Breaking up AT&T?

Craig Moffett of MoffetttNathanson was on CNBC recently and said that he eventually expected AT&T to be broken up. This caught my eye since I was at Southwestern Bell and have strong memories of the first divestiture of AT&T into separate companies.

The AT&T divestiture happened on January 1, 1984, when the AT&T parent long-distance company was separated from seven ‘baby Bells’ that operated the last-mile networks. That divestiture was government ordered by judge Harold H. Greene and was aimed to break up the monopoly that AT&T held in the long-distance market. The divestiture had the desired effect and long-distance rates tumbled in ensuing years to the point that people today make unlimited long distance calls on a cellphone without giving it a second thought. But pre-divestiture, long-distance rates average more than 10 cents per minute and were a major hindrance to interstate commerce.

In proof that you can’t hold a monopoly down, AT&T gathered back a lot of the pieces from divestiture over time, acquiring BellSouth, Ameritech, Southwestern Bell, and Pacific Telesis – four of the seven Baby Bells. However, it is not those acquisitions that led Craig Moffett to predict the second breakup of AT&T. Since Divestiture AT&T, along with Verizon has come to dominate the cellular business. AT&T did that through organic growth, but also by acquiring Cellular One and Leap Wireless. In 2011 AT&T tried unsuccessfully to merge with T-Mobile. The company has also bought into the satellite business with the acquisition of DirecTV. AT&T expanded its fiber network through the acquisition of Centennial. The biggest new venture for the company was the acquisition of the content creator Time Warner, that included Turner Broadcasting and Warner Bros. Studios.

Moffett believes that big diversified companies don’t do as well in the market as would the individual component companies. He also believes that at some point that Comcast will separate from NBC/Universal. When companies like AT&T and Comcast acquired content providers, both touted the huge benefits that could come from being both a content provider and a content purchaser. But those cross-benefits have never materialized to the extent envisioned by the original purchase.

Moffett said that at some point somebody will force AT&T to split off valuable assets like Warner Communications. He believes that likely will be the result of pressure from Wall Street and investors, but it’s also possible that a breakup could be urged by the government. Several Democratic candidates have mentioned breaking up the big telecoms as part of their platform.

Watching big companies over time is a big preoccupation in the telecom industry. Everybody in the industry watches the big corporations buy and sell companies like moving chess pieces. The acquisitions are always big news, but the impact in the industry mostly involves the employees affected by the mergers and acquisitions and not the rest of us.

Companies like AT&T seem to have little choice but to grow through acquisition. Wall Street drives companies to continue to increase earnings and it was a big shock for AT&T management when they were dropped from the Dow Industrial average in 2015. AT&T’s core business is still telecom. The company ignored the landline business for a long time, and you could barely find mention of the business in their annual reports a decade ago. The company has lately been touting the advantages or expanding its last-mile fiber network. For many years AT&T thought of itself as a cellular company, but that business has gotten more competitive and prices have dropped. Much of the company’s core business now earns infrastructure level returns – which are nice and steady and spinoff cash, but which don’t create the kind of returns that stockholders want to see.

Moffett’s prediction reminds us that the stories of the big corporations are never finished. They merge and buy companies to grow, and over time split or retract when they get too large. Big companies reinvent themselves when their industries undergo big changes. It wouldn’t be surprising to see AT&T repeat this cycle several times over the rest of this century.

Verizon Announces Residential 5G Roll-out

Verizon recently announced that it will be rolling out residential 5G wireless in as many as five cities in 2018, with Sacramento being the first market. Matt Ellis, Verizon’s CFO says that the company is planning on targeting 30 million homes with the new technology. The company launched fixed wireless trials in eleven cities this year. The trials delivered broadband wirelessly to antennas mounted in windows. Ellis says that the trials using millimeter wave spectrum went better than expected. He says the technology can achieve gigabit speeds over distances as great as 2,000 feet. He also says the company has had some success in delivering broadband without a true line-of-sight.

The most visible analyst covering this market is Craig Moffett of Moffett-Nathanson. He calls Verizon’s announcement ‘rather squishy’ and notes that there are no discussions about broadband speeds, products to be offered or pricing. Verizon has said that they would not deliver traditional video over these connections, but would use over-the-top video. There have been no additional product descriptions beyond that.

This announcement raises a lot of other questions. First is the technology used. As I look around at the various wireless vendors I don’t see any equipment on the market that comes close to doing what Verizon claims. Most of the vendors are talking about having beta gear in perhaps 2019, and even then, vendors are not promising affordable delivery to single family homes. For Verizon to deliver what it’s announced obviously means that they have developed equipment themselves, or quietly partnered on a proprietary basis with one of the major vendors. But there is no other ISP talking about this kind of deployment next year and so the question is if Verizon really has that big of a lead over the rest of the industry.

