Monday, April 30, 2018

Sprint, T-Mobile US HHI Numbers Probably Will Not Pass Antitrust Muster

The proposed merger between Sprint and T-Mobile US is going to have a tough time getting around its score on the Herfindahl-Hirschman Index (HHI), a basic tool used by the U.S. Department of Justice and most other antitrust authorities globally.

The Justice Department will generally investigate any merger of firms in a market where the Herfindahl-Hirschman Index (HHI), a test of market concentration, exceeds 1000 and will very likely challenge any merger if the HHI is greater than 1800.

The U.S. market has an HHI of about 2500.

Three years ago, the very same proposed transaction would have occurred where the U.S. market had an HHI score of about  2,766. But following a merger of Sprint and T-Mobile, the score would be 3,252. Is the market less concentrated now?

The last time Sprint and T-Mobile US tried to merge, three years ago, Craig Moffett of MoffettNathanson calculated that the wireless industry currently had an HHI score of 2,766.

But following a merger of Sprint and T-Mobile, the score would be 3,252. That suggests an increase in concentration of about 486 points. So it did come as a surprise that regulators signaled opposition to the merger.

Do you think the market is significantly less concentrated than three years ago? AT&T and Verizon are probably slightly down, share wise, from three year ago levels. T-Mobile clearly is up, but Sprint is down, MVNOs are flat and others have lost share.


To be sure, the U.S. market is not as concentrated as many other markets. Looking at the biggest 36 mobile markets globally, analyst Chetan Sharma found that the average HHI score of a typical market ranks 3440 on the scale.

Developed markets have an HHI of 3270. So the U.S. market, with an HHI of 2500, lies between “heavily concentrated” and “moderately concentrated.”

The point is that DoJ is looking at a major market concentration move in a market that already is, by its own tests, moderately to heavily concentrated.

While it always is conceivable that U.S. Department of Justice attorneys will ignore their traditional horizontal merger guidelines, it seems unlikely that any proposed horizontal merger increasing the Herfindahl–Hirschman Index by more than 200 points will have an easy time being approved.

The agency generally considers markets in which the HHI is between 1,500 and 2,500 points to be moderately concentrated, and consider markets in which the HHI is in excess of 2,500 points to be highly concentrated.  

Transactions that increase the HHI by more than 200 points in highly concentrated markets are presumed likely to enhance market power under the Horizontal Merger Guidelines issued by the Department of Justice and the Federal Trade Commission.  

The proposed AT&T acquisition of T-Mobile would have resulted in AT&T having 43.7 market share, resulting in an HHI of 3,335, an increase of 951 points. It is no surprise that the merger effort failed.

The AT&T acquisition of T-Mobile also would have raised HHI scores by more than 200 points in 94 percent of markets. By some estimates, HHI would have increased by more than 416 points.

The latest proposed merger should probably find an HHI change of close to that 416 points, not much less than the 486 points that lead the DoJ to oppose the AT&T acquisition of T-Mobile US.

Will Federated Networks Ever Have "Owners Economics?"

Have you noticed how there is a need for, and move towards “federation” in most parts of the computing and telecom ecosystem? In other words, interoperability between networks and systems is federation.

Federation can be defined as “a group of computing or network providers agreeing upon standards of operation in a collective fashion.”

The typical practical arrangement is when two distinct, formally disconnected, telecommunications networks, that may have different internal structures, interconnect. When different messaging platforms agree to interconnect, that is another example of federation.

In computer systems, to be federated means users are able to send messages from one network to the other. In other words, federation means interoperability.
You can see this federation trend in the move to multicloud computing, when resources from multiple cloud providers can be amalgamated to work as one functional system.

Google Project Fi federates mobile device access from Wi-Fi, Sprint or T-Mobile US networks.

That last example suggests possibilities, as telecom networks increasingly become virtualized. Federation is about the creation of virtual networks, incorporating multiple physical network assets.

So here’s a big question. Will “owner’s economics” change in a world where federation is widely possible? Up to this point, it has been difficult for mobile virtual network operators to compete, sustainably, with competitors who own their own infrastructure.

