Greetings from the Mile High City, the Queen City, and the City of Angels. Attendees of the Mobile World Congress – Americas were greeted with multi-story advertisements on buildings touting original content, a reminder that investment in the TMT (telecom/ media/ technology) industry can take many forms (more on this thought when we discuss Verizon’s earnings below).
I had the chance to visit with many of you during the show, and, to a tee, no one was excited to be there. “Where’s T-Mobile?” was a frequent question, usually accompanied by comments about slow vendor payments or delayed decisions. Several had “I told you so” comments about Nokia (a summary of their bad week is best captured here). And the fact that Sprint had the main entrance exhibit led some to theories that should be reserved for late night cable news.
Here’s the bottom line from the show: Our industry is changing – a lot. The mobile handset, and the licensed and unlicensed spectrum that it connects to, will be life-changing for nearly all who use it. And software (largely not present at the show), not hardware, will define value. More on this can be found from our earlier column “About This Thing Called 5G” in which we define the 5G value statement as “More software… doing more things… faster and better.” Networks matter (we will see their importance below), but software fuels their engine.
This week, we will look at three companies who announced earnings (Verizon, Comcast, and Charter) and examine the differences in growth strategies versus their industry peers. As a reminder, T-Mobile and AT&T announce earnings next week, with CenturyLink, Sprint and others following later in November.
Verizon: Consumer Wholesale and Fiber Save the Day and Seed Long-Term Options
Verizon reported strong earnings on Friday, led by wireless service revenue growth and expense reductions (nearly 14,000 fewer employees in the last 12 months alone). They had balanced growth in wireless net additions between consumer and business (and phone net adds in both, including public sector, which was a shot across the bow to AT&T’s FirstNet initiative). And, while they paid down debt, they continued their significant capital spending ($12 billion year-to-date at the consolidated level, up slightly from 2018).
Rather than go deep on each statistic, let’s summarize three areas where Verizon and AT&T are pursuing markedly different strategies:
Wireless wholesale revenues (found in the Wireless Historical Financial Results) are a key source of Verizon’s revenue growth. Verizon’s consumer income statement shows growth from multiple sources:
Approximately $376 million of 3Q 2018 to 3Q 2019 growth comes from the Operating Revenues – Other line. Excluding equipment, operating revenues grew $808 million. Around half of the operating revenue growth is coming from wholesale, which has no CPGA and carries ~65% EBITDA margins.
Interestingly, segment EBITDA grew a paltry $77 million driven largely by an increase in equipment subsidies ($136 million in 3Q 2018 vs. $341 million in 3Q 2019). Excluding the equipment subsidy, the $77 million EBITDA growth becomes $282 million (a good proxy for wireless service margin growth). So, while revenues grew by $808 million year-over-year, only 35% translated into cash.
Verizon’s CFO Matt Ellis addressed the equipment subsidy in the earnings call, saying:
So on the wireless cost of service side, I mentioned that the phone net adds split was fairly even between Consumer and Business. Business had a more than 10% increase in phone gross adds. … a lot of our Business customers are still on a subsidy model rather than device payment model, so I think you see the impact of that.
Adjusting for subsidies, and assuming a very conservative 65% EBITDA margin on wholesale revenues, it’s highly likely that 85% or more of the 3Q 2018 to 3Q 2019 EBITDA improvement came from wholesale ($376 million * 65% = $244 million EBITDA/ $282 million = 87% of EBITDA growth comes from Other revenues).
On a 2Q to 3Q 2019 sequential basis, the impact of wholesale is even more dramatic, with the unit accounting for $312 million of the $466 million non-equipment revenue growth (67%) and, assuming EBITDA margins of 65% on the $312 million, more than 120% of the subsidy-adjusted sequential EBITDA growth ($119 million + $42 million of increased subsidy = $161 million).
What this says is that the negative margin impact from consumer retail growth (and write-downs thanks to new unlimited pricing plans) is being covered by consumer wholesale. Without cable and Tracfone, the story line would have been very different.
