Home » Google

Category Archives: Google

Four Earnings Questions

** Note – I will be at MWC-Americas on Wednesday (all day) and Thursday morning.  Please send a note to sundaybrief@gmail.com if you would like to catch up.  Thx, Jim **

opening pic

Greetings from Lake Norman, NC (picture of a recent sunrise is shown).  This week, we will discuss four questions that should be asked during earnings calls (which start this Thursday with Comcast, followed by Verizon and Charter on Friday and AT&T and Google the following Monday, Apple and Sprint (likely) on Oct 30 and CenturyLink on Nov 6).  Please note that these questions are not in priority order.  Here’s four questions we’d like to see answered in upcoming earnings calls:

 

1. To Apple: If Goldman Sachs is correct, and the Apple Card truly is “the most successful credit card launch… ever” how will Apple use these new relationships to increase the iPhone renewal rate?

To AT&T and Verizon:  Do you anticipate that Apple’s new credit card will disintermediate the store purchase and financing experience?  If that occurs, and customers finance their new device through Apple directly, how will that impact revenues, margins, and churn?

We received a strong indication of Apple Card’s success from Goldman Sachs this week when CEO David Solomon revealed on his earnings call that “we believe [the Apple Card] is the most successful credit card launch ever.”  Solomon went on to disclose:

…we have seen a pretty spectacular reception to the card as a product. The approval rates early on have been lower, and I say that that’s a decision, obviously, Goldman Sachs is making as the bank, but we’re doing that in concert with Apple. And it is because we’re quite vigilant from a risk point of view, of not being negatively selected out of the box. Meaning, over time, we’ll start to see better credits appear and the approval rates will go up, where we’ve seen an enormous inbound, we’ve issued a considerable amount of cards. We’ve just been through our first bill cycle, which went smoothly, and so from an operational point of view, it’s gone well

Apple announces earnings on October 30th.  It’s likely that they will not actively promote anything until after the Holidays (if demand is good, and they are throttling activations through selective credit scoring, probably not best to get promotionally aggressive).  However, if Apple attracts 10 million US card holders in the first year (we would not be surprised if this happens), you have to think that the ability to finance select transactions at 0% a.p.r is inevitable.

 

2.To AT&T: Do the list of divestitures you are working on with Elliott Management include unprofitable local phone exchanges?

 To each of the other local exchange providers (particularly rural):  How will you more effectively compete against a clustered (and therefore operationally efficient) cable industry?  Was your concern over valuation when you considered clustering in the past unfounded given the deep losses that have occurred in broadband acquisition over the past decade?

 

We briefly discussed this in the TSB focused on the Elliott Memo.  In our note, we described the diseconomies of scale arising from island or isolated exchanges in North Carolina.  To prove that the Tar Heel state was not a fluke, we show below the local telephone provider exchange map of South Carolina (link here):

sctba pic

In contrast with this menagerie of local exchange properties, cable broadband providers in South Carolina consist of the following (from the South Carolina Cable TV website and company websites):

  1. Spectrum/ Time Warner Communications: 72 million population covered
  2. Comcast Communications: 600,800 population covered
  3. Comporium Communications: 305,000 population covered
  4. Horry Broadband Cooperative: 205,000 population covered
  5. Northland Communications: 164,000 population covered
  6. Atlantic Broadband: 133,000 population covered
  7. Hargray Communications: 106,000 population covered

With a total population of about 5 million, to have more than 3.6 million (72%) covered by just three providers and more than 4.3 million (86%) by seven providers shows why cable broadband has an advantage:  they have clusters which produce economies of scale.

What is the critical importance of owning and operating the telephone exchange in Florence, SC (population just under 38,000) for AT&T?  Why not pursue a structure with other phone companies in Northeast South Carolina that mirrors the one proposed by Apollo Management for combining Dish and DirecTV assets?  What efficiencies (and increased business opportunities) could be realized from greater exchange consolidation?

How bad is it (likely) for AT&T?  Look at the May 29 Frontier sale announcement of their Northwestern exchanges, where they disclosed that the sold properties passed 1.7 million locations yet Frontier only had 350,000 consumer and business customers (20% relationship penetration).  Does AT&T (U-Verse/ Internet) have a relationship with 30% of the homes passed in Florence?  What are the value prospects, and how do they fit into all of the other things that AT&T is managing?

As for the other local exchanges, how long can they compete with the new T-Mobile, who, like we discussed in last week’s TSB, is promising 100 Mbps fixed wireless service to the vast majority of the United States (including Florence) in a few years?  Is it too late to change?

 

3.To T-Mobile: How do you continue to drive increased postpaid retail gross additions?  How much of it is driven by new device launch promotions (iPhone 11/ 11 Pro/ 11 Pro Max) versus increased 600 MHz footprint?

 

We have reported for the last several weeks on the lack of availability of both the iPhone 11 and the iPhone 11 Pro Max at T-Mobile (note: in last weeks report, most of the iPhone 11 issues were driven by specific colors).  Here’s the data for this week:

iphone 11 availability as of Oct 20

T-Mobile has really been selling a lot of iPhone 11 devices.  Their shortages on the 256GB storage level have been ongoing for three weeks, leading us to believe that this may be a supply chain miss (and perhaps a sign of economic good times).  Not surprisingly for Apple, the iPhone 11 in green (and, to a lesser extent, purple) is harder to come by than more standard red, white, and black.  Now the chart for the iPhone 11 Pro:

iPhone 11 Pro availability as of Oct 20

This is also an interesting chart for T-Mobile.  As we have pointed out several times, Magenta does not have a $0 option for either the iPhone 11 Pro or the iPhone 11 Pro Max.  Our guess is that the T-Mobile shortage is continuing for all but the 512GB memory model for two reasons:  a) greater upgrades within the T-Mobile base (presumably to get the 600 MHz coverage and all of the other iPhone features), and b) some movement from other carriers (Sprint?) to Magenta.  These are educated guesses (not stabs in the dark) and should not take away from any 600 MHz progress as a factor.

