300th Episode with Craig Moffett on Convergence and Fiber Economics

Full Transcript

Don Kellogg 0m10s

Hello, and welcome to the three hundredth episode of the week with Roger, a conversation between analysts about all things telecom, media, and technology by Recon Analytics. I'm Don Kellogg, and with me as always is Roger Reitner. Roger, 300.

Roger Entner 0m23s

Happy three hundred, and I'm so excited for having a phenomenal guest.

Don Kellogg 0m28s

Yeah. So we mentioned last week on the pod that we'd have a fantastic guest, and we're pleased to welcome Craig Moffett onto the podcast. Craig is a cofounder, partner, and senior analyst at MoffettNathanson, and dare I say, a legend in the industry. So, Craig, welcome.

Guest Speaker 0m43s

Thank you very much. And, Roger, congratulations on 300. That really is a remarkable milestone.

Roger Entner 0m48s

Well, it's just tenacity. Right? We keep going. Right? And refuse to give up.

Roger Entner 0m54s

But thank you for for being on the on the show. This is awesome. I'm very happy to be here. So I think an awesome topic to talk about, you know, and you've done a lot of work on it. We've done a lot of work on it.

Roger Entner 1m8s

It's convergence and the cost of deploying fiber. Right?

Guest Speaker 1m13s

It's one of my favorite topics. It's actually where I probably cut my teeth in the industry longer ago than I would care to admit back when I was at Boston Consulting Group and working for a client who I guess the statute of limitations is over for saying but I can't name who my clients were. At the time, it wasn't even Verizon yet. It was Nynex and Bell Atlantic back in the 1990 early nineteen nineties, modeling the economics of what later became FIABs.

Roger Entner 1m39s

Awesome. I helped them in the early two thousands to roll it out and to think through that. So we're like the well, I don't think we're godfathers. We were like early contributors and fans of this.

Guest Speaker 1m54s

But somewhere, there is certainly an acknowledgment for the fact that we're both old.

Roger Entner 1m58s

That is true. Well, you became wise. I became foolish. Right? So when I look at this, there's like a continuum of technologies where they make sense.

Roger Entner 2m8s

There's fiber, there is FWA, cable floats in here. Right? And there's Starlink. And we won't talk about Starlink because we talked about this in excess. We'll have you back to talk about that.

Guest Speaker 2m23s

That sounds good.

Roger Entner 2m24s

Where does it make sense to build fiber, and where does it not make sense to build fiber?

Guest Speaker 2m29s

Well, Roger, it won't be a surprise that the most important factor, and this dates back, we knew this thirty years ago when we were modeling fiber, the most important factor is density. That, by the way, is a lesson that it seems like we're relearning over and over again the in bead program when you read that they're suddenly figuring out that, hey. Wait a second. Maybe it would have been better to have satellite serving all these super low density places. It's something that we probably, you know, should have been obvious thirty years ago, but we keep having to relearn it.

Guest Speaker 3m3s

I think the fiber industry, the overbuilding industry and by the way, when I refer to fiber overbuilding, I think of any operator that is building in a market that already has a high speed terrestrial network that includes cable. So if you're the first fiber builder in a market that already has cable, some people don't call that overbuilding. I do. So just

Roger Entner 3m27s

Yeah. No. No. Wherever there's somebody before. And by the way, cable is also a fiber builder.

Guest Speaker 3m32s

Absolutely. Cable is building you know, people forget. I think it was, what, 2005 when hurricane Katrina hit New Orleans. Yeah. So much of the cable system was destroyed now over twenty years ago.

Guest Speaker 3m46s

Cox rebuilt that system with fiber. So the cable industry has been rebuilding with fiber and building with fiber for at least twenty years now.

Roger Entner 3m55s

Yeah. And all the out of markets are fiber markets. What is it? Charter is like You

Guest Speaker 4m1s

know, on the edge out and the subsidized hard off markets and things like that are all fiber.

Roger Entner 4m7s

Yeah. Charter has like last number I saw was 11. Probably, we're now, like, 14% of Charter's customers are fiber customers.

Guest Speaker 4m17s

Yeah. They would have to be. Right? Because they've been growing their footprint at 2% plus per year, and all of that excess growth, if you figure household formation is under one percent, that's going on with HFC. That means more than 1% every year is being added to the footprint in fiber, and they've been doing it for ten years.

Guest Speaker 4m36s

So, yeah, it's gotta be in the double digits at this point.

Roger Entner 4m39s

And they're savaging the DSL markets where they're not the overbuilder because there's nobody else but DSL. And within a year, they tell us they have 40% share.

Guest Speaker 4m49s

Yeah. It makes perfect sense. But that gets back to this sort of this density question because I think the most important question is not the greenfield markets because, ironically, almost all the deployment in very rural markets is fiber these days, but that's largely because it's being subsidized to be that way. We can come back to what else is out there. It's not just FWA and satellite.

Guest Speaker 5m15s

It's also there's a big risk market that people forget over unlicensed spectrum that's probably deteriorating or disappearing really quickly now that better solutions for very rural customers are coming along. But the part of the market I think that's the most interesting is the overbuilding of cable that's happening in what is now the sort of shoulder market. So it's certainly not the case that all the early fiber building that was done or fiber overbuilding that was done was done in the densest markets, but it's probably not far from being the case. Right? It would have been crazy to have done anything else.