The other major item that escaped the headlines is the continued fiber build in 60+ metropolitan areas outside of the Verizon franchise territory. At the end of the earnings conference call, Matt Ellis called out the fiber impact:
… we’re rolling out more fiber as you know in our One Fiber initiative that is going to give us more opportunities to sell into those customers as they move off of legacy products and our fiber build has continued to gain momentum, increased at a little bit in the third quarter of over 1500 route miles a month on average in the quarter. So we’re getting to a good momentum there and that will open up additional opportunities for us as we go forward to replace those legacy volumes
An additional 20-40K route miles of fiber hitting the market starting early next year will not go unnoticed and marks a very different strategy from AT&T out of region, particularly for enterprise customers. As that number accumulates (1,000 route miles means a lot in a city like Birmingham, AL or Albuquerque, NM). And, as others have correctly noted, that fiber is largely being connected to wireless telecom infrastructure and not commercial real estate (an entirely different build/ approval process). But it’s different because it’s a highly leverageable asset (this is a “Fiber Always Wins” case to quote a previous TSB article). We have previously talked about the impact of the CenturyLink build (4.7 million fiber miles), but not as much about Verizon’s One Fiber initiative.
Lastly, the Disney+ announcement stands in stark contrast to what we will likely hear from Time Warner/ AT&T executives on Tuesday (free HBO for AT&T customers started to leak last Friday – see CNBC article here). Content production with telco cash balances is unknown territory. At a minimum, it dilutes management focus from strategic items like infrastructure buildout and fiber competitiveness. More likely, it locks in AT&T to HBO and other Time Warner content at the expense of other options (as opposed to Verizon’s “playing the field” strategy of Apple Music last year, Disney+ this year and next, and maybe something entirely different in 2021). We will have more to say on Disney+ vs. HBO Max in next week’s TSB (and why we think Apple Music lessons learned over the last 12 months prepares Verizon for a very successful Disney partnership), but it’s worth thinking about the value of the content selection option.
Bottom line: Verizon’s earnings message was focused on wireless service revenue and content deals. Their profit growth is increasingly being driven by non-retail sources, however. One Fiber could translate into Enterprise market share gains if management quickly re-builds their local out of region wireline capabilities. Verizon looks less and less like AT&T each day.
Comcast: Residential Broadband Dominates – Now What About Wireless?
Comcast had a record quarter on broadband growth, with 359K residential and 20K business net additions. The 359K figure is the highest net additions for any quarter since Q1 2017 (first quarters tend to be promotion-driven and Comcast was in the middle of DOCSIS 3.1 adoption at that point – no such tailwinds existed in 3Q 2019), and according to Brian Roberts, the most for any third quarter in ten years. Penetration of homes and business passed grew to 48.2%. Not only did subscribers grow, but total revenue grew as well, as existing customers upgraded their service and promotions expired.
The talk track for Comcast residential broadband is this: New homes/ dwellings are being built, and Comcast is grabbing disproportionate market share from AT&T U-Verse (Chicago/ Houston), Century Link (Seattle/ Utah) and Verizon FiOS (Boston/ Philadelphia/ Washington DC). In turn, existing customers are increasingly satisfied with their products and services, which keeps non-mover churn in check and increases bundling and upgrades. We are ready for all broadband challengers.
To put the 359K net additions in context, Verizon’s FiOS unit grew 30K net new additions from Q2 to Q3 2019 yielding a 12:1 advantage. Including DSL, Verizon lost 5K residential customers in 3Q 2019, a figure 10K worse than 3Q 2018.
Over the last nine months, Verizon broadband has gained 9K residential broadband customers (106K net FiOS less 97K DSL losses). This compares to total residential broadband gains at Comcast of 893K – a 99:1 advantage (!).
One would think that Brian Roberts and Michael Cavanagh could drop the microphone and walk away. And, had Altice not deployed a very aggressively priced MVNO on Long Island using Sprint’s network, most analysts would have changed their questions to deal with softball topics like theme parks, debt and buybacks. The Altice deployment drove many questions, including (TSB paraphrases of the questions based on the conference call transcript):
Doug Mitchelson at Credit Suisse: How are you leveraging your base to get better pricing, and can Comcast implement strand mounts (a la Altice) to improve their cost/ GB and improve Verizon’s coverage?
Brett Feldman at Goldman Sachs: Is the MVNO unit driving up technical and product support costs in the quarter and how will continued growth impact the fourth quarter?