AT&T’s shortages (basically out of everything that is not gold colored) are likely much more weighted to upgrades.  A lot of changes have happened in AT&T’s network since the iPhone 7 (along with the 6S, most likely phone upgraded to the 11), and the business upgrade cycle is also in full swing (spending any available budget to improve corporate liable handsets).  There may also be a small amount attributable to the FirstNet initiative as their LTE band was not included until last year’s models (XR, XS, XS Max were the first models with LTE Band 14).

Similar trends are seen with the iPhone 11 Pro Max:

iPhone 11 Pro Max availability as of Oct 20

The backlog seems to be much more manageable here than for T-Mobile.  It’s likely that T-Mobile’s iPhone 11 Max shortages are attributable to supply chain/ forecasting, and nothing more.

 

4.To all carriers (especially CenturyLink): If low latency applications are critical to the value creation of 5G (basically keeping 5G more than a special access open expense reduction), what is your edge data center strategy?

This is a particularly important question for the large wireless carriers (including T-Mobile) and enterprise focused companies such as CenturyLink (who now owns Level3 Communications).  It’s hard to remember, but there was a time when both AT&T and Verizon (and Sprint and T-Mobile) owned several data centers apiece for internal use – both AT&T (Brookfield Infrastructure partners – $1.1 billion – 2018) and Verizon (Equinix – $3.6 billion in late 2016) sold their data center assets.  Investing in dozens (hundreds?) more could be necessary, however, if no closer solutions exist.

Also of interest with respect to edge is the entrance of Pensando Systems, who announced last week that they raised an additional $145 from Hewlett Packard Enterprise and Lightspeed Partners  to fund their edge computing interests.   Pensando has now raised a total of $278 million dollars (3 rounds in 3 years) with a high degree of interest from a wide variety of potential partners.  More on the startup (certainly a candidate for our next “Companies to Watch”) in this CNBC article (John Chambers of Cisco fame is their Chairman).

Also of interest are companies such as Qwilt, an edge video server company that has raised over $65 million from various partners including Cisco, Accel Ventures, and Bessemer.  Verizon has deployed Qwilt as their application edge delivery platform.

Understanding edge strategies is critical with the increase in over the top solutions (such as last week’s Hulu 4K device announcement which broadens their base to include Amazon Fire Stick, Microsoft’s Xbox One, and the LG WebOS TV platforms).  More capabilities will lead to higher expectations and even higher consumption.

 

TSB Follow Up

Several of you issued lengthy replies to last week’s TSB.  There is no doubt that strong feelings exist supporting maintaining equal outcomes of data packets.  There’s equal certainty that others see S.B. 822 (California Net Neutrality bill) as a stepping stone to more activist state proceedings with respect to cable unbundling (which would clearly deter new incumbent investments in the Golden State).  We decided not to go there in last week’s TSB (our focus was on how wireless companies would treat throttling) but see how and why the ghost of Brand X is more than a mirage to many of you.

One item that I think is undebatable – Congressional action would clearly eliminate the newfound love of Federalism that is breaking out in many state legislatures.  We will write more on this in the future, but we at TSB offer up the following bill parameters for consideration:

  1. Establishment of a minimum residential achieved average upload and download speed (wireless and wired) above which regulations would be loosened if not eliminated (we would propose 200 Mbps for 2020 (200 Mbps for stand-alone Hotspot; 100 Mbps per smartphone or tablet) and 500 Mbps for 2025 (500 for stand-alone Hotspot; 250 Mbps per smartphone or tablet) with agreement to establish the 2030 speed at no less than 700 Mbps). Residential averages would be evaluated by no less than two independent 3rd parties at a zip code level.

 

The rationale behind this is twofold:  a) Regardless of bit prioritization practices, the presence of 200 Mbps for 4-5 simultaneous users clearly provides a healthy broadband baseline.  This would be based on achieved as opposed to advertised speeds.

 

This also provides the ability to have lower speeds but uses market mechanisms to drive the mix.  If AT&T wants to offer Gbps speeds for $90/ month, then they will have a smaller fraction achieving this higher speed than if they offered the same product at $50.  The market will reach an equilibrium.

 

This would also greatly encourage the adoption of 5G services across wireless carriers.  If 50% of the base is wireless and achieving LTE speeds of 100 Mbps, they would need to be offset by 50% of the base experiencing average speeds of more than 300 Mbps.

 

It would also create a competitive mechanism assuming either telco or cable did not achieve the figure in the first measurement.  Some incremental capital expenditure would also occur (and this can be done prior to having a larger infrastructure spending bill if that is desired).

 

  1. Tighter enforcement of Type II provisions and regulations. Unbundling provisions in the 1996 Telecom Act have been watered down to a large extent, with telcos (and, to some extent, the business arms of the cable companies) replacing the harmful operational effects of unbundling with 60-month term discounts on traditional special access services.

 

If Type II were properly enforced (penalties properly monitored and assessed), there would be more impetus to be classified as an information service.  This is a fault of all regulators – state and federal – and should be addressed.

 

  1. Adding core control to Type II provisions for wireless providers. If national or regional wireless providers do not step up their game and have market-leading data infrastructure, they should allow others to disintermediate them (core control allows a rural-focused MVNO to set up infrastructure in the slower market and use the faster speeds in more metro areas).  This “nuclear option” would certainly spur innovation among the wireless carrier community and perhaps spawn a previously unthinkable concept – spectrum/ network sharing.

These are very measurable, practical legislative remedies which refocus objectives to weighted average usage (including testing price elasticities to a greater extent) and increased competition.  We clearly believe that the current approach will create a patchwork of network procedures as well as full employment for telecom attorneys.

 

Next week, we’ll look for clues from Comcast, Charter, and Verizon’s announcements, as well as some previews for AT&T’s earnings and the Time Warner analyst day (Oct 29).  Until then, if you have friends who would like to be on the email distribution, please have them send an email to sundaybrief@gmail.com and we will include them on the list.