In a virtualized environment, will that remain true? As we have seen with cloud computing, the economics of “owned capabilities” compared to “rented capabilities” already has changed.

Will something like that happen with telecommunications capabilities? In other words, might it be possible, in the future, to create virtual (federated) communications networks that rival physical networks? And could the economics change, to where a virtualized network has economics at least equal to those of a physical network?

It is an empirical question that will be answered in the concrete, at some point. In other words, might an app, platform or service provider be able to create virtual networks that cost less than using owned physical facilities?

Probably more to the point, can new capabilities be created out of federated networks that are driven by value, not cost? Can federated networks do things that single networks cannot? And does that increase in value offset the costs of using a virtualized approach, even if more costly than a physical network approach?

We must wait to see how the economics play out.


Source: Nokia Bell Labs

Sprint T-Mobile Merger Approval Still Seems Questionable

The odds that the new proposal to merge Sprint and T-Mobile US could clear antitrust review do not seem to be different from three years ago, when the same deal was proposed, and was opposed by the U.S. Department of Justice.

In fact, DoJ opposition to the vertical merger of AT&T and Time Warner assets, would suggest the conditions are worse. Win or lose their antitrust lawsuit, has DoJ signaled it is more open to big horizontal mergers of the specific type it opposed just three years ago, and when it has opposed a vertical merger many believe should not have caused concern?

Some argue this is not an instance of “consolidation.” That is simply untrue. We would move from four facilities-based tier-one suppliers to just three, at least until some time in the future, when one of the three experiences a major business reversal and some tier-one supplier in the internet ecosystem decides it needs such assets.

“This isn’t a case of going from four to three wireless companies,” says T-Mobile US CEO John Legere. He notes the existence of other mobile virtual network operators and a few regional providers. But the merger most certainly reduces the number of facilities-based tier one providers from four to three.

Some might say the proposed deal occurs “under different conditions” than three years ago, when the U.S. Department of Justice signaled its opposition. True, Comcast is an operating MVNO now. Charter is set to enter the market as well, as an MVNO. But market share has not budged much.

Also open to question are the consumer benefits. Virtually all equity analysts have been in favor of consolidation from four to three mobile companies because that would reduce competition and allow all the firms to raise prices.
T-Mobile and Sprint also argue the deal will not lead to layoffs. What merger of this size has not been accompanied by “merger synergies” that reduce headcount?

“This combination will also force AT&T, Charter, Comcast, Verizon, and others to make investments of their own to compete,” T-Mobile US and Sprint say. Those firms already are investing as fast as their capital budgets will allow.

It is not surprising that Sprint and T-Mobile US argue their proposed merger will lead to faster 5G investment, U.S. 5G leadership, lower prices, create jobs, higher economic growth and more competition in the mobile business. What, after all, would they say?

“This combination will create a fierce competitor with the network scale to deliver more for consumers and businesses in the form of lower prices, more innovation, and a second-to-none network experience – and do it all so much faster than either company could on its own,” said John Legere.

No offense; Legere is a brilliant competitor. But many of the claims are questionable, some would argue. “Faster investment?” It is hard to argue that the four facilities-based mobile service providers are not investing in 5G platforms as fast as their capital budgets will allow.

To be sure, a merged Sprint and T-Mobile US could invest more rationally, but likely not faster: they are running as fast as they can, right now. And by “rational,” less. That is why the argument in favor of the merger includes the claim that “neither company standing alone can create a nationwide 5G network.”

For that claim to make sense, one would have to believe T-Mobile could not have done so on its own. I doubt anybody believes that, given T-Mobile’s own recent claims about how fast it is moving, and how much spectrum it has to do so.

Sprint is challenged financially, to be certain. But it never has been clear Sprint would survive as a stand-alone company in any case. Most observers would likely agree “somebody” eventually will buy Sprint.

But is T-Mobile the “best” buyer, if industry lines are blurring, as Legere argues? Would not a tier-one app provider, a major device supplier, a big cable TV company better match the profile of an industry that now fuses content, apps and access? In that sense, the proposed Sprint merger with T-Mobile creates a bigger horizontal mobile company, to be sure.