Jennifer Fritzsche at Wells Fargo: How does Comcast view the upcoming CBRS (Preferred Access License), C-Band and Millimeter Wave spectrum auctions? Will spectrum ownership be a part of Comcast’s strategy going forward?
Craig Moffett at MoffettNathanson: How will Comcast use eSIM (specifically dual SIM/ dual standby) to improve their wireless cost structure?
Michael Rollins at Citi: How are you changing your bundling message to reflect your wireless offering? Is Comcast experiencing difficulties in retailing wireless? Could a media + wireless bundle drive more subscriber growth?
Of the ten questions in the Q&A, five (at least partially) dealt with the MVNO business. Putting this into perspective, wireless represents slightly more than 2% of the 3Q 2019 Cable segment revenues and just over 1% of total 3Q 2019 corporate revenues. Even on a growth basis, wireless was $90 million out of $561 million growth from 3Q 2018 to 3Q 2019 (16%) and would be an even smaller number if we excluded the advertising revenue drop. Why so much interest in wireless?
Part of the answer could be the natural inclination to focus on those areas of the business that are dragging down EBITDA. Xfinity Wireless lost $94 million in EBITDA in the quarter or about $18.50/ month/ average subscriber. This figure is substantially better than the $178 million lost in Q3 2018 ($66.29/ month/ average subscriber) but largely unchanged from the $88 million lost the previous quarter ($19.60/ month/ average subscriber). Upticks are hard to stomach even if they are explainable, and it’s likely that the iPhone 11 launch impacted 3Q EBITDA.
Another answer is to look at wireless growth in light of the large High Speed Internet base. Comcast had 1.689 million average wireless subscribers in Q3 against 25.811 million average High Speed Internet subscribers. Assuming 1.8 Xfinity wireless lines per household (a figure below 2.0 assumes that By the Gig is more popular with individual/ single line users than other family-focused plans), 1.689 million subscribers would translate into roughly 940,000 households or below 3.7% penetration of current Xfinity household accounts (note that as the 1.8 lines per household grows, the penetration level shrinks). That’s an underwhelming figure given the 2.5 years Comcast has been actively marketing retail wireless services.
Nearby is the updated net additions growth chart for both Comcast and Charter. It’s very interesting to note that since 4Q 2017, net additions have been running in a very tight range. In fact, the four quarter net additions rolling average ranges from a low of 196 to a high of 214 – a close-fitting cluster. It appears that Comcast is being more deliberate in their growth strategy (in effect placing a 200K quarterly growth governor) in anticipation of additional events.
We have no doubts that Comcast has a long-term wireless strategy, and that it involves increased licensed and unlicensed spectrum ownership and operation at some point. But we also understand that every piece of content added to the package hurts gross margin and keeping up with AT&T/ HBO and Verizon/ Disney is going to be difficult without offload.
Bottom line: Comcast blew away High Speed Internet performance expectations which changed analyst focus to wireless, specifically unlimited plan profitability. Absent the collapse of the Sprint/ T-Mobile merger, there’s a lot of planning ahead.
Charter: Like Comcast, But with Less Content and no Europe Exposure
Charter also posted very strong growth with 351,000 net High Speed Internet additions and 276,000 wireless subscriber net additions. This translated into $4.1 billion in EBITDA which included $145 million in total mobile EBITDA losses (across 656,000 average monthly subscribers, this equates to a loss of $74/ average customer/ month which is better than where Comcast was after their first five quarters of service). Charter has grown slightly faster in their first five quarters of wireless service than Comcast did, and CEO Tom Rutledge believes that their sales productivity is just getting started.
In the Question and Answer section of their earnings call, Charter reiterated their increased growth trajectory on both wireless and broadband, and also reiterated that they are looking at CBRS across a wide variety of fronts (rural wireless expansion, highly congested areas, etc.). Charter did not mention any changes in their wireless strategy with respect to business, but it’s likely that small business expansion (< 20 lines) will continue to grow in 2020.
Bottom line: The tone of the Charter call was completely different than Comcast, reflecting the differences between the companies. Content discussions focused on retransmission agreement progress (with an acknowledgement that customers will see some increased costs), and there were absolutely no international discussions. It’s clear that Charter wants to a) hit the 1 million subscriber milestone by the end of the year, and 2) continue to realize scale efficiencies in wireless.