 

Have a terrific week… and GO CHIEFS!

 

Third Quarter Earnings – What Could Dislodge Wireless?

opening pic

Greetings from Charlotte, North Carolina (picture is, from left, Frank Cairon, formerly of Verizon Wireless and Ryan Barker, currently with Verizon Wireless enjoying some good Mexican food on Friday in the Queen City with yours truly).

 

This week’s TSB examines the short-term dynamics that could impact wireless growth in the third quarter and through the end of the year.  At the end of this week’s TSB we will also briefly examine the current state of litigations and investigations active and pending (T-Mobile/ Sprint, Facebook, and Google).

 

Follow-up to Last Week’s CBRS article

 

federated wireless logoBefore diving into earnings drivers, a quick shout out to Federated Wireless, who raised $51 million this week in a mammoth C Round financing (full announcement here).  Existing investors American Tower, Allied Minds, and GIC (Singapore sovereign wealth fund) all participated in the round, and Pennant Investors (Tim McDonald, formerly of Eagle River (Craig McCaw), will be joining the Federated board) and SBA joined with fresh cash.  Kudos to Federated CEO Iyad Tarazi for his continued leadership and perseverance.  With $51 million in additional cash, spectrum sharing gets a global boost.

 

In addition to this news, the FCC also has placed the approval and scheduling of the Private Access License auction (this is the dedicated band that gets priority over the General Authorized Access band) on the docket for June 2020 (FCC Commissioner Pai’s blog post is here).  It’s generally assumed that this means a C-Band auction will come at the end of 2020/ beginning of 2021 (although this week’s news that Eutelsat has withdrawn from the C-Band Alliance has some believing that there may be a side deal afoot).  The PAL auction timeline is in line with expectations, and it’s likely participants will include some new(ish) entrants.

 

Third Quarter Earnings – What Could Dislodge Wireless?

Speaking of expectations, there’re not a lot of dramatic changes expected in the wireless arena.  Consensus has T-Mobile leading the postpaid phone net additions race (no surprise), with Sprint struggling to keep pace, AT&T in the 0-300K range for postpaid phone thanks in large part to FirstNet gains, and Verizon, excluding cable MVNO revenues, growing their retail postpaid phone base only slightly.  With the exception of FirstNet (and a few quarters of decent Verizon growth), this is a pretty consistent story dating back to early 2017.  What events could change the equation and dislodge the current structure?

 

  1. More rapid AT&T postpaid phone net additions led by FirstNet. Here’s how AT&T CFO John Stephens summarized the relationship between spectrum rollout and FirstNet deployments at an investor conference in early August:

 

We had some AWS-3 and some WCS spectrum that we had, so to speak, in the warehouse that we hadn’t deployed. We had 700 spectrum, Band 14 from FirstNet, which the government was requiring us to deploy. And then we got a whole new set of technologies that were coming out, 256 QAM and 4-way MIMO and carrier aggregation. They were particularly important to us because of our diverse spectrum portfolio. So we got the FirstNet contract and we had to touch a tower, have to go out on the network. And we decided, with this contract, now is the time to, so to speak, do everything. Put all the spectrum in service, that’s the 60 megahertz. In some towers, it’s 50, some towers, it’s 60, but it’s 60 megahertz of new spectrum that was generally unused that we’re putting in.

 

This has the effect of increased costs, but also improved network performance.  With 350,000 net additions already from FirstNet (Stephens disclosed this in the same conference), it’s entirely possible that they could post a 100K net add surprise due to increased coverage and deployments.  In turn, improved wireless bandwidth, while driving up costs, should lower churn in areas like Detroit (RootMetrics overall winner in a tie with Verizon – first since 2012), and Boston (first RootMetrics overall win in Bean Town since 2017).

 

  1. Faster cable MVNO growth. While this week’s news was on Altice’s aggressive $20 unlimited price point for existing customers (great analysis on their strategy here), both Charter and Comcast see a lot of mover activity in the third quarter.  This would seem to be a very good time to present their wireless offer.  Here’s a chart of net additions by both Charter and Comcast for the past two years:

cable mobile net additions trend chart

While the two largest cable providers accounted for ~390K growth in 2Q (and over 1.3 million net additions growth over the last four quarters), there’s a strong likelihood that this figure could grow even greater as the attractiveness of the wireless bundle pricing takes effect.   Both Spectrum and Comcast are maturing their service assurance processes, and those efforts should lower churn.

 

Comcast also made a number of changes to their “By the Gig” plans which encourage this option for multi-line plans that use 2-6 Gigabytes per line per month (and therefore use a lot of Xfinity Wi-Fi services).  It basically amounts to a prepayment for overage services, but could be attractive for certain segments/ demographics (full details on these offer changes are here).  Charter did not follow the Comcast changes described in the link and their “By the Gig” pricing continues to be $2/ mo / gigabyte higher.

 

Both Comcast and Charter are running into Apple iPhone announcement headwinds if next week’s headlines meet expectations (no 5G, no CBRS, no special financing deals, better camera).  If the changes do not increase willingness to upgrade/ change to an iPhone, it’s going to be very difficult to craft a cable plan (even $20/ mo.) that will buck the trend.  The upgrade cycle will be extended to 4Q 2020, when both the carriers and Comcast/ Charter will have full access to 5G.

 

Our prediction is that Charter and Comcast will have 475-500K net additions in the third quarter thanks to a combination of lower churn and higher gross additions (led by increased moving activity).  Altice’s offer will add another 70K net additions in September, with those gains coming from Sprint retail and, to a lesser extent, T-Mobile retail and wholesale (Tracfone).

