed networks, even mobile networks might eventually consider such a pattern, if capital intensity were to increase for future very-dense networks. 

As we have already seen, perhaps a consortia of mobile operators might join together to create and own the physical plant. The odds of such developments increase with potential infrastructure intensity and cost.

As the U.S. spectrum auction of 3.4-GHz spectrum nears an end, $22 billion already has been bid on those assets, with an average price per MegaHertz per person of about 65 cents. Some will argue that directly increases capital investment by the winning entities. Others will agree, but enumerate those investments separately from other mobile operator capex. 

How much capital investment mobile operators invest each year depends on how one counts such investments. Often, physical plant and spectrum investments are tallied separately. 

Consider a CTIA report on U.S. mobile industry capital investment. Between 2016 and 2020, for example, “capex” is said to range from $25 billion to $30 billion per year. 

source: CTIA

Spending for spectrum licenses ranges is tracked separately, it appears, as it appears spectrum license spending has accelerated since 1994. 

Prior to 1994, spectrum license spending might have been tracked by decade. Between 1994 and 2001 CTIA used an eight-year interval. Since 2002 CTIA uses a five-year interval. 

That indicates a quickening of the tempo of license acquisitions; more competition for licenses as well as bigger spectrum allotments (more frequent auctions; more licensees; more capacity per auction). 

source: CTIA 

It appears CTIA, for purposes of tracking capital investment, does not include spectrum purchases in its chart on U.S. mobile industry “capital investment.” CTIA shows cumulative “capex” between 2016 and 2020 as a cumulative $138.5 billion total. 

CTIA also shows spectrum licenses paid for in the 2017 to 2021 period as $116 billion. So obviously, spectrum license spending, though “capex” for accounting purposes, is separated from spending on spectrum licenses, which also is categorized as “capex” by accountants. 

Standard and Poors includes spectrum licenses in the “capex” category, for example.  

So actual mobile capex often is portrayed as higher or lower, depending on the assumptions. “Network capex” often is separately portrayed from “spectrum capex.” Both have grown. 

CTIA also says cumulative capex including 2001 to 2020 is $601 billion, an average of about $30 billion per year. That tends to correlate to separate tracking of network capex and spectrum license capex. At the same time, cumulative spectrum license fees are said to represent a cumulative $200 billion in spending.   

Spectrum licenses represent perhaps 25 to 35 percent of mobile operator assets in the United States and perhaps 10 percent to 20 percent of assets for other mobile operators globally.  

Overall, some might argue that total capex is rising. This chart from Standard and Poors Market Intelligence shows U.S. mobile operator capex (Just the three largest facilities-based providers) rising over time. As S&P includes spectrum spending in capex totals, that suggests spectrum spending is rising. 

source: Standard and Poors 

But secondary transactions also happen. Including those investments, U.S. mobile operators have cumulatively spent about $601 billion in capital investment since 1994, according to the CTIA.

Spectrum license purchases since 1994 have amounted to at least $211.5 billion, not including the cost of spectrum licenses that have traded hands in the form of direct purchases or company acquisitions and mergers. 

Assume that about half of all spectrum originally acquired at auction then is resold on secondary markets, primarily as firms with those licenses are acquired by other firms. That would imply an additional $106 billion that mobile operators have spent on spectrum assets. 

source: CTIA 

If so, then spectrum represents about $317.5 billion in spectrum capex, or roughly 53 percent of total capex.

The larger point is that the value of access network assets might be changing. Such assets are viewed as desirable investments by private equity and institutional investors. At the same time, a greater share of access network plant might be viewed by telcos as not providing sustainable business advantage, or at least not enough value to consider 100-percent ownership of such assets.

At the same time, if capital investment intensity continues to rise, there will be greater business logic to considering partial divestment of access network plant.

That might well raise new questions about the value of structural separation--not as imposed by regulators--but as business choises made by the telcos themselves. It is conceivable that market participans themselves might revalue access infrastructure in ways that lead to more structural separation, as a way of managing capex while maintaining value.

Are "Fairness Doctrines" Helpful or Possible?

“Deplatforming” is among the latest problems to arise in the content business. But some would say there are troubling issues around bullying or misinformation. All generally pose issues about First Amendment freedom of speech, with the caveat that the First Amendment has been held to apply only to the federal government. 

There is, so far, no generally recognized individual “right to free speech” on any social media platform, for example. Nor can private firms be prevented from espousing their own views, with no regard for balance or “ fairness.”