Next week, we will incorporate AT&T’s Time Warner Cable analyst day, AT&T quarterly earnings, and T-Mobile earnings into the discussion. Until then, if you have friends who would like to be on the email distribution, please have them send an email to email@example.com and we will include them on the list.
Greetings from the Queen City, where the IT scene is red hot even though cooler fall temperatures have finally arrived. I was pleased to be the guest of San Mateo-based Aryaka Networks at the 2019 Orbie (CIO of the Year) awards on Friday. It was great to catch up with many folks in attendance including Karen Freitag (pictured), a Sprint Wholesale alum and the Chief Revenue Officer at Aryaka.
This week, we will dive into drivers of wireline earnings. At the end of this week’s TSB, we will comment on several previous briefs (including the AT&T Elliott Memo fallout) in a new standing section called “TSB Follow Ups.” We close this week’s TSB with a special opportunity for reader participation.
Wireline Earnings: Enterprise, Expense Management, and Extinction
One of my favorite things to write about in the TSB is wireline – that forgotten side of telecom and infrastructure that serves as the foundation for nearly all wireless services. Wireline is a case study in competition, regulation, cannibalization, innovation, and a few other “-tions” that you can fill in as we explore the following dynamics:
Residential broadband market share (measured by net additions). Before the Sunday Brief went off the air in June 2016, cable was taking more than 100% share of net additions. This means that customers were leaving incumbent telco DSL (and possibly FiOS) faster than new customers were signing up. At the end of 2018, cable continued its dominance with 2.9 million net adds compared to 400 thousand net losses for telcos (see nearby chart. Source is Leichtman Research – their news release is here). If this trend holds through the end of the month, it will mark 18 straight quarters where cable has accounted for more than 95% of net additions (Source: MoffettNathanson research).
To be fair to the telcos, all of 2018’s losses can be attributed to two carriers: CenturyLink and Frontier. We have been through the Frontier debacle twice in the last three months and will not retrace our steps in this week’s TSB (read up on it here).
But CenturyLink is a different story, with losses coming in areas like Phoenix (where Cox is lower priced), Las Vegas (Cox lower priced except for 1Gbps tier), and legacy US West areas like Denver/ Minneapolis/ Seattle/ Portland (Comcast has lower promotional pricing). Even as new movers are considering traditional SVOD alternatives like Roku and AppleTV in droves, there’s a perception that the new CenturyLink fiber product is not worth the extra cost.
A good example of the perception vs reality dichotomy comes from the latest J.D. Power rankings for the South Region:
While these ratings reflect overall satisfaction with the Internet service, it’s very hard for new products to overcome old product overhang (and DSL experiences can create long memories).
But superior customer satisfaction (749 is a decent score for telecom or wireless providers regardless of product) does not guarantee market share gains. AT&T (Bell South) has continued to improve its fiber footprint, invested heavily in retail presence, and improved the (self-install) service delivery experience. Even with that, it’s highly likely that AT&T’s South region lost market share to Comcast (2nd place) and Spectrum (4th place). Why is a three-circle product outperforming a five-circle product?
The answer lies in several factors: Value (see comments above about promotional pricing and go to www.broadband.now for additional information), Bundled products (which links back to value – bundle cost may be significantly cheaper), and Legacy perceptions (DSL overhang mentioned above, tech support overhang, install overhang).
For more details, let’s look at two very fast-growing areas: Dallas, TX and Hollywood (Miami), FL. Nearby is a chart showing online promotional pricing for AT&T, Spectrum (Charter) and Comcast. There are some differences on contract term (AT&T has contracts in Dallas; Spectrum does not. Comcast has a 2-yr term with early termination fees to get the $80/ mo. rate for their triple play in Miami). Both zip codes selected above have over 50% served by AT&T fiber. AT&T is more competitively priced than Spectrum in Dallas, and extremely competitive with Comcast especially at the mid-tier Internet only level (promotional rate gigabit speeds are $70/ month with no data caps).
With superior overall customer satisfaction and competitive pricing, why does AT&T continue to tread water on broadband and lose TV customers? Are cable companies out-marketing Ma Bell? Is there a previous AT&T experience overhang? Are AT&T retail stores creating differentiation for AT&T Fiber (compared to minimal showcase store presence for Spectrum or Comcast)?