 

  1. T-Mobile’s 600 MHz coverage (and gross add) improvements. As T-Mobile, Sprint, and the state Attorneys General try to find a resolution to their quagmire, T-Mobile keeps on deploying 600 MHz spectrum.  Here’re their reported (through Q2 2019) and estimated (Q3/Q4) progress:

t-mobile 600 MHz chart

Every new device on the T-Mobile.com website (and every T-Mobile store) is 600 MHz/ LTE Band 71 capable.  Many older devices are not, however, and that is preventing greater market share gains in secondary and tertiary geographies (many/ most BYOD Android devices from AT&T, for example, will have the 700 MHz but not have the 600 MHz band).  The expected Apple announcement represents a slight headwind for T-Mobile as well.

There’s a natural gross add/ upgrade path that follows 145 million coverage growth over a 12-month period.  Assuming T-Mobile keeps their churn rate at 0.8%/ month over 3Q (2.4% of the ending branded postpaid 2Q base would be ~1.07 million disconnections), they have grown their 600 MHz marketable base by 20 million from Q1 to Q2 and by another 50 million from Q2 to Q3.  If they just grew their penetration in the 600 MHz band by 1.5% for Q1-Q3 incremental POPs (or 0.5% penetration for the entire estimated 3Q 2019 footprint), they would negate the entire estimated branded postpaid churn for the rest of the country.  This ex-urban/ rural growth opportunity is unique to T-Mobile and would at the expense of AT&T and Verizon.

Offsetting the 600MHz growth is small cell progress.  T-Mobile committed at the beginning of the year to deploy 20,000 incremental small cells in 2019, but that guidance was withdrawn in their Q2 Factbook with H1 growth of only 1,000.  That leaves a very large backlog of in-process capital (excluding capitalized interest, total capital spending was $3.48 billion vs an estimated spending range at the high end of $5.4 – $5.7 billion).  T-Mobile should spend at least $2.2 billion in capital spending in the second half of 2019 on 5G, 600 MHz, and other initiatives and will undoubtedly be left with a lot of in-process small cell deployments.

 

  1. Sprint’s prepaid and postpaid churn. There’re a lot of headwinds for Sprint in the third quarter – overall 2Q churn trends are higher than previous year’s, and no one expects the seasonal respite to last long.  It’s likely that 3Q postpaid churn could exceed 1.85%, led by postpaid phone churn of a similar level (look for late September promotional activity).

 

Prepaid churn is a bit tougher to forecast and will also be tracked closely.  If the postpaid churn comes in below 2Q levels, check the prepaid recategorizations to postpaid (they were 116K in Q2 and 129K in Q1).  Both prepaid and reclassified postpaid accounts will be transferred to Dish assuming the merger goes through, but it makes the postpaid headline number more palatable.

 

The 30-day guarantee promotion, according to most reports, has been an ineffective switching tool.  (Sprint’s 5G rollout success has been much more impactful).  Expect Sprint to say very little until there is clarity on the litigation, and to post greater than expected losses in prepaid subscribers as they preserve their marketing dollars for a post-merger world.

 

Litigation Tracker:  Why the Facebook, Google, and T-Mobile/ Sprint cases are not all the same

Speaking of litigation and investigation, we were very dismayed that media sources are choosing to lump the New York-led Facebook investigation announced Friday, the to be announced Texas-led Google investigation, and the on-going T-Mobile/ Sprint (TMUS/S) litigation into one mega-story.  While there are some similarities, the upcoming Google action is broader than Facebook and more bipartisan than the TMUS/S complaint.

As most of you who are following the TMUS/S suit know, the states of Oregon and Illinois recently joined the original 14 states and the District of Columbia to block the merger.  (As an aside, Fox Business is reporting that the state AG group is focusing on the inexperience and shaky financial condition of Dish Networks as opposed to what would have been an uphill market concentration battle).  In fact, in the original TSB concerning the lawsuit (here), we were surprised by the absence of Illinois.  The figure below shows who is involved in what (underlined states are named in the Facebook litigation):

litigation tracker chart

To recap, there are 40 states + the District of Columbia named in the National Association of Attorneys General comment letter to the FTC (filing here), and at least 30 of them are joining a Texas-led lawsuit against Google to be announced early this week.

The makeup of the states in the FTC letter is very bipartisan:  14 Republican and 26 Democrat attorneys general.  All of the states involved in the TMUS/S litigation are also named in the FTC comment letter.  To contrast, of the 17 states in the TMUS/S litigation, only Texas (AT&T HQ) is Republican and the other 16 are Democrat.

As we stated in the TSB on the AG lawsuit, very few sparsely-populated states, regardless of political affiliation, are participating in the T-Mobile/ Sprint litigation.  Fourteen of the fifteen least densely populated states in the US (data here) are named in the Google/ FTC letter.  However, only two of the fourteen (Colorado, Oregon) are participating in the TMUS/S litigation.  The promise of a rural solution outweighs the benefits of a fourth carrier in metropolitan and suburban areas.

The Facebook investigation is also widely bipartisan with five Democrats and four Republican states represented.  Florida is involved in the Facebook investigation but none of the other two legal activities.  In addition, there are nine states (including New Jersey, a Democrat stronghold) that are not currently participating in any legal activity.

Bottom Line:  Concerns about Google’s anti-competitive practices are supported by a large number of state Attorneys General and are very bipartisan.  A subsegment of Democrat AGs (and TX) is also a part of the T-Mobile/ Sprint lawsuit.  And an even smaller subsegment plus Florida is a part of the recently announced Facebook investigation.

Next week, we will highlight some wireline trends and talk about overall profitability across the telecommunications sector.  Until then, if you have friends who would like to be on the email distribution, please have them send an email to sundaybrief@gmail.com and we will include them on the list.

 

Have a terrific week… and GO CHIEFS!

Deeper: CBRS – Share and Share Alike

cbrs top 10 slide

The following articles provide a good overview of CBRS.  There are a lot of things changing with CBRS each day, and we will be sure to keep the list updated as new developments occur.

