Tegus Sponsored Deep Dive #7: Revisiting cable (part #2: FWA)

This is part #2 in my tegus sponsored deep dive revisiting the cable sector. You can find part #1, which lays out why I’m revisiting cable now, here. You can also find the three tegus interviews I did that formed the basis for this series here:

• Former executive of Century Link / Lumen

• Former Executive at Charter Communications

• Former Employee at T-Mobile

There are plenty of threats to cable and the broadband markets. Regulation, fiber to the home (FTTH) overbuilding, potential future competition from satellite internet (like SpaceX), etc. However, there is one threat that looms particularly large, and that’s fixed wireless (AKA fixed wireless access, or FWA). FWA is the one technology I see that has the potential to completely disrupt the cable business model (in last year’s deep dive, I noted “at its most extreme, fixed wireless is the uber long term bear case for cable,” and I think that holds true today), so in today’s post I want to dive deep into FWA and why it’s such a threat.

Heading into 2022, cable was pretty dismissive of fixed wireless. For example, if I go back to CHTR’s Q3’21 earnings call, they were asked about competition from fixed wireless and responded, “we're seeing that the competitive environment doesn't appear to be significantly different than it has been.”

That changed over the course of 2022. T-Mobile’s FWA business saw accelerating growth; adds increased ~6x from Q2’21 to Q2’22!

And you saw cable companies (begrudgingly) admit that FWA was impacting them (though they still suggest the impact is limited to the lower end of the market) and analysts admit “that everyone probably underestimated fixed wireless coming into the market and the impact it would have.” And that impact definitely shows up in the numbers; in 2022, Comcast reported broadband subscriber losses in a quarter for the first time ever (table below from Comcast’s FY22 trending schedule; note the massive fall off in subs from 2021 (the left most numbers) to 2022 (the right most)), and total broadband sub adds were ~20% what they came in 2021.

Comcast wasn’t alone in reporting sub drops; basically every other cable company went negative on adds by the back half of 2022. And adds won’t be getting better in the very near term; at a recent conference Comcast noted “there's a number of analysts that have us losing subscribers in the first quarter. And as I sit here right now, it's probably the right assumption.”

~18 months ago, it was an open question if FWA could take real market share in broadband. No one doubted FWA had a place in some niches / edge cases (super rural places where building fiber didn’t make a lot of sense, users (like the quite elderly) who had almost no data, speed, or reliability needs, or maybe mobile food trucks that didn’t need much data but would move constantly), but many (including myself and the cable companies) thought FWA’s limitations meant that it wouldn’t work for ~95% of the population and it would not be able to take serious share.

That debate has been put to rest. TMUS (and VZ) have proven that fixed wireless can compete with and take share from cable companies.

The debate going forward is now how competitive fixed wireless can be and how much share it can take. This is the most important question for cable / telecom going forward.

At it’s most extreme, fixed wireless is a complete cable killer. Think about a wireless company for a second. Adding an incremental customer (i.e. a new cell phone user) costs them basically nothing; if they add just one new user, it’s not like they need to go build out new towers or buy new spectrum. The customer just turns on their phone and they’re good to go.

The uber bear case for cable companies is right there. Wireless companies already have nationwide networks built out; if they can start adding fixed wireless customers at no incremental cost and no detriment to their networks like they do new cell phone customers, then fixed wireless has a massive edge on cable. This would be a completely converged wireless world; basically, in this world you would go to your cell phone provider and just get an “unlimited data everything plan” that would provide you both your cell phone service and your high speed home broadband. Because the wireless company provides both services over the same network and they upgrade / maintain the network together, the incremental cost of providing you home broadband is zero, and they eventually destroy the home broadband / cable market (who offers home broadband that you don’t need since you’re getting it for “free” with your wireless service). Wireless service will have replaced home broadband the same way it replaced home landline telephone service. Note that this model is basically how TMUS described all the fixed wireless customers they added in 2022:

It was just a great success again in 2022. 2 million of those 2.6 million customers came in 2022. And a little bit of a difference, I think, to what perhaps other players in the industry are doing, we said that with the merger and as we roll out this massive capacity on the network that you would have fallow capacity. And when you have fallow capacity that you've already burdened the mobile build with all of the CapEx to get this rollout this fabulous network, it makes a tremendous amount of sense to fill up some of this massive highway traffic that's unused with high-speed Internet. And for us, it became the first killer use case of 5G.