Bottom line: Cable will still win a majority of net adds despite lower customer satisfaction and higher prices. Why AT&T cannot beat cable especially in new home (AT&T fiber) construction areas is a function of marketing, operations and brand mismanagement.
Enterprise spending – Did it return to cable instead of AT&T/ Verizon/CenturyLink? We commented last week on AT&T’s expected gains in wireless enterprise spending thanks to the FirstNet deal. How that translates into wireline gains is an entirely different story. Here’s the AT&T Business Wireline picture through 2Q 2019:
While these trends are not as robust as wireless and operating income includes a $150 million intellectual property settlement, AT&T management described Business Wireline operating metrics as “the best they have seen in years.” What this likely indicates is that AT&T’s legacy voice and data service revenue losses (high margin) are beginning to decelerate (at 14.6% annual decline, that’s saying a lot – Q1 2019 decline was 19.2% and the 2Q 2017 to 2Q 2018 decline was 22.0%!).
Meanwhile, Comcast Business grew 2Q 2019 revenues by 9.8% year over year and is now running an $8 billion run rate (still a fraction of AT&T Business Wireline’s $26.5 billion run rate but a significant change from Comcast’s run rate in 2Q 2016 of $5.4 billion). Spectrum Business is also seeing good annualized growth of 4.7% and achieved a $6.5 billion annualized revenue run rate. Altice Business grew 6.5% and is now over a $1.4 billion annualized revenue run rate. Including Cox, Mediacom, CableOne and others, it’s safe to say that cable’s small and medium business run rate is close to $13 billion (assuming 33% of total business revenues come from enterprise or wholesale). That leaves a consolidated enterprise and wholesale revenue stream of ~ $5.5 billion which is more than twice Zayo’s current ARR.
The business services divisions of cable companies are repeating the success of their residential brethren. They are aggressively pricing business services, using their programming scale to grab triple play products in selected segments such as food and beverage establishments and retail/ professional offices. And, as Tom Rutledge indicated in last week’s Bank of America Communacopia conference, they are starting to sign up small business customers for wireless as well. I would not want to be selling for Frontier, CenturyLink or Windstream in an environment where cable had favorable wireless pricing and the ability to use growing cash flows to build a competitive overlay network.
Enterprise and wholesale gains are important for several reasons. In major metropolitan areas, segment expansion gets cable out of the first floor (think in-building deli or coffee shop) and on to the 21st floor. To be able to get there, cable needed to have a more robust offering. Comcast bought Cincinnati-based Contingent services in 2015, and Spectrum also improved its large business offerings. They are not fully ready to go toe-to-toe with Verizon and AT&T yet, but with some help from the new T-Mobile (all kidding and previous John Legere lambasting aside, a new T-Mobile + cable business JV would make perfect sense), things could get very difficult for the incumbents.
Moving up in the building is important, but there’s another reason to expand from the coffee shop: CBRS (if you are new to TSB, the link to the “Share and Share Alike” column is here). Given that 2-3 Gigabytes/ subscriber of licensed spectrum (non-Wi-Fi) capacity are consumed within commercial offices per month, there’s a ready case for MVNO cost savings as Comcast, Altice, and Spectrum Mobile continue to grow their wireless subscriber bases.
Further, since many enterprises are going to be introduced to LTE Private Networks soon, there’s a threat that Verizon and AT&T (and Sprint if the T-Mobile merger closes) stop provisioning cable last mile access out of their regions and only provision wireless access. As we have discussed in this column previously, the single greatest benefit of 5G LTE networks is the ability to control the service equation on an end-to-end basis for branch/franchise locations. It represents a compelling reason to move to SD-WAN, and allows cable to deepen its fiber reach and build more CBRS (and future spectrum) coverage.
Bottom line: Cable continues to grab share in small, medium, and enterprise business segments as they move from connecting to the building to wirelessly enabling each building. CBRS presents a very good opportunity to do that. Even with cable’s entre into the enterprise segment, it will still be dominated by AT&T (with Microsoft and IBM as partners) and Verizon for years to come.