  1. One of the best government sites is the NTIA CBRS page here. Included in this is the original NTIA October 2010 Fast Track Report cited in this week’s TSB
  2. The FCC Report and Order on CBRS from 2015 is here.
  3. This Ruckus on-line tutorial is a terrific place to start if you need a 20-minute high level overview of CBRS.
  4. The CBRS Alliance has a page devoted to certified devices that is updated frequently. The latest certified devices are here.
  5. Light Reading had a recent article summarizing AT&T’s, US Cellular’s and Verizon’s outdoor interests. It’s here.
  6. Charter’s SVP Technology Craig Cowden’s address at MCWA in 2018 is here (20 min). A good summary of the value of CBRS to the cable industry.  Charter’s FCC application for their CBRS tests is here.
  7. Adam Koeppe’s June interview with RCR Wireless concerning 5G, CBRS, MIMO, Beamforming and more is here. The Light Reading article referenced in the TSB is here.
  8. A good summary of MIdCo’s 2018/ 2019 trial with Telrad and Federated Wireless from Fierce Wireless is here.
  9. Telecompetitor interview with Federated Wireless CEO Iyad Tarazi is here. In this article, Iyad describes Federated’s role in the Kinetic Edge Alliance, which we ran out of time to discuss and is a very interesting intersection point between computing, proximity, and network sharing.  More on that group here.

CBRS – Share and Share Alike

opening pic

Greetings from Lake Norman/ Davidson, North Carolina, where the college football season has started.  We took in the Davidson College home opener and the Wildcats (red jerseys) defeated Georgetown 27-20 to a crowd of more than 2,300.  It was an exciting part of the Labor Day weekend and a good win for the Cats.

 

This week’s Sunday Brief focuses on the potential of Citizens Broadband Radio Service (CBRS) to change the telecommunications landscape.  We will also have an update on C-Band spectrum auction news.  First, however, a quick follow up to last week’s article on AT&T’s system and network architecture changes.

 

Follow-up to last week’s AT&T article

We had greater than expected interest concerning last week’s TSB including receiving several background articles that we had not uncovered in our research.  One of the most important of these was a blog post by AT&T Senior Vice President Chris Rice on their Domain 2.0 developments that was posted on August 21.  In this article, Chris describes their major architectural change:

 

We started on a path for a single cloud, called AT&T Integrated Cloud (AIC). This was our private cloud, meaning we managed all the workloads and infrastructure within it. Originally, AIC housed both our network and several of our “non-network” IT workloads and applications.

But we quickly learned it wasn’t optimal to combine both types of workloads on a single cloud. It required too many compromises, and the IT and network workloads needed different profiles of compute, network and storage.

We opted for a better approach: Create a private cloud for our network workloads, optimize it for those workloads, and drive the software definition and virtualization of our network through this cloud approach and through the use of white boxes for specific switching and routing functions.

 

The change to last week’s article is subtle but not insubstantial:  AT&T’s network cloud (formerly AIC) is optimized for network traffic loads and functions (but still built on white box/ generic switching and routing), while non-network functions are operated in the public cloud through Microsoft and IBM.

 

john-donovanIt’s important to note that the executive champion of this cloud strategy, John Donovan, is going to be retiring from AT&T on October 1 (announcement here).  We have included an early speech he gave on AT&T’s Domain 2.0 strategy in the Deeper post on the website.  John brought engineering discipline to AT&T’s management, and, while the parlor game of his replacement has begun, the magnitude of his contributions to Ma Bell over the past 11+ years should not go unnoticed.

 

CBRS – Share and Share Alike

When we put together a list of Ten Telecom Developments Worth Following in mid-July (available on request), we were surprised by a broad range of CBRS skepticism in the analyst community, especially given the breadth of US wireless carriers playing in the CBRS alliance.  “Nice feature” or “science experiment” was the general reaction.  Many of you chose instead to focus on the C-Band auctions, which are important and addressed below.

 

After some reflection, we have come to the conclusion that the most important feature of CBRS is neither its mid-band position (3.5 GHz), nor the mid-band spectrum gap it fills for Verizon Wireless (more on that below), but the fact that at times all of the spectrum band can be shared.  Customers receive the benefits of an LTE band without a costly auction process.

 

If you are intimately familiar with CBRS, you can skip the next couple of paragraphs.  For those of you new to TSB or the industry, here’s a copy of the slide we used to describe CBRS in the Ten Telecom Developments presentation to start your education:

cbrs top 10 slide

 

 

The commercialization of shared LTE bands is pioneering and one of the reasons why it has taken nearly a decade to move from concept to commercialization (the original NTIA report which identified the CBRS opportunity is here).  This does not appear to be a singular experiment, however, as Europe is proceeding with shared spectrum plans of their own in the 2.3-2.4 GHz frequencies (more on that here).

 

To enable this sharing mechanism in the United States, a system needed to be developed that would prioritize existing users (namely legacy on-ship Navy radar systems) yet allow for full use of the network for General Authorized Access (GAA) users when prioritization was not necessary (opening up to 150 MHz of total spectrum for GAA which could power 5G speeds for tens of millions of devices nationwide).  A great primer on how CBRS generally works and how spectrum sharing is performed is available here from Ruckus, a CommScope company and one of five Spectrum Access System providers.

It’s important to note that the Environmental Sensing Capability (which determines usage by priority) and the Spectrum Access System (which authorizes, allocates and manages users) are two different yet interoperable pieces of the CBRS puzzle.  And, while the ESC providers have been approved by the NITA (CommScope, Google, and Federated Wireless), the SAS providers have not been approved (more on that here in this Light Reading article).  While all of the SAS providers have not been made public, it’s widely assumed that they include the three ESC providers mentioned above.

 

Delays in the SAS approval process have not kept the CBRS Alliance from heavily promoting a commercial service launch on September 18 (news release here).  This event will feature FCC Commissioner Michael O’Rielly, Adam Koeppe from Verizon, Craig Cowden from Charter, and others who will celebrate the Alliance achievements to date and place the development as a central theme going into 2020.

