Cable’s wireless ambitions…one step closer

We have been arguing for a couple of years that investors have the 5G threat backwards.  The opportunity for wireless carriers in the fixed broadband market is small and the threat to Cable often exaggerated.  However, the opportunity for Cable companies entering wireless is much greater, and it is largely ignored by investors as an opportunity for Cable or as a threat to the Wireless carriers (we touched on it in our tome on Cable’s wireless economics HERE (now a little dated), and in a past global weekly review HERE, and Blair has commented on it HERE).

We have focused on this theme less in the last year because wireless subscriber growth had been slower than we expected for Cable, and the thesis wasn’t likely to get much traction until they picked up the pace.  We even pulled the value of the wireless opportunity out of our price targets for Comcast (worth $5) and Charter (worth $98).  We had never included it in our valuation for Altice because they hadn’t announced a strategy when we were focused on the issue, but a successful Wireless MVNO would be worth at least $3 for them (we have a detailed wireless model for each of them; let us know if you need them).

We noticed something in the new iPhone that renewed our excitement.  It could pave the way for Cable to move beyond the MVNO, with a much better wireless product, and much better wireless economics.

The new iPhones can use CBRS spectrum…

This year’s crop of iPhone’s included functionality for CBRS spectrum.  CBRS encompasses 150MHz that sits between 3.55GHz and 3.7GHz (Band 48, in the parlance of the standards setting bodies).  By way of reminder, this is an odd band that is currently being used by Navy and Coastguard radars.  The Navy isn’t getting out of the band, but they have agreed to share the band on novel terms that could become a template for repurposing other bands of spectrum that are being used by the government.

The government users will have the first claim on the lower 100MHz whenever they need it.  70MHz of priority access licenses (PALs) will be auctioned in seven 10MHz blocks in June 2020.  The winners of these licenses will have priority access to the spectrum covered by their licenses whenever the government users aren’t using it.  The PAL licenses will be dynamically allocated in 10MHz slices across available frequencies between 3.55GHz and 3.65GHz.  Since the Navy only uses 10-20MHz at a time and intermittently in limited geographies, the PAL licensees should have access to their spectrum pretty much whenever they need it.

The remaining 80MHz will be unlicensed.  In addition, any of the 70MHz of licensed spectrum not being used by the PAL licensees at any point in time can be used by unlicensed users too.  The unlicensed spectrum will be made available for the first time next week.

The CBRS band will have power limits that restrict the reach of the spectrum – like WiFi, but better.  This makes it not so good for traditional macro site deployments, but it works well for densely packed small-cells[1]We use small-cell here as a catch-all that includes indoor and outdoor DAS systems, micro-cells, pico-cells, and access points in homes and offices..  The wireless carriers have limited small-cell deployments today, mostly in high-usage areas like stadiums, malls and transport hubs, with some outdoor deployments covering denser portions of big markets.  We estimate that this infrastructure may only cover 15% of their usage today, though it will grow as they continue to deploy small cells.

Because of the power limits and the shared access requirements, the wireless carriers haven’t been very enthusiastic about the CBRS band.  Among the four national carriers, only Verizon has been making a vocal push for the band to be repurposed.  Verizon’s focus isn’t surprising given that they are the most spectrum constrained of the carriers and would likely welcome capacity from any source.  In addition to Verizon, among the communications services companies we cover, Cable has been focused on CBRS[2]CBRS has support from the broader ecosystem.  The CBRS alliance includes everyone from Google to Federated Wireless, who will manage the spectrum allocation process, to most of the equipment OEMs, handset makers, service providers, chipset makers, and others (LINK)..

…and the CBRS band may be ideal for cable companies to deploy…

Cable companies already have very dense small-cell networks that more or less blanket the areas where they have customers.  They operate as WiFi access point networks today.  Every home or office that uses Cable broadband has one or more access points, and the industry has a couple million more outside of homes in public areas.  Taken together, the cable companies might have as much as 75 million access points that could be refashioned into 4G and 5G small-cells over time.  CBRS could work exceptionally well in an architecture that is as dense as Cable infrastructure would allow.