But the strategic need still remains: ownership of apps, content and access assets are the future. This merger creates scale in access, which is helpful, to be sure. But the move does not address the future moves that then would become necessary: merger again with a big firm in the apps, platform or device area.

In fact, the scale of the new company likely reduces the odds many firms could envision such an acquisition. And there is a good argument to be made that if the industry is changing (access, platform, apps and content all fusing), as the two firms claim, this move does not help them do so. It simply creates more scale. That is helpful, but leaves unanswered the question of how to strategically reposition in a market that devalues "access only" or "content only" providers.

“There is no way we were going to build four national 5G networks,” some might argue. Many would argue that is not true. One way or the other, all four facilities-based providers are moving to 5G, as fast as they can.

And there are other ways for either T-Mobile or Sprint to get some assets (backhaul for small cells) they might need, if in fact each requires as much small cell support as do AT&T and Verizon. Those are open questions.

The proposed new company would not--despite the claims--get the country  to 5G any faster. AT&T and Verizon are moving as fast as they can, no matter what.

As for the claim that such a merger improves chances of “competing with China in the global 5G race,” other questions have to be raised. Where is the competition? Access? Or is the competition fundamentally in the platform, apps and infrastructure supply areas? If so, having another big supplier of “access” does not help much.

In fact, a combined new firm will obviously spend less on infrastructure than the stand-alone companies would have. That is why they want to merge, in part.

Also, the new firm does not create new platform, app or content capabilities. It is, by definition, a horizontal merger of like access assets.

The two firms argue their merger will lead to more competition in rural markets, beyond the fixed network providers. That could happen, at the margin. But more competition in rural markets, by wireless substitution, is coming, anyway, and not just from mobile suppliers.

The merger might, as Sprint and T-Mobile US argue, lead to creation of a better competitor in the enterprise mobility area. That also is possible, to an extent.

The main points, though, are that the stated rationales might be quite the opposite of the probable impact in some cases, are neutral in many other instances and helpful at the margin, as the companies claim.

But it seems unlikely that much as changed in terms of likely antitrust opposition. Much, observers say, will hinge on AT&T getting antitrust approval to acquire Time Warner. But that is a vertical merger that absolutely does not reduce competitors in the market, quite unlike a horizontal merger that will reduce competition.

And DoJ resistance even to the vertical merger, after declaring opposition, three years ago, to the same proposed Sprint merger with T-Mobile, does not bode well for the same transaction, three years later.

Sunday, April 29, 2018

Internet.org Sponsored Access Helped 100 Million People

Facebook’s Internet.org efforts “have now helped almost 100 million people get access to the internet who may not have had it otherwise,” according to Mark Zuckerberg, Facebook CEO.

Ironically, this form of sponsored access is among the practices many “network neutrality” supporters want to outlaw. After all, giving users “no charge” access to a suite of popular apps--but not the “full internet”--does not “treat all bits and apps the same.”

That is among the objections some of us have to strong forms of network neutrality. Consumers should have full access to all lawful apps. That is among the original ways that neutrality originally was conceived, and still makes sense today.

But it has never made sense to some that, in addition to that openness, access providers should be legally barred from offering sponsored access, in the same way that internet.org works, or the way “toll-free calling” has been lawful in the voice business.

App providers routinely use content delivery networks--paid-for services not available to every app provider--to provide better quality of service. In principle, in at least some cases, it makes sense to allow use of content delivery mechanisms to provider better quality of service over access connections, as well as over backbone networks.

There are other advantages for app providers, as well as end users. Caching content at the edge of the network not only provides better user experience, but cuts the cost of internet transit services app providers must buy to support app access.

When content is locally based, app providers do not have to buy as much capacity across the internet backbone. That might be one reason internet transit revenue has dwindled in recent years, even as consumer bandwidth demand skyrockets.

source: TeleGeography

Friday, April 27, 2018

Sometimes Unhappy Customers Continue to Buiy

Two and three decades ago, it was somewhat hard to explain why buy rates for linear video were so high (up to 95 percent of U.S. households), but satisfaction routinely was so low. That sort of flies in the face of our general belief that happy or satisfied customers are good predictors of buying behavior.