 

CBRS Use Cases:  Not Everyone is Waiting for Private Licenses

The myriad of CBRS use cases mirror the different strategies for telecommunications industry players.  Here are four ways carriers are using CBRS in trials today:

 

  1. CBRS as a last mile solution for rural locations (AT&T and rural cable providers MidCo Communications mapMidCo and Mediacom Communications). In the MidCo configuration, outdoor Citizens Band Service Devices or CBSDs (see picture above) are placed in proximity to potential (farm) homes passed (see nearby map of MidCo territories in the Dakotas and Minnesota).  Per their recent tests, MidCo was able to connect homes up to eight miles from the outdoor CBSD.  They estimate that CBRS will add tens of thousands of homes and businesses to their footprint (they serve 400K today so every 10K new customers is meaningful).  Good news for an over the top service like Hulu, Netflix, and YouTube TV and bad news for DirecTV and Dish.

 

AT&T has been testing CBRS as a similar “last mile solution” in Ohio and Tennessee using equipment from several providers including CommScope (ESC, SAS) as well as Samsung (network).  These trials are expected to wrap up in October.  If AT&T can find a more effective last mile solution for copper-based DSL in rural areas, revenues and profitability will grow (by how much depends on Connect America Fund subsidies and service affordability).

 

AT&T has been mum on their trial progress to date.  In June, however, AT&T asked the FCC to allow them to turn up antenna power in these markets to test various ranges and speeds (bringing the power allowances to a similar level of the WCS spectrum that AT&T already owns and operates in the 2.3 GHz spectrum frequency).  This was met with strong opposition by a coalition of providers, and it’s not clear that AT&T’s request was ultimately granted.  More on the AT&T request and response can be found in their FCC filing here, and in this June 2019 RCR Wireless article.

 

  1. CBRS as an additional LTE service for cable MVNOs (Altice, Charter, Comcast). It’s no secret that cable companies are eager to continue to grow their wireless presence within their respective footprints (and corporate is equally as eager to improve profitability and single carrier dependency).  CBRS would add a secure option that is seamlessly interoperable with other LTE bands to create an alternative to their current providers (Verizon, Sprint, etc.).  It also provides a new “secure wireless” service for small and medium-sized businesses which can be deployed with Wi-Fi.  A cheaper alternative for out-of-home wireless data?  Count cable in.

 

We spent some time a few weeks ago talking about the evolution of Verizon plans, specifically how their cheapest unlimited plan now includes no prioritized high-speed data (article here).  Is CBRS a better alternative to deprioritized LTE?

 

The short answer is “not yet.”  LTE Band 48 is only available across the most expensive devices, and, presumably, if customers can shell out $1000-1500 for a new device, they can probably afford extra LTE data allowances above 22-25 Gigabytes (see previous article linked above).  Notably, the new Moto E6 (budget-minded Android device) includes neither CBRS nor Wi-Fi 6 (specs here).  The new ZTE Axon 10 Pro phone does not include Band 48 or Wi-Fi 6 (specs here).  The OnePlus 7 Pro, however, does include Band 48 but not Wi-Fi 6 (specs here).  And, if rumors are to be believed, the upcoming Apple device announcement in a couple of weeks will disappoint everyone – no 5G, no CBRS even though the new device will likely support the 3.5GHz spectrum band in Japan, and likely no Wi-Fi 6 (which is why the Apple Card will likely be used to offer attractive financing options).

 

This leaves cable companies with a good selection of Samsung and Google devices that can use CBRS (Galaxy Note 10, Galaxy S10 5G, upgraded Pixel 3X and 3XL, and likely the upcoming Galaxy 11 release).  For cable to win on this front, they may need to provide plan incentives to influence the pace of upgrades and request this band for Moto and low-end Samsung devices.

 

  1. CBRS as a mid-band LTE outdoors/ public venue solution for Verizon. It is no secret that Verizon is going to use CBRS GAA as a part of their carrier aggregation solution (see this August 2019 article from Light Reading for more details). Such a solution is usually not designed for greater throughput in rural markets alone – Verizon clearly sees some form of CBRS as a portion of their overall licensed/ un-licensed solutions portfolio.

 

The question that will be answered in the next quarter or so is whether CBRS is valuable enough to be a part of Verizon’s licensed portfolio (e.g., they buy 20 or 30 MHz worth of private CBRS licenses, or whether they use the GAA portion in the same License Assisted Access (LAA) manner as they use 5.0 GHz Wi-Fi).  It’s likely that if CBRS is important, they will be at the auction table.

 

It appears that the C-Band (3.7-4.2 GHz) license quantity and auction schedule is in flux, if the latest report from Light Reading (and the corresponding New Street Research analysis references in the article) is true.  This also impacts Verizon’s near-term interest in CBRS PALs.

 

Verizon’s interest is important as they can drive manufacturers to quickly include Band 48 in devices.

 

  1. CBRS as an indoor and/or private LTE solution for wireline. One lesser-discussed option for CBRS is as a private LTE indoor solution for enterprises and building owners.  While we touched on this option for cable companies (who will undoubtedly drive business ecosystem development), this could also have ruckus cbrs routerinteresting implications for companies like CenturyLink (Level3), Masergy, and Windstream.  Ruckus, a traditional Wi-Fi solutions provider now owned by CommScope, already has an indoor unit for sale here (picture nearby).  The implications for in-building coverage are significant because 3.5 GHz does not overlap with existing deployed frequencies (including existing 2.4 GHz and 5.2 GHz Wi-Fi solutions), and, as a result, will not increase interference that deploying an AWS (1.7/ 2.1 GHz) or EBS/BRS (2.5 GHz) might create.  With solutions as cheap as industrial Wi-Fi and minimal interference concerns, there might be more value created with CBRS indoors than outdoors.

 

Bottom line:  CBRS is a real solution for rural broadband deployments and will attract the interest of large and small rural providers.  CBRS will be important to wireless carriers when Samsung and Apple join Google and Facebook in building a robust ecosystem.  This is a good/great but not an industry-changing technology.  If the first commercial applications are successful in 2019 (the odds are good), demand for Private Licenses will be significant (if not, expect more pressure to resolve C-Band spectrum allocation issues quickly).