We expect the Cable companies to acquire PAL licenses when they are auctioned next year.  They won’t need much - maybe 20MHz.  Spectrum is shared by all users within reach of a cell site or access point.  Because Cable companies have so many access points, so densely packed, they “reuse” the spectrum for more than a typical carrier network would.  In addition, they will use their licenses in conjunction with the unlicensed portion of the band, giving them up to 150MHz at any point in time[3]A cable company with a 20MHz PAL license will be able to use their 20MHz in conjunction with the 80MHz that is dedicated to unlicensed users, giving them access to 100MHz channels.  The 80MHz of GAA spectrum will be shared with other unlicensed users, but because of the power limits and the functionality of the spectrum management system administered by Federated Wireless and others, many access points will have full use of the 100MHz channel much of the time.  In fact, a cable company deploying CBRS will also have access to the other PAL licenses when they are not being used, giving them access to up to 150MHz..

…it could transform their wireless offering…

Cable’s wireless customers should experience blazing fast 4G and 5G access whenever they are in reach of one of their CBRS access nodes.  When an access point has access to the full 150MHz, it should be able to deliver LTE speeds of hundreds of Mbps and eventually Gigabit 5G speeds.  This should be substantially faster than the Verizon network, which would still operate as the fallback network when the customer isn’t in reach of a Cable access point.

Cable companies had hoped to deliver a similar benefit with wide-area WiFi, but it never worked very well outside the home.  There were a few issues.  Phones connect to WiFi access points when there isn’t a strong enough signal to deliver data (access point overreach). The handoffs between WiFi and the carrier networks have never worked well, resulting in dropped calls and data sessions. And WiFi is often congested, with performance that is often worse than on the carrier’s LTE networks. Because of this, as wireless customers have migrated to unlimited plans and stopped worrying about managing their usage, they often switch off WiFi, particularly when they are away from their home network.

CBRS deployed with LTE or 5G will work better than WiFi for the Cable companies because they will primarily be using licensed spectrum, reserved for their use most of the time, with a standard that is either the same as or built to integrate seamlessly with whatever their MVNO carrier is using.

This all means that the Cable companies should be able to offer a better wireless product than they do today, and a better wireless product than their MVNO partner can offer.  To use Charter and Verizon as an example, Charter’s customers will have access to the Verizon network, with whatever spectrum is deployed in it, all of the time.  They will also have access to the Charter CBRS network when it is available.  Charter can set its protocols such that customers only use the CBRS network when it can deliver better performance than the underlying Verizon network.

Charter has been doing well with Verizon’s wireless offering, sold at a discount.  They ought to do better with a product that is better than Verizon’s sold at the same discount.  They may even be able to increase the magnitude of discount.

…it could also transform the economics of their wireless offering…

We estimate that Cable customers consume 5GB on Verizon’s network at a cost of $3 per month, on average. The cable companies could cut that by two-thirds by deploying their own network, boosting EBITDA per sub by ~$10 per month, on average. At 15% penetration of households, Comcast and Charter would generate close to $3BN in EBITDA from their wireless business. The improved EBITDA margins would boost the value of the wireless business considerably.  This is what gets us $5 in present value for a wireless business at Comcast and $98 at Charter.

The improvement isn’t quite as meaningful for Altice because they are starting with a lower cost per GB than their larger brethren.  They are also starting with much lower pricing ($20 for unlimited vs $45 for Comcast and Charter).  We estimate $6 in EBITDA per sub per month today (excluding fixed costs), increasing to $10 with the benefits of CBRS.   At 20% penetration (we assume higher terminal penetration for Altice given much lower pricing), Altice would generate $2BN in Wireless EBITDA and the business would have a present value of $3.

Deploying spectrum will serve another, more important purpose for Altice though. Their MVNO with Sprint, or T-Mobile if T-Mobile’s deal is approved, will expire in 4-7 years. They could face higher prices or lose access to an MVNO altogether when their current agreement expires. By deploying spectrum, Altice ensures that their wireless business has a terminal value.

We have assumed that Cable companies keep pricing flat as their economics improve, but they could very well price more aggressively and take more share, faster.  There would be a trade-off between higher penetration and lower pricing, so this may not impact the value we have ascribed to the wireless business at the cable companies, but it could seriously increase the damage that the Cable companies do to the existing national wireless carriers.