In fact, even unhappy customers will continue to buy products for which there are no effective substitutes.

The answer is that, despite their mixed feelings, consumers did not actually have choices. If you wanted access to content diversity, the only choice was subscription TV.

That is no longer the case, so we should not be at all surprised that consumers are abandoning linear services for streaming alternatives that replicate much of the value, at far lower prices.

Linear video entertainment subscriptions have gotten more expensive for all U.S. consumers, but arguably most so for households with smaller incomes.


Though prices have virtually always risen annually, the issue is whether the rate of increase is in line with the growth of prices generally.


Nobody disputes the evidence that linear video entertainment subscription prices have risen faster than underlying inflation. According to SNL Kagan, the price increases have outstripped median household income growth since 2007.




All products have life cycles. Three decades ago, consumers might not have been completely happy, but there were no real alternatives. That is one explanation for persistently low satisfaction scores, but high buying rates.


That is not the case any more. Consumers do have choices, and the range of choices is growing.  So consumers are voting with their wallets to use streaming services that cost less. No big surprise, there.

U.S. Mobile and Internet Access Prices Have Fallen for 2 Decades

Between 1997 and 2017, the cost of mobile service in the U.S. market dropped 50 percent, according to the U.S. Bureau of Labor Statistics.

According to the U.S. Bureau of Labor Statistics, prices for internet services and electronic information providers were 22 percent lower in 2018 versus 2000.

Between 2000 and 2018, internet access services experienced an average inflation rate of -1.36 percent  per year. “In other words, internet services costing $50 in the year 2000 would cost $39.10 in 2018 for an equivalent purchase,” says the BLS.

The overall inflation rate was 2.07 percent during this same period, BLS says. Likewise, prices for mobile services also have fallen.

There often is a big difference between posted retail prices for internet access and the packages consumers actually buy. Available plans can be expensive or cheap, but what matters are the buying habits of actual customers, who may or may not pay those posted prices.  

These days, as so many U.S. consumers are on some sort of plan featuring a product bundle, there might not be very many customers actually paying the “stand-alone price” for internet access.

In fact, internet access prices actually have fallen over the last two decades, not increased.  

Tuesday, April 24, 2018

Taiwan Regulator Warns of Dangers of Excessive Competition in Mobile Market

Up to a point, competition is good for consumers, leading to better prices and greater value. But there is such a thing as “ruinous competition,” which drives prices below the point at which suppliers can sustain themselves or invest in new products and features that provide more value.

That is a problem in the Taiwan mobile market, says  National Communications Commission (NCC) spokesperson Wong Po-tsungj. The problem is price wars.

Chunghwa Telecom, for example, recently introduced a 4G service plan costing NT$499 monthly (about US$17 a month) for unlimited access to mobile Internet and unlimited phone calls between Chunghwa Telecom subscribers. That basic plan was matched by competitors.

“If telecoms simply want to boost their market shares and revenue by luring subscribers from competitors, rather than with innovative business models, it would not be positive for the development of 5G in the nation,” Wong said.

“What they are doing does not help to make the pie bigger,” said Wong. And though it is easy to criticize firms for making profits, those profits are what allows firms to invest and innovate, as well as stay in business.

That tension between actions and policy that support sustainable competition, or ruinous competition; or competition versus investment, is at the heart of all thinking about ideal mobile market structure.

Few believe any more that telecom is a natural monopoly. But few would deny that sustainable retail markets are likely to be oligopolies. The real questions tend to be over the shape of such oligopolies. How much sustainable competition--especially facilities-based competition--is possible on a sustainable basis?

That is the problem NCC sees with the current price wars.

Fixed Wireless Now Generates as Much Global Revenue as Business Voice, Unified Communications

As vital as business voice and unified communications capabilities might be, they do not drive huge service provider revenues, globally.

By some measures, the global market for all business voice and unified communications, including the value of business access to support voice and UC, is perhaps US $28 billion a year.