Expense Management and Productivity Improvement. Flat to slowly declining operating costs in an environment where revenues are declining more precipitously is a recipe for increased losses. Even with some of the capital and operating expense being shared with 5G/ One Fiber initiatives, the reality is that lower market share is leading to diseconomies of scale and both are going up a cost curve right now.
That’s why reducing operating expenses is not a spreadsheet exercise – operating in a territory originally engineered for 80-90% market share that is now at 30-40% share requires increased efficiency. Connecting to neighborhoods is hard and connecting through neighborhoods to individual homes is even harder. Combine this with a change in technology (fiber vs twisted copper) ratchets the degree of difficulty ever higher.
One of the great opportunities for all communications providers is using increased computing (big data) capabilities to quickly troubleshoot issues and recommend remedies. For example, customers who go online or contact care and are “day of install” should have a different customer service page than someone who has been a 4-month regular paying customer or a 2+ year customer who is shopping around.
There’s no doubt that the online environment has been improved for every telco, and also no doubt that many more issues in the local service environment require physical inspection and troubleshooting. But when telcos move to correctly predicting customer needs through online help 95+% of the time, the call center agent will go the way of the bank teller and the gas pumper: Convenience and correct diagnosis will trump in-person service.
For those of you who are regular followers of TSB and read last week’s column, there’s also the issue of territory dispersion. Without retreading the information discussed last week, one has to ask if there are trades to be made in the telco world (or spinoffs) that make sense to do immediately (Wilmington, North Carolina, a legacy Bell South and current AT&T property would be a good example using last week’s map).
Bottom line: There won’t be any dramatic changes to the wireline trends – yet. But, as 5G connectivity replaces cable modems and legacy DSL (particularly to branch locations) and as cable expands its fiber footprint to include in-building and near-building wireless solutions (starting with CBRS), the landscape will change. And there will be a lot of stranded line extensions if wireless efforts are successful.
Randall Stephenson met with Elliott Management this week, according to the Wall Street Journal. At an analyst conference prior to the meeting, Stephenson offered somewhat of a hat tip to Elliott Management, saying “These are smart guys.” The AT&T CEO also stuck by his decision to move John Stankey into the COO role, noting “if you’re going to go find somebody who can do both, right, take a media company that has transitioned to a digital distribution company and pairing it with the distribution of a major communication company, and you want to try to bring these two closer and closer together and monetize the advertising revenues, all of a sudden, that list gets really, really short.” If Elliott’s decision to go public with its criticism is based on the Stankey announcement, I wonder how that logic was received in New York last Tuesday.
Apple iOS 13 fails again, this time failing to display the “Verified Caller” STIR/SHAKEN (robocall identifier standards) on Apple devices until after the called party has answered. Kind of defeats the point, right? More in this short but sweet article from Chaim Gartenberg at The Verge – we agree with the T-Mobile quote in the article “I sure hope they get this fixed soon.” Don’t hold your breath, as Apple is running on its 12th year of not allowing developers to access the incoming phone number.
The Light Reading folks have a very good chronicle of what’s going on with CBRS trialshere. Sharing can work, but it takes a lot to do it. The value has to be clearly present to increase carrier attention and participation.
Eutelsat can’t seem to make up its mind. On September 3, they dropped out of the C-Band Consortium (Bloomberg article here) and last week they seemed to backpedal based on this FCC memo. Time for Commissioner Pai to save the day!
Next week, we will cover some additional earnings drivers. Until then, if you have friends who would like to be on the email distribution, please have them send an email to firstname.lastname@example.org and we will include them on the list.
One last request – we are currently on the hunt for some of your favorite titles that chronicle telecom/ tech history. No title is off limits. Currently, we have three that have made the cut:
We’ll talk more about it in the week’s Sunday Brief, but the first quarter was very good to the wireless and cable providers. Not as good to the remaining wireline telcos, thanks to the dividend cut and corresponding stock price hit that CenturyLink took in February.
The comments are on the slide so I will not repeat them. But, when we announced in August 2012 that we were entering the “golden era” of value creation for wireless, we weren’t kidding. They have a lot of ground to make up (2009-2011 were very good to Google, Amazon, Microsoft, and Apple) but the wireless and cable sector seems to be on their way to creating more value in 2013 and perhaps erasing all of 2012’s losses.