 

What makes CBRS great is dynamic spectrum sharing among carriers.  Should that continue with all new frequencies auctioned (including C-Band), you should expect to see competitive pressures grow in the sector, particularly with private LTE/ indoor applications.

 

Next week we will provide the first of two third quarter earnings previews, focusing on wireless service providers ahead of the Apple event.  Until then, if you have friends who would like to be on the email distribution, please have them send an email to sundaybrief@gmail.com and we will include them on the list.

 

Have a terrific week!

 

Deeper – Fiber Always Wins (Until it Doesn’t)

Mission Hills notice on Google Fiber

The following articles provide a good overview of the role fiber plays in the telecommunications ecosystem today (there’s some good details on Google Kansas City and Louisville setbacks as well):

  1. An RCR Wireless article that details Verizon’s “integrated engineering process.” It also has some details on the fiber deal with Corning.
  2. Corning (Bob Whitman) and Verizon (Glenn Wellbrock) conversations from YouTube part 1 and part 2.
  3. Altice press release announcing first homes on Long Island to receive 960 Mbps symmetrical speeds for $80/ month.
  4. Google missing their deployment goals for Kansas City, KS and Mission Hills, KS
  5. Telecompetitor AT&T Fiber penetration article in which they talk about doubling market share. The map see in the article (AT&T Fiber cities) is here.
  6. Cincinnati Bell 2Q investor penetration showing that AT&T’s goal of achieving 50% market share is entirely possible with the right content bundles (they are at 44% without the content – see page 7).
  7. Blair Levin’s and Larry Downs Harvard Business Review article on Why Google Fiber failed and how it’s really a success in disguise. Rationalization gone awry.  This might have been the case if their more recent deployments had not damaged their brand.
  8. Local (Louisville Courier-Journal) coverage of Google’s decision to pull out of the Louisville market is here.
  9. Google’s plans to expand their Webpass (60 GHz wireless hub and spoke) service for Multi-Dwelling Units Austin is outlined here.
  10. The Dallas Morning News rant on Frontier’s failures is a classic and outlined here.

Fiber Always Wins (Until it Doesn’t)

opening pic

Greetings from Davidson, North Carolina, where the hazy days of summer will soon give way to the bustle of orientation.  There’s plenty to cover in this week’s TSB, and our main topic will be on the importance of fiber in the telecom industry.  But first, a brief comment on Verizon’s wireless pricing changes.

Before diving into commentary, a quick reminder that we will be posting a “Deeper” section for each TSB on the website.  Here is the Deeper section from last week’s TSB (State AG’s case against the T-Mobile/ Sprint merger).  We usually post these by Tuesday evening.

 

Verizon’s Bouquet of Plans – Something for Everyone?

When Verizon announced their pricing plan changes on Friday, August 2, the collective response was “Why?”

verizon plan description pic

A deeper inspection of each change explains Verizon’s strategy.  They are continuing to lead with a plan that includes no fixed allotment of high-speed data (the Go Unlimited legacy plan was changed as well to “Day 1” deprioritized data).  Verizon also put a stake in the ground:  the minimum rate for any individual 5G plan (or first line in a multi-line scenario) is $80/month.  Why one would purchase a Start Unlimited plan for $80, and forego a fixed allotment of high-speed data (and Hotspot capabilities) offered by the Do More Unlimited (50GB, 15GB can be Hotspot) or Play More Unlimited (25GB) is a head-scratcher.  Bottom line: the 5G pricing tier starts at $80/month with no 4G LTE premium data allotment.

The other interesting development (outside of the $5/ line/ month across the board price decrease) is the continuation of 720p HD video streaming on Verizon’s best plan.  Customers can get 1080p 4G LTE streaming for an additional $10/ month (total cost of $100/ month with taxes & fees versus T-Mobile’s 1080p Magenta Plus plan which includes 50GB of premium data as well as taxes and fees for $85/ month or AT&T’s 1080p Unlimited & More Premium plan which includes 1080p but only has 22 GB of premium data allotment for $80/ month).  It baffles me that Verizon sells devices with amazing screen solutions like the Samsung Galaxy Note 10 (3040 x 1440 pixel screen – Quad HD+) and the Apple iPhone XS Max (2688 x 1242 pixel resolution), advertise (and win) speed test awards, yet limit video streaming to ~5-8 Megabits/ second.  In the past, the argument was that smartphone resolutions weren’t good enough to tell the difference, but manufacturers have caught up and 5G adoption is still several quarters away for most wireless customers.

Bottom line:  There are likely no broad implications to the wireless industry from Verizon’s recent pricing changes.  This should not trigger a response from AT&T, T-Mobile or Cable (although making Full HD (1080p) standard on all pricing plans would be a very un-carrier move for T-Mobile).  It’s likely that if AT&T were to change their pricing plans as a part of a broader 5G announcement, they would align monthly rates to speeds (e.g., Cricket Wireless is cheaper but download speeds are capped).  Big Red is in a holding pattern until they can fully deploy 5G (more on that below).

 

 

Fiber Always Wins (Until it Doesn’t)

While I was running Access Management for Sprint ($3+ billion annual expenses at the time), we would frequently discuss fiber opportunities:  bankruptcy-driven system purchases, sub-duct IRU availability (sometimes packed with fiber), joint builds with other carriers, fewer fibers with Dense Wave Division Multiplexing, etc.  Our motto, developed by our lead engineer, was “Fiber Always Wins.”  Fiber drove our investments, first connecting to the incumbent Local Exchange Carrier (LEC), then to commercial buildings, data centers, wireless switching centers, small cells and cell towers.