In 1949 the Federal Communications Commission promulgated the fairness doctrine, which required the holders of broadcast licenses to present controversial issues of public importance and to do so in a manner that was honest, equitable and balanced. It was abolished in 1987. 

The objective was to encourage a “diversity of viewpoints” in broadcast media. Some seem to believe a new fairness doctrine would be a good thing. Maybe not. 

In the internet era, it is hard to make the argument that the ability to express opinions requires any help. There is no scarcity of voices or platforms. 

“But we already see some of the problems with maintaining “honesty, equal and balanced” discussion. Who decides what is honest? How is balance maintained. Nor is it simple to operationalize “balance.”

The old adage about “two sides to every argument is flawed. There always are multiple “sides.” 

As a practical matter, how would the licensing body determine what “fairness and balance” is, in practice? Who can define “public interest?” 

And, as we saw when the fairness doctrine was enforced, how will speakers change their behavior to minimize the impact? One way broadcasters limited their obligations was to reduce, not increase, the amount of political content they carry. 

So rather than more political speech “in the public interest,” there was less coverage of political issues. Also, as television has matured, it has become a medium organized around entertainment, not “public affairs.” More regulation would likely result in less coverage, not more, as advocates of a new fairness doctrine might argue is necessary. 

There is a deeper issue: whose rights are we talking about? Until the era of broadcast media, it has been “speakers” who had First Amendment rights of free speech. The fairness doctrine flipped that on its head, being promulgated because it was viewers and listeners who had the rights. 

It is hard to manage a protection or promotion process for audiences, rather than speakers. Nobody knows what the audience “needs or prefers” in a broad sense. So some referee--some government entity--gains control over political speech, which is precisely what the First Amendment is designed to prevent. 

Promoting or protecting freedom of speech is a tough issue, but fairness doctrines are unlikely to help. Many problems are caused by private firms or individuals and groups seeking to silence some views.

However well-intentioned, those efforts often are arguably misguided. “Cancelling” speech does not promote freedom or a fuller airing of views. “Fairness” doctrines might sound good, but have not worked and arguably are not necessary at a time when so many ways exist for people to “speak.” 

There are, one might argue, many problems with political culture. Fairness doctrines will not fix those problems.

If Capex Intensity Increases, How does Thinking About Network Ownership Change?

Infrastructure investors are playing a bigger role in access network capital investment, often for the same reason they invest in airports, seaports and other forms of long-lived infrastructure. They see “moats” and stable, long-term demand with predictable cash flows. 

As did other investors of 30 to 40 years ago, connectivity infrastructure and the cash flows built on it are seen as relatively stable sources of free cash flow bolstered by their relative scarcity. It started with cell towers but increasingly is moving towards optical fiber access networks, small cell network providers and data center infrastructure.  

All of that raises new questions about where value lies in the connectivity business. To use the obvious analogy, money can be made operating a seaport or airport as money is to be made moving goods from manufacturers to end users and retail buyers. 

The best connectivity industry examples are wholesale access platforms, where one entity owns the infrastructure and retail service providers all use the one infrastructure. 

The business choice between facilities-based versus leased network access is an issue in other contexts as well.  At scale, the former tends to offer better economics than the latter. That is why the leading mobile service providers in almost-all markets own their own networks. 

At lower volumes, and especially “outside” the core geography, leased access  typically offers better economics. 

Hybrid models seem to be developing, though, where the access infrastructure is partly owned by a service provider and an infrastructure investor. The advantage is lower capital investment by the service provider, at the cost of shared revenues and profit. 

Monetization opportunities will often depend on the ability to sell infrastructure access to multiple buyers, where traditionally a network has been virtually exclusively for the use of the network owner or its wholesale customers. As data centers arguably do best when they are “carrier neutral,” access assets might in some cases also benefit from multi-customer business models. 

Indoor coverage by small cell networks provides an example. It will make sense for all mobile operators to take advantage of an indoor coverage network, for example. 

It is not yet clear just how far changes in the physical platform could evolve. It might be fair to say that if value tends to migrate elsewhere in the ecosystem or value stack, away from the “access” function, such moves create new possibilities. 

source: Kearney

As networks increasingly are virtualized, it is conceivable that a new division of labor could develop in some markets, with asset owners providing physical facilities and other participants providing the service enablement or services. 