Of that $28 billion or so of revenue, perhaps $10 billion consists of internet and data access circuits, implying that the value of hosted voice, unified communications apps and phone systems is about $18 billion annually.

So compare that to just one other product, namely fixed wireless internet access, admittedly a niche.

The global fixed wireless access market will grow 30 percent in 2018 and will generate $18 billion in service revenue, according to ABI Research, boosted by new use of 5G platforms to supply fixed wireless.

The point is that the fixed wireless revenue segment already is about the same size revenue contributor as business voice and unified communications.

It might not be easy to compare the relative value of those revenue streams. Businesses require voice capabilities and unified communications, while consumers also "need" internet access. The point, though, is that the fixed wireless internet access business (mostly consumer) is about the same size revenue contributor as all business voice and UC (phone systems, hosted voice, unified communications solutions).

You might therefore characterize the fixed wireless segment--though still a niche within the broader service provider business--as roughly as big a contributor as business voice and UC.

ABI Research forecasts worldwide fixed wireless broadband market to grow at a compound annual rate of 26 percent to generate $45.2 billion worth of revenue globally  in 2022.

Tier-one service providers launching 5G fixed wireless include Verizon plans an initial 5G fixed wireless network covering around 30 million U.S. households.

AT&T and Charter also are carrying out 5G fixed wireless broadband tests in select markets in the United States.

In Europe, Orange, Elisa, and telecom infrastructure company Arqiva are performing 5G fixed wireless trials as well. Australia’s Optus is planning for a 5G fixed wireless service launch in 2019.

Verizon is Optimistic About 5G; Others are Skeptical. Both are Right

As always, public company perceptions of upside from new markets is conditioned by their own view of ability to capitalize on those new markets. In many regions and markets, heavy recent investments in 4G mean service providers have less appetite for another round of investment to support 5G, as you would expect.

In other regions, where profits in the 3G and 4G era have been tough, there is some caution about revenue and profit upside from 5G. In addition, specific contenders have their own business reasons for seeing potential or danger. In other words, skepticism about the size of the 5G market opportunity, as well as optimism, are specific to particular firms, in particular geographies, for reasons having to do as much with current business environment as anything else.

For U.S. service provider Verizon, a relatively-small fixed line footprint, and the ability to compete in new markets with mobile-based alternatives to fixed network service, are a clear upside.

“We continue to be very excited about the opportunities that 5G has,” said Matthew Ellis, Verizon CFO. “I don't think in the U.S. we've seen people pulling back from 5G at all.”

Simply, Verizon sees many ways to grow its business with 5G. Residential broadband, fixed and mobile, is part of the expectation. But the biggest upside is expected from new internet of things and business-to-business apps, many of which can only be supported by ultra-low-latency networks.

In many quarters there is concern or belief that 5G will require huge capital investment boosts, beyond what was required for 4G. That is debatable. Verizon, for example, does not see capex exceeding the typical and expected annual range.

Though some seem convinced 5G capex will be double or triple what 4G cost, others believe 5G capex could actually be lower than required to create 4G. Some believe 5G could be built with only modest single-digit increases in overall capex. And yet others believe 5G will even cost less than 4G cost.

A variety of new approaches, including open source, virtualization, huge increases in use of unlicensed spectrum, shared spectrum and aggregated spectrum, new radio technologies and ability to leverage existing fixed network investments all will contribute to cost profiles for 5G using lots of small cells and new amounts of backhaul.

On the revenue side, Verizon has ability to compete for new revenues outside its current fixed network footprint. Also, the U.S. market is among those expected to lead in early internet of things deployments.

Also, unlike some other markets, the U.S. has a relatively robust supply of fiber and capacity “deep into the access network,” provided by two or more access providers in nearly every significant market (cable, telco, independent ISP, mobile infrastructure companies,  independent business bandwidth suppliers).

That means small cell backhaul facilities are arguably better positioned, in terms of facilities in place and the number of partners to supply such backhaul.

The point is that a range of opinions exist--for good reasons--about upside for 5G. There is no single pattern that fits every market.

Monday, April 23, 2018

Connectivity Not Necessarily a "Commodity?"