We're now we're riding about 2.6 million customers that have fabulous economics, again, because the network is already there, but it's follow capacity.

So that would be your ultimate doom argument for cable. However, that ultra doom case does not appear to be in play. The reason for that is simple physics: wireless networks are capacity constrained at some level, and they simply can’t handle every wireless consumer getting their broadband from fixed wireless. There’s actually no (current) disagreement on that capacity constraint point; you can read any transcript from a wireless company and see that they are aware they can’t take every single broadband consumer over fixed wireless (I really like this interview w/ TMUS’s president of technology; it’s quite balanced in discussing the opportunities and limitations of wireless capacity).

But just because the ultimate doom case can be dismissed doesn’t mean fixed wireless can’t do a ton of damage to cable companies. Remember, the wireless networks are already built out; adding an incremental user on to them is basically costless if the network has capacity. It doesn’t matter if that user is a fixed wireless user or a normal mobile user; if the network has extra capacity, throwing them on is going to be profitable.

When thinking about extra capacity and FWA, I think there are two questions to consider: one of economics, and one of sustainability.

The first questions on of economics. Let’s say a wireless company has some extra capacity but not unlimited amounts. Here’s a quote from FYBR in December; they note the top quartile of their households consumer 2 terabytes a month in data, and fixed wireless operators are reporting the average FWA consumer is using 300-400 gigs and growing.

If we just take our Consumer business, which targets average households, so let's say, 5 people in a household with maybe 7 to 10 cellular devices, 3 or 4 4K TVs, a couple of play stations, someone in office doing Zoom calls and all the other things that we all do in our houses, that generates an average usage for us of 1 terabyte a month, and at the upper quartile would be 2 terabytes a month. That simply isn't deliverable with FWA box. And in fact, of course, some of the mobile operators have shared that usage data was about 300- to 400-gig a month. And that's growing logarithmically

The average U.S. smartphone user takes ~31 gigs of data/month (that’s over both wifi and cell, but let’s just say it’s all on cell networks for our sake). Some quick math shows that FWA users consumer ~10x the data of a wireless user.

Wireless companies thus have a choice: if you have extra capacity, do you want to allocate that capacity to serving one fixed wireless customer or 10 wireless customers?

The answer to that generally depends on pricing…. if you were getting $1k/fixed wireless sub and $10 for a wireless customer, you’d prefer the one fixed wireless customer!

But that’s obviously not the case. Comcast is the largest cable company in America and charges <$70/month for broadband. AT&T’s ARPU for wireless is $55/month. The math of what to focus on is pretty obvious.

Again, that supposes there are some limitations on capacity. That tends to be the case in urban and suburban areas. But wireless companies tend to buy their spectrum in nationwide blocks, which means they have ~the same amount of spectrum in New York City as they do in rural America. Spectrum needs to have a lot of ecosystem around it to work (towers need to be built up, fiber needs to go to towers, etc)., but one thing you could see is wireless companies focus on running a “normal” wireless business in denser parts of the country while using fixed wireless to fill excess capacity in more rural / less dense parts of the country. This is basically the path AT&T is going on:

I feel like a little bit of a broken record. And we've been on record from the get-go that fixed wireless is not going to be a priority of ours. We're going to prioritize our investments in fiber, our investments in wireless, but we're not going to prioritize fixed wireless where it could make sense for us areas like some of the rural areas where we don't expect fiber to ever be deployed, and we don't believe it would interfere with our wireless services could make sense as a catch product in an area where we're trying to get fiber to. But outside of those 2 circumstances don't really think it's going to be a priority for us because once you start to factor in the tax on your network along with the customer acquisition cost, and the price point at which the product is being offered, you look up and you probably -- the returns on that product are just not that attractive.