Meanwhile, Verizon was building their FiOS network which at peak (2015, prior to the sale of former GTE properties to Frontier) passed 20 million homes.  While the current number of homes passed is not published, most analysts peg the post-Frontier total at 15.5-16.1 million homes.  Given Verizon’s current FiOS internet count of 5.84 million, they have approximately 37% share, with about 5-10% subscribed to copper or wireless alternatives and the remainder (50-55%) subscribing to cable.  FiOS comprised 87% of the total 2018 Consumer Markets division revenue.

Verizon’s FiOS fiber experience and the acquisition of XO Communications led to the formation of the One Fiber initiative (2019 RCR Wireless article on Fiber One is here which includes details of Verizon’s multi-year purchase agreement with Corning).  This cross-company cataloguing and joint planning is reshaping how Verizon invests billions of annual capital dollars nationwide.  As a result, as they stated in their most recent 10-K, “We expect our “multi-use fiber” Intelligent Edge Network initiative will create opportunities to generate revenue from fiber-based services in our Wireline business.”

Bottom line:  FiOS fiber deployments throughout the Northeast are extremely valuable and provide Verizon with a time-to-market advantage.  Nationwide, the fiber that is being deployed for 5G will improve wireline/ Ethernet/ cloud competitiveness.  One division feeds the other.

 

ACSI Internet service provider results by year.png

Customers love FiOS.  The nearby American Customer Satisfaction Index (ACSI) Internet Service Provider satisfaction survey results clearly show that both Verizon and AT&T (U-Verse fiber now passes more than 12 million homes) are preferred over cable.  Notably, Verizon’s lead over Altice (Long Island cable provider) is back to seven points.  This is one of several factors driving Altice to deploy fiber to the home (the only US cable company to commit to a 5 million home buildout).

Similarly positive results can be seen from the most recent J.D. Power Residential Wireline Survey, with Verizon taking top honors for both Internet and Television service.  Fiber speeds result in higher spending per household, and the long-term maintenance costs for fiber have been proven to be much lower than copper.

 

When does fiber not win?  Overbuilders using fiber are having mixed results.  The most heavily marketed competitive fiber provider over the past ten years, Google Fiber, has been a dud – there’s no two ways about it.  Experts like Blair Levin and Larry Downs (Harvard Business Review article here) believe that Google was using the Fiber effort simply to prod incumbents to accelerate their broadband deployments, but that’s inaccurate – they are still investing, expanding, and trying to be a disruptive broadband provider in their markets.

Google Fiber is generating the majority of Alphabet’s “other revenues” ($162 million in 2Q) and contributes in a small part to the content commitments needed to make YouTube TV affordable (Google TV customers are less than 75,000 and YouTube TV likely has more than 1 million paying subscribers).  Analysts estimate that Google Fiber probably serves no more than 475,000 broadband subscribers today, with 20% of those in their first market – Kansas City.  Google’s efforts to expand in Louisville using nano-trenching (two inches into the asphalt as opposed to six for micro-trenching) failed miserably and, as a result, Google Fiber ceased operations in Louisville last April (CNET article here).  Their acquisition of Webpass, which uses 60GHz spectrum to transport wireless data between fiber-fed buildings, is finally beginning to expand beyond their legacy markets to Austin (see blog post here).

Google Fiber would best serve its shareholders by either a) selling the asset or b) initiating efforts to wholesale fiber to wireless providers and business-focused managed service providers.  The value of deployed fiber in highly desired suburban neighborhoods would be welcomed by a variety of telecommunications companies.

Which brings us back to Verizon.  Could they have a joint build and/or fiber swap relationship with Google?  At least enough to enable Webpass (through Verizon’s XO communications asset) to expand rapidly to 60/80/100 markets? That would shake things up considerably, and multi-dwelling units are not a target of Google’s 5G initiative.  It isn’t lost on most telecom industry followers that Google decided not to initiate any builds in Verizon’s legacy Northeast telco territories.  Given their fairly strong corporate relationship (Pixel, YouTube TV), why not deploy together?

While Google was busy doing their initial rollouts, Frontier was acquiring the former GTE properties from Verizon.  They did this in two primary transactions:

  1. In 2009, Frontier announced the purchase of 4.8 million access lines from Verizon (all former GTE except for West Virginia which was a legacy Verizon asset) for $8.6 billion (~$1800/ line). Verizon received $5.3 billion in Frontier stock from the transaction – the remainder was in cash and assumed debt;
  2. In 2015, Frontier completed the purchase of GTE properties by purchasing the assets of California, Florida, and Texas (which included 1.6 million FiOS Internet and 1.2 million FiOS TV subscribers) for $10.5 billion ($9.9 billion in cash and $0.6 billion in assumed debt).

At the end of 2016, Frontier’s service territory looked like this (5.4 million customers including 4.1 million broadband lines spread across 29 states):

frontier service area map 2017.png

Since that time, the rest of the broadband industry has continued to grow, but Frontier has managed to lose 1.1 million customers (> 20% of its base) including 0.5 million broadband customers (> 12% of the base) in a mere 10 quarters.   Customer churn is rising and currently exceeds 25% on an annualized basis.  Even though the fiber base is being upgraded to 10 Gbps capacities, the FiOS base continues to lose customers.

We talked at length about Frontier’s woes in the TSB titled “What’s Changed Over the Past Three Years?” Last week’s stock performance (down an additional 27%) has significantly narrowed Frontier’s options.

Bottom line:  When does fiber not win?  When it’s the secondary/ tertiary focus of the company (Google), and when it’s hampered by the pressures created by aging copper plant (Frontier).

Billions of dollars of value have been created from fiber deployments, and local exchange providers (AT&T and Verizon) should be able to leverage previous fiber deployment for small cell and other high-bandwidth deployments.  Fiber wins… most of the time.

Next week’s issue will explore the industry implications of Apple’s new credit card (CNBC teaser article here).  If you would like a copy of either the Top 10 Trends or the IoT Basic presentation discussed in last week’s TSB, please let us know at the email below and we’ll get you a copy.

Until then, if you have friends who would like to be on the email distribution, please have them send an email to sundaybrief@gmail.com and will include them on the list.

Have a terrific week!