Some statements properly need to be qualified and understood in context, even if our typical phrasing lacks that nuance. We might agree there are a few instances where "connectivity is not a commodity." But that characterization hinges on our understanding of the term, in context, in all its range of connotations.

When observers say that connections to the internet (business or consumer) are a commodity, that typically is phrased as access or connectivity being a “dumb pipe”

The other implied terms span a range: low value, low price, slow-growing, low profit margin, declining growth, saturated market or any other modifier that suggests the connectivity is old, stagnant, saturated, hard to differentiate, lacking growth opportunity, is not innovative,  or some similar notion.

So when a telecom executive implies that “connectivity is not a commodity,” that claim also has to be parsed. Stefano Gastaut, CEO of Vodafone IoT, does not believe that connectivity will be a commodity for a long time.

There are many nuances here. The claim is not unqualified. Gastaut does not claim connectivity services will never be a commodity, only that it will take some time for that to happen.

Allso, note his own context. He is the head of a business unit that sells internet of things services that arguably “are not yet common or necessarily a commodity,” using the sense of “a market that is young and fast-growing.”

Gastaut noted that Vodafone now has 65 million IoT connections, including 14.4 million connected vehicles and five million medical devices. He said IoT is a “sizeable business growing very nicely.”

So Gastaut seems to imply an understanding of the term “dumb pipe” that includes the size of the IoT connectivity business opportunity and the growth rate. But it also is fair to note that IoT sensor connections often generate revenue of between 15 cents per month to $1.50 a month.

Irrespective of profit margin, that is not a lot of incremental revenue, even in high volume.   
He also suggests that connectivity services are not like other utility services (water, electricity) that cannot innovate.

Observers might agree with some of the possible characterizations, using the broadest possible range of connotations for what “dumb pipe” means, in a business sense.

Aside from the clear “technology” function of internet access (it is, in fact a dumb pipe allowing access to any lawful third party app using the public internet, or any lawful private app for which a specific user has access right), one might use an expansive understanding of the range of meanings around “dumb pipe.”

Some of us would use a narrower understanding. IoT connectivity might or might not yet be a commodity. It might not yet have slim profit margins, (though that is most likely). The number of connections clearly is expected to grow at high rates.

Moreover, the range of ecosystem roles IoT connectivity providers assume--in addition to connectivity--is not yet clear. So such providers might well operate apps, platforms and devices in addition to connectivity, and those roles will have varying revenue, growth and profit profiles. Not all will appear to be “commodity” driven in terms of business dynamics.

The point: perhaps some forms of connectivity (market segments, specific apps and geographies or routes) are not presently “commodities,” using a broad sense of what the term “dumb pipe” means.

Those specific instances are few. The word “commodity” can be understood as “a useful thing” or a product that can be bought or sold, or exchanged for other products, or, as its intended usage in telecommunications implies, commodity is a product with no “unique selling point.”

In other words, a ton of raw sugar can be priced and sourced from any number of suppliers, geographies and plant sources, at identical or similar prices. The sugar is chemically identical, irrespective of the source.

So, in context, connectivity services are not “completely” commodities. Network access in France is not a full commodity in the sense that access providers in North America can supply the demand. Capacity across the Atlantic is not fully interchangeable with capacity across the Pacific, or between North and South America.

There are geographic barriers, business model barriers, regulatory obstacles and also some differences in product packaging and tariffs, in any country with competition.

All that noted, in the internet era, the value of access is sharply affected by the shift of value to other parts of the ecosystem (devices, platforms, apps). And the universal trend is for retail prices per bit to decrease over time. That is what Moore’s Law is alll about.

Gastaut’s claim--with all possible modifiers and in the context of his own business unit--that connectivity is not a commodity is correct in some sense. Whether it also is “true” is perhaps another matter.

To my knowledge, no forecaster of internet of things revenue believes much of the revenue will be earned by connectivity suppliers, in that role.

"Tokens" are the New "FLOPS," "MIPS" or "Gbps"

Modern computing has some virtually-universal reference metrics. For Gemini 1.5 and other large language models, tokens are a basic measure...