That strategy could have interesting implications for cable companies. A few years ago, CABO traded for a large multiple premium to the rest of the cable companies. There were a few potential reasons for this, but one of them was that CABO’s largely rural meant CABO was much more insulated from FTTH upgrades than other cable players. Today, however, investors could look at CABO’s rural footprint and wonder if CABO will face more competition from FWA than their peers.

Of course, CABO would counter that they have plenty of experience with fixed wireless, and they don’t consider it a threat to their current business. On top of that, CABO would counter that the wireless offering in many of their markets is terrible; if phone companies can’t provide reliable cell service in those markets, how can they hope to provide FWA (which requires lots of capacity, low latency, and high reliability)?

The other question is of sustainability. Data usage continues to grow at unbelievable rates; CABO noted on their Q4 call that, “Our average data usage has grown at a CAGR of 26% over the past 5 years to nearly 640 gig bits per month in the fourth quarter of 2022, with downstream usage 15x higher than upstream.”

Many cable operators would argue that fixed wireless cannot keep up with the continued demand for data consumption. Sure, fixed wireless can work today for filling up some excess spectrum, but as data demands continue to rise wireless companies are not going to be able to keep up with the demand for faster speeds. In fact, as data usage grows, wireless companies might look at their limited capacity and the economics trade-offs discussed above and say, “hmmm….. maybe we should start dropping some of those fixed wireless users because we really need that spectrum/capacity to service our more valuable wireless users.”

Comcast put it pretty succinctly recently:

Fixed wireless is the newer competitor. They are clearly having their moment right now, adding close to 1 million subs per quarter. Maybe that degrades, maybe it doesn't in the near future, but I think there is a limited runway for that to go penetrate. So, taking a customer base and saying, "Who's going to suffer less reliability, higher latency, deprioritization on a network when wireless traffic takes over? And who's going to be willing to tolerate that?" We think that's probably a shrinking pool over time. And so, not to say it's not having its moment right now, but it's a real finite base you can sell into. It's disrupting us a little bit at the lower end of the market. But back to your prior question, we're going to go out of our way to protect the existing base, move the existing base forward and compete rationally in certain segments where we're seeing some disruption.

And here’s how CHTR put it:

Fixed wireless access, I can make the argument, I will, it's not anything really new. It's wireless substitution, but somebody has decided to market it in a cheaper way. And it's interesting, but it really is just wireless substitution. And so it will suffer from the exact same thing that's always been the case is a lack of capacity, by definition, an inferior product. And as data usage goes up, I think that's going to be pressured and customers will want to focus on having a quality product.

And on top of that, when you combine our broadband product, which is better, and our mobile product, which is the fastest, and you put it together, you still save money, whether that's at promotion or at retail. So yes, there's somebody new inside the marketplace, and they're marketing it in a different way, but it is wireless substitution. It won't ever be as good. It doesn't actually save customers money. And I think over time, they say the truth will prevail.

So the cable companies / fiber providers are betting that current FWA trends are not sustainable. And I think that’s a good bet. Basically, an investment in cable is a bet on two things

1. Consumer data usage will continue to grow rapidly

2. Wireless providers will not be able to grow capacity faster than consumer data usage.

I think both are good bets.

History suggests that capacity is an “if you build it, they will come” scenario. ~30 years ago people wouldn’t dream of needing more than 1 MB/sec from their interest (dial up was ~56k, so 1 MB would be 20x faster). Today no one could live on that speed. AR and VR are eventually coming, and they’re going to take huge amounts of data, and as more TV watching switches to streaming (and we continue to demand clearer pictures) and more pieces of our home get connected to the internet, data usage is only going to go up. Fixed wireless might work for some consumers today, but it wouldn’t work for that top quartile of consumers FYBR mentioned in the quote earlier… and where the top quartile of data users are today is where the public is in 3-5 years.

The other risk is that wireless providers somehow manage to grow their capacity faster than consumer data usage. If consumer data usage growth ~25%/year but wireless companies somehow figure out a way to grow their capacity by 50%/year, then they’ll be flush with capacity and can offer consumers unlimited everything.