The other big question is delivery distance. The quoted 2,000 feet distance is hard to buy with this spectrum and that is likely the distance that has been achieved in a test in perfect conditions. What everybody wants to understand is the realistic distance to be used in deployments in normal residential neighborhoods with the trees and many other impediments.

Perhaps the most perplexing question is how much this is going to cost and how Verizon is going to pay for it. The company recently told investors that it does not see capital expenditures increasing in the next few years and may even see a slight decline. That does not jive with what sounds like a major and costly customer expansion.

Verizon said they chose Sacramento because the City has shown a willingness to make light and utility poles available for the technology. But how many other cities are going to be this willing (assuming that Sacramento really will allow this)? It’s going to require a lot of pole attachments to cover 30 million homes.

But even in Sacramento one has to wonder where Verizon is going to get the fiber needed to support this kind of network? It seems unlikely that the three incumbent providers – Comcast, Frontier and Consolidated Communications – are going to supply fiber to assist Verizon to compete with them. Since Sacramento is not in the Verizon service footprint the company would have to go through the time-consuming process needed to build fiber on their own – a process that the whole industry is claiming is causing major delays in fiber deployment. One only has to look at the issues encountered recently by Google Fiber to see how badly incumbent providers can muck up the pole attachment process.

One possibility comes to mind, and perhaps Verizon is only going to deploy the technology in the neighborhoods where it already has fiber-fed cellular towers. That would be a cherry-picking strategy that is similar to the way that AT&T is deploying fiber-to-the-premise. AT&T seems to only be building where they already have a fiber network nearby that can make a build affordable. While Verizon has a lot of cell sites, it’s hard to envision that a cherry-picking strategy would gain access to 30 million homes. Cherry-picking like this would also make for difficult marketing since the network would be deployed in small non-contiguous pockets.

So perhaps what we will see in 2018 is a modest expansion of this year’s trials rather than a rapid expansion of Verizon’s wireless technology. But I’m only guessing, as is everybody else other than Verizon.

The Valuation of a Cable Customer

SANYO DIGITAL CAMERA

Craig Moffett of MoffettNathanson recently set a valuation of an OTT customer from Sling TV at a quarter of the level of a normal Dish Networks customer. Since almost every small cable provider in the industry is interested in their valuation, I thought I’d talk today about Moffett’s numbers and how they might relate to cable valuation for small cable operators.

First the numbers. Moffett said that a normal Dish Networks cable customer is worth $1,100. That valuation reflects both the operating margin on Dish’s cable business as well as the average expected time that a cable customer stays with the company. Valuation in the industry in general is based on a multiple of operating margin – revenues less operating expenses. I don’t know what Moffett used as a multiple in this case since the valuation of Dish is muddled by the fact that they also own a mountain of spectrum.

Moffett set the value of a Sling TV customer (also operated by Dish Networks) at only $274. This low valuation tells us several things. First, the margins on Sling TV has to be significantly less. The company is obviously setting a low price to attract customers. And while Sling TV has a much smaller channel line-up than the big bundles at Dish Networks, Sling TV includes a lot of the most popular (and expensive) channels such as ESPN and Disney. I would also think that the valuation reflects a much higher churn for Sling TV. Customers are free to come and go easily and can buy service one month at a time. This contrasts to many Dish customers who get low prices by signing up for 1-year or longer contracts.

There are also other cost characteristics that are different for a satellite customer compared to on online customer. For instance, for a satellite customer Dish has to cover the cost of the satellite networks, the cost of the receivers used by customers. Sling TV has to instead just pay for transport of programming through Internet. Both parts of the business have to cover advertising and the cost of billing and back office. But it seems like Sling TV would have lower costs since customers must prepay by credit card. It’s hard to know which has a cost advantage, but I would guess it’s Sling TV. But Dish has millions of customers and would have some significant economy of scale.  

How do these valuations compare to the valuations of small cable providers? The big difference between terrestrial cable providers and Dish is having to provide a fleet of technicians in trucks and maintaining a landline network of some sort. Small cable operators also have to operate a headend and always face upgrades to keep up with the latest innovations in the industry. These costs are far more costly per customer for a small cable operator than what Dish is paying. I would think that due to economy of scale that Dish also has an advantage on costs like customer service, billing, etc. The equipment costs for customers are probably similar for Dish and terrestrial cable operators.