…and it likely won’t cost much to deploy

Cable’s WiFi Access points are cheap today: less than $100 for units used at home, and less than $1,000 for enterprise-grade units.  Adding CBRS spectrum might increase that by 10-15% today (LINK), and as unit costs fall over time the incremental cost should be negligible.  The cable companies swap out 10-20% of their access points every year, and the cost is captured as part of ongoing CPE capex. They may accelerate the upgrade of access points driving higher capex for a couple of years; however, the incremental capex would come with a quick and material payback.

We think CBRS licenses are likely to be cheap too. The auction will be on a timeline similar to that of the C-Band auction[4]The date for the CBRS PAL license auction has been set for June 2020.  We don’t have a date yet for the C-Band auction, but we should have an FCC order before the end of the year, and the spectrum is likely to be sold by year-end 2020., which we think will be hard fought, with all the national carriers plus Dish wanting the spectrum.  We would expect the national wireless carriers to concentrate their resources on the C-Band auction because it doesn’t have power limits and because the licensing regime mirrors what they are used to more closely.

It’s difficult to know what the price CBRS will fetch, given the uniqueness of the band, but at $0.25 / MHz-POP Comcast and Charter might spend around $1.5BN each (and $0.25 / MHz-POP may prove too high by a factor of 10).  If that seems like a lot, consider that Comcast routinely spends around $3BN a year on small acquisitions for which they are never held to account.  This investment would come with a quick and clear payback.  To put it in more basic terms, we believe Comcast, Charter and Altice could pick up the spectrum they need in mid-2020 without materially impacting what they have available to spend on share repurchases (and we probably have higher share repurchase estimates than you do).

The cable companies will need to get handsets capable of using CBRS into the hands of their customers in order for customers and the companies to see the benefits described above, but this is already underway.  There are 10 handsets currently in the market that are CBRS capable, now including the iPhone 11.  We expect most new devices to include the band going forward.  The process of seeding the base is already underway.  In addition to a CBRS radio, handsets will need dual-SIM / dual-standby in order to switch between a cable CBRS network and a carrier network seamlessly; this has also been included in iPhones since the iPhone 10 (it is also in the iPhone 11), as well as other prominent Android based devices.

Some things still need to be worked out

It took the Cable companies much longer to launch and ramp their MVNOs than we expected, in part because wireless is a more complex business than we (or they) realized.  The challenges of figuring out procurement, supply chains, insurance, retail distribution, billing, customer care and other aspects of the business all took time.  Deploying and operating a CBRS network will be more complex still, with its own set of lessons.  This will be compounded further by the complexity of integrating two networks – one owned and one leased – with dual-SIM capabilities embedded in handsets.

It will undoubtedly take time, and there will be bumps along the way.  But the direction of travel is clear, and we don’t think investors are paying close enough attention.  Sooner or later, the connectivity market will be a single, converged market with today’s wireless and wireline companies all competing for share.  The cable companies occupy a much smaller portion of that converged market today than the wireless companies.  They have a good deal more opportunity to attack, and a much smaller position to defend.  Moreover, we think they have a strong asset advantage and a time advantage over the other market participants.

Cable companies really might have the most robust 5G wireless networks in the next five years.  This is something we think very few investors take seriously.  In other words, if we are wrong, Cable doesn’t have much to lose.  If we are right, they have a good deal to gain, and it will mostly come at the expense of Verizon and AT&T.

Our updated valuation comp sheets can be found HERE.

1. We use small-cell here as a catch-all that includes indoor and outdoor DAS systems, micro-cells, pico-cells, and access points in homes and offices.
2. CBRS has support from the broader ecosystem.  The CBRS alliance includes everyone from Google to Federated Wireless, who will manage the spectrum allocation process, to most of the equipment OEMs, handset makers, service providers, chipset makers, and others (LINK).
3. A cable company with a 20MHz PAL license will be able to use their 20MHz in conjunction with the 80MHz that is dedicated to unlicensed users, giving them access to 100MHz channels.  The 80MHz of GAA spectrum will be shared with other unlicensed users, but because of the power limits and the functionality of the spectrum management system administered by Federated Wireless and others, many access points will have full use of the 100MHz channel much of the time.  In fact, a cable company deploying CBRS will also have access to the other PAL licenses when they are not being used, giving them access to up to 150MHz.
4. The date for the CBRS PAL license auction has been set for June 2020.  We don’t have a date yet for the C-Band auction, but we should have an FCC order before the end of the year, and the spectrum is likely to be sold by year-end 2020.