As we presently see with wholesale-only infrastructure, one entity might provide the access functions, while all retail service or app providers pay to use that infrastructure. Though this might always make more sense for fixed networks, even mobile networks might eventually consider such a pattern, if capital intensity were to increase for future very-dense networks. 

As we have already seen, perhaps a consortia of mobile operators might join together to create and own the physical plant. The odds of such developments increase with potential infrastructure intensity and cost.

SMB IT Spending Rebounding to Normal Levels

Small and medium business information technology spending is expected to rebound to pre-Covid levels, says Analysys Mason.

IoT Already a Material Revenue Driver for Some Service Providers

Internet of things already is a potentially-significant revenue driver for connectivity providers, as the number of machine-to-machine devices continues to grow. Not all IoT device connections will require direct mobile or fixed network connectivity. Many devices will use some unlicensed local access technology, with backhaul to the internet provided by a mobile or fixed connection. 

source: OECD 

Still, mobile IoT connections are a material driver of subscriptions in many markets. “Overall, M2M/embedded mobile cellular subscriptions grew by over 30 percent in one year (the Q2 2018-Q2 2019 period) in countries where data were available,” says OECD researchers.

IoT connectivity revenues already are a material  revenue source for many service providers, though estimates might vary. 

Units Sold and ARPU Remain the Mainstay, Even for IoT, Edge Computing Opportunities

There are some unstated maxims for connectivity businesses where it comes to revenue growth. As for many other consumer products, the core revenue model is built on units sold and price per unit. 

When, in a single market, few have subscriptions, the growth strategy is “get more subscribers.”

When subscriber adoption is high in any single market, the growth strategy often is “get new subscribers in other countries.” 

When that is deemed infeasible for any reason, the logical revenue growth strategy is either “take market share from other suppliers” or “change the value proposition so we can charge higher prices.”

Only when none of those strategies seems to work will connectivity businesses ever seriously consider shifting the business model and moving into adjacencies and taking on additional roles in the ecosystem. 

source: OECD 

At some point, as markets saturate,  “get more subscribers” will become a challenging strategy, with more effort shifting to “get higher prices.” Only at some future point, when neither of those strategies is working very well, will connectivity providers willingly and seriously look at ways to leverage additional roles in the ecosystem in meaningful ways. 

In the meantime, there is some opportunity to do so in the internet of things and edge computing markets, for example, even if the primary revenue upside is likely to be found in the traditional “add more subscribers” area. Higher average revenue per unit sold is unlikely to be higher than mass market ARPU for either IoT or edge opportunities, though. 

Compare What Matters When Assessing Broadband or Mobility Behavior

If one spends a lot of time looking at statistics on global broadband and mobile trends, one sees lots of comparative data about take rates, subscriptions, prices and speeds. One also sees a lot of data comparing various indicators across countries. 

There always are methodological issues when trying to compare broadband or mobile prices across countries, however. Living costs and income vary, so prices vary in ways that reflect those differences. Currency differences also operate. 

There are big distortions created when researchers select particular service plans to compare. Results are different when choosing the lowest-priced plan; the lowest-usage plan; the highest-priced plans; highest-speed or highest-usage plans or some “median” plans (retail price, usage allowance or speed). 

So distortions are inevitable across countries when rate of adoption, typical speeds, typical prices and usage allowances vary widely. Consumer buying patterns also matter.

source: OECD 

Researchers must decide, before comparison, whether the service plans consumers actually purchase are a relevant sorting criteria. In other words, some argue it makes most sense to look at the plans consumers actually purchase, not the plans they could buy. 

That is abundantly clear when looking at buyer behavior in Organization for Economic Cooperation and Development countries, where a high of 100 percent and a low of about 35 percent of customers buy “service bundles” rather than any “single-product” plan. 

That of course distorts results when researchers compare only single-product plans. In other words, where many to most consumers buy bundles, the “single-product” plans tell us close to nothing about the actual prices consumers pay. 

Those plans exist, but do not represent actual buying behavior. 

If researchers insist on using those plans for comparison, they can compensate by using methods such as purchasing power parity to normalize real prices. Done that way, actual prices can vary substantially from what the data appears to show. And that has been true most of the last decade. 

In fact, when adjusting for purchasing power, broadband prices globally are quite uniform, different studies show. 

Comparing single-product plans might be useful when most consumers buy them. The exercise does not tell us very much when relatively-few consumers buy them.