I have analyzed the books of a number of small triple play providers in recent years and if costs are allocated properly to products I haven’t seen one that has a positive margin on the cable TV product. While small cable systems generally charge more than Dish Networks they also pay more for programming. But the main reason that small terrestrial cable operators lose money is the work load associated with supporting cable TV. I’ve done detailed time studies at clients and have seen that in a triple play company that way more than half of the calls to customer service and the truck rolls are due to cable issues. If a small company allocates expenses properly between products, then cable is almost guaranteed to be a loser.

What does that mean for valuation? It’s probably obvious that if one of the major product lines of a company is losing money that the negative earnings pulls down the overall valuation of the business. Said more plainly, if the cable business at a small company is losing money, then that part of the business has no value or even a negative value. This is a conversation I have with clients all of the time, and most small cable providers have at least thought about the ramifications of dropping their cable product.

It’s not quite as easy as it sounds, because if somebody drops cable then they need to also pare expenses that were used to support cable. For a small company that means cutting back on customer service and field technician positions – something that small companies are loathe to do. Small carriers also worry that cutting cable will cost them overall customers, particularly if they are competing against somebody else that offers the triple play. It’s definitely a tough decision, but I’ve heard that as many as fifty small telcos have ditched traditional cable.

I’m also seeing for the first time that many new network operators are launching new markets without cable TV. Or they are instead looking at models where some external vendor like Skitter TV sells cable to customers.

Unfortunately, the cost of programming is still climbing fast and the margins on cable keep worsening for small cable operators. I expect that some time within the next five years or so we will reach a flash point where the collective wisdom of the industry will say that it’s time to ditch cable – and at that point we might see a flood of small companies exiting the business. But I don’t know of a harder decision to make for a small triple play provider.

Expect Higher Data Prices

Numismatics_and_Notaphily_iconThe US already has some of the highest Internet prices among developed nation. This is due largely to the lack of competition in most markets, meaning that there is no downward price pressure. Brian Fung of the Washington Post reported on a recent congressional hearing where Craig Moffett, a well-known market analyst, said that prices for Internet access are likely to climb in the future.

Moffett’s reasoning is that cable companies, who have most of the data customers in the country, are losing cable customers (or are running out of the ability to continually raise cable rates), and so they are going to have little option but to raise data rates.

The large cable companies, who together control the majority of the data customers in the country, are mostly publicly traded companies (except for Cox) and they are very much driven by the need to have profits climb quarter over quarter, year over year.

For many years the revenues and the profits of the large cable companies have been driven by two phenomena—the quickly growing data market and continual large cable rate increases. While data customer penetration rates are still growing, the rate of growth has slowed down and the vast majority of homes that want and can afford high speed Internet access already have it. And so the cable companies are no longer going to see the steady boosts to their bottom line that comes from double digit growth in very high margin data customers.

The cable companies have also been living off cable rate increases. They loudly blame cable rate increases on increases in programming costs. But the truth is that they have almost always raised cable rates more than what was needed to just cover higher programming costs, and so each rate increase added to the margins from cable and went straight to the bottom line.

But we are now seeing what I call consumer rate fatigue with cable rates. As cable keeps getting more expensive we are going to see more cord cutters, and even more cord shavers. Cable rates have climbed to the point where the cable companies should now hesitate when thinking of raising rates more than needed to cover cost increases. One only has to do the math to see that raising cable rates only 7% per year will increase an $80 monthly bill to over $100 per month in only four short years. The math is finally catching up to the cable industry, and at a time when there are finally online alternatives appearing for content.

Without those two historic bottom line drivers the cable companies are left with having to raise data rates if they want to grow profits to meet investor expectations. The cable companies are all dabbling with new revenue streams such as home power management, security systems, WiFi phones, and other new products. But these new products don’t have the same kinds of high margins as Internet data, nor is it likely that cable companies will sell enough of these new products to make a real bottom line difference.

The industry has been somewhat spoiled for the last couple decades due to having the powerful triple play bundle of voice, video, and data. While the margins on video aren’t great, the other two products have extremely high margins. The bundles have allowed the cable companies to have relatively high penetration rates of all three services. But nobody expects the new products to do nearly as well as the triple play services. Rather than having a few products with very high penetration rates, the cable companies are likely to end up with a product portfolio containing numerous products, each with a relatively small 5–10% penetration. That is going to make them into very different companies than today.

And so expect to start seeing data rates raised every year. This has already begun with the base rates for many cable companies and the large telcos like Verizon. I would expect the rate increases to be small at first but to climb over time to feed the bottom line expectations.

This is all counterintuitive. Most of my clients have 70–80% margins on Internet service today and it’s probably higher than that for the large cable companies. It takes some chutzpah to raise the rates on a product that is already that profitable. But I completely agree with Moffett and I think this is inevitable that data prices will rise, considering the lack of competition in most of our markets.