I’m (obviously) not a network engineer, but I’m just not sure how wireless companies could magically grow their capacity like that. All of these companies have access to the same engineers, and I’ve yet to see a company suggest that fixed wireless can handle unlimited data and that a wireless network would never run into capacity constrains. I think this quote from TMUS discussing their fixed wireless product versus fiber is pretty telling (it’s also telling they continue to deploy fiber in select markets):

Great. Well, first on broadband, it's kind of stating the obvious. When somebody who is a fiber provider, so as you know that product is not as good as our product. It's kind of like the people at Ferrari pointing a finger at the world's best-selling car, Toyota saying, "We're faster. We have the faster car." Yes, but Toyota is the world's best-selling car, and that's because -- and if you look in the case of T-Mobile 5G home broadband, because it's perfectly suited to what people want.

And although it has less overall potential for capacity than a strand of fiber, which is patently obvious, it's radically simple. It's low cost. It's transparent. It's portable within tens of millions of households. And it has the speed and capacity that allows people to do what they want.

Again, TMUS is the leader in fixed wireless. They’re not out here saying that they’re going to compete to take the whole market. They’re noting their capacity is lower than fiber and that their fixed wireless solution is perfectly situated to grab people who aren’t as picky when it comes to speed / capacity / reliability. I think you can see that in the results too; TMUS’s FWA adds actually declined from Q3’22 to Q4’22 (from 578k to 524k). That’s still a blistering pace, but given TMUS’s marketing blitz, expanded availability, and a still new service that is yet to really start experiencing churn, that decline in adds suggests that the product is experience some type of constraint (either in capacity, so TMUS is controlling adds, or in demand, so the consumers who really wanted FWA have already switched and the market is starting to saturate).

It’s also important to remember that physics are physics, and every player has access to top tier engineering talent. So when you see AT&T refusing to commit to FWA except for the most rural locations, or VZ still leaning into fiber buildouts despite a successful FWA launch, or even the quote above from TMUS about fiber being a better product and the limitations of the FWA product, that should tell you a lot about where the wireless players see the future. Consider what CMCSA said at a recent conference:

You take all the work our engineers do to study this stuff up and down and as do others, and everybody obviously talks their own book to a degree. So you got to all sort through that for yourselves. But I think one of the things that's clearly the case is it taps out when you get to the kind of speeds we're talking about and the capacities we're talking about. It's a fine approach to market. It can work if you have lower needs. But how long are -- if the typical and average need of a consumer keeps going up, the person who might be satisfied today may not be satisfied tomorrow.

If you take one example, Thursday Night Football moved from network television to Amazon Prime, viewership dropped in half, yet that night became the most streamed -- the highest demand on our networks in the U.S. versus what had previously been Sunday Night. So you imagine where the world is going. It's going to continue to put stress and strain in such a way that I really don't think the kind of consumer that -- the average consumer is going to find that, that's going to work.

And as you all know, it's not the most economically efficient way for the wireless companies to use that spectrum. So the day will likely come where they have to reprioritize how they use their spectrum, and that's going to change the experience, plus it's a differentiated experience in one neighborhood to the next on one street -- house on the street versus another house on the street, that's all the testing that we've done. And those are some of the learnings that -- we want to stick to the strategy of playing the long game against a high demand for high-quality, best-in-class, lots of latency, lots of durability, lots of reliability, lots of speed, lots of capacity. And that's what our road map is in broadband. And to devalue that by -- we will, with selective offers and what's right for different segments, answer some of the competitive threat, but we're not going to lose sight of the bigger prize over the long term.

So yes, fixed wireless has taken more share than I think all but the biggest wireless bulls thought they could. But it’s showing signs of slowing, and even without that I think there are big long term questions about its capacity and sustainability.

And there is one other thing to consider: the bear case for cable is fixed wireless has unlimited capacity…. but the difference between fixed wireless and WiFi is basically zero, and cable has fiber much, much deeper where consumers really want it than a typical wireless player. What if the real threat isn’t fixed wireless eating cable, but cable’s mobile business eating wireless?

I’ll talk about all of that and wrap this series up in part 3. I look forward to seeing you then.