Guest Speaker 5m54s

So if you assume that the densest parts of the country have been already built and that we're moving to progressively less and less dense markets as the good stuff has been picked over, what's left is starting to be sufficiently low density that it becomes really hard to see how anyone earns a return. The other interesting takeaway that comes from an analysis by the let me share some numbers, by the way. I think about it in terms of deciles of the country. So the densest 10% of the country by population lives in census block groups that average about 700 homes per road mile. The second is only about one fourth that dense.

Guest Speaker 6m39s

The second densest 10% is in a 167 homes per road mile. Most of the FCC data that you look at, and and this is confirmed by most of what you hear from the companies as well, Says, you know, the bulk of the activity that's going on right now is probably happening in the fifth decile. So call it roughly the fortieth to fiftieth percent of the country in terms of density because there's a fair amount of fiber that I as I've I said before, has been done out at the very low density part of the country where it's been subsidized. So take that out, and what's left puts you probably in the fifth decile. That's about 76 homes per road mile, call it.

Guest Speaker 7m19s

If you believe the numbers for what's going to be built over the next few years by add up what AT&T has said they're gonna build, add up what Verizon now that they own Frontier is going to build, what all the independents, the Chantels, and the Unities, and the so on, and the Brightspeeds. Add it all up, and it would say we're going to be in the seventh or even eighth decile within another, call it, four years. And if you look at the number of homes that are being built, we're gonna not just maintain the same number of homes passed every year, but it's actually going to accelerate. One of the really interesting observations that comes out of that is, well, if that's true, the seventh decile is about 47 homes per road mile, so it's about 40% lower density. So you'd have to build 40% more road miles to move from the fifth to the seventh decile.

Guest Speaker 8m18s

And then it drops again by about half, so you you're talking about more than doubling the number of fiber miles that are gonna have to be installed every year in order to maintain the same number of homes passed. That's before you get to the cost implications. But you have to wonder where's the labor gonna come from to double the number of fiber miles being built in The United States at the same time that there is all this demand for fiber to new data centers and their supply chain constraints on fiber because of data centers and the equipment to deploy fiber because of data centers and so on?

Roger Entner 8m54s

Yeah. No. I'm I'm with you. By the way, interesting backdrop. Don and I worked with a I don't know how small we can call a small rural telco.

Roger Entner 9m6s

We helped build a rural telco with about 50,000 subs, their fiber strategy, where to build and where not to build. And it was largely a density issue. Where it became interesting was where they had their own backhaul. They had backhaul and they connected several cell sites. And then it became interesting to get to some households that you would not have built if you wouldn't have had already the backhaul because you could simply spread out into the neighborhood and pick it up, and there are pockets like that.

Roger Entner 9m42s

But yeah, overall, you're absolutely right. I was at Fiber Connect this year, the data centers are buying up literally everything. They have set asides for Bead, so that there is like fiber left over for Bead deployments, but that's, like, what, 10%, 20% of that. Then Verizon and AT&T have their quotas that they bought already from Corning, and everything else gets bought up by the data center guys.

Guest Speaker 10m11s

That's naturally very inflationary. Right? It drives up prices, and it also drives up prices of labor. You know, that's always been one of the things I've been fascinated by is that the labor supply to do this has historically been, if not predominantly immigrant, there's certainly been a large immigrant component of the fiber deployment labor supply. And at a time when we've closed the borders to new immigration, it makes it much harder to see where is the labor force going to come from to meet these demands.

Guest Speaker 10m43s

We'll talk about the cost implications of all that in a second, but it just raises questions about what's the feasibility of even doing this.

Roger Entner 10m51s

Yeah. And and what a lot of people don't recognize is that, like, 75% is, like, good old trenching and other basic construction costs.

Guest Speaker 11m2s

Exactly right.

Roger Entner 11m3s

Fiber by itself is not that big.

Guest Speaker 11m5s

No. That's right. The fiber piece of it is relatively small. The vast majority of the costs are labor costs.

Roger Entner 11m11s

Yeah. Hard hand labor costs.

Guest Speaker 11m13s

It's still just good old fashioned construction projects. And I didn't mention, but the other factor that really impacts the cost of deployment and feasibility of earning a return is whether it's aerial or whether it's buried. It's obviously much cheaper to hang fiber from telephone poles than it is to have to bury it. So not surprisingly, years and years and years ago, as I said, I've been doing this a long time. Years ago when I was working with what later became Verizon, part of the reason that Verizon was able to be first with what later became Fios is that so much of the Northeast isn't just dense.

Guest Speaker 11m52s

It's also aerial because it was built such a long built meaning the houses were built such a long time ago before that anybody was burying infrastructure. Things, generally speaking, if a neighborhood was developed after about 1970 or so, it became more popular just because it looks better to bury the infrastructure. So, you know, the suburbs of Atlanta, say, or places like that tend to be buried, whereas a lot of the Northeast tends to be aerial. And so it's just a lot cheaper to deploy in the Northeast than anywhere else. And so Verizon was able to make the economics work.

Guest Speaker 12m27s

Even Verizon, I think, in those days admitted

Roger Entner 12m30s

Oh, it was very picky. They were very picky in 2000. Like, Boston, they stopped building out because it became no longer by the way, the price difference is a 100,000 versus $300,000 per mile. By the way, what was really interesting, I met with other construction companies, and by the way, we work with a company called we license our data to a company called BCTI. They are a software company or fiber overbuilder.

Roger Entner 13m1s

They're using our data, our NPS data to target neighborhoods, but they have all the fiber miles, they have like they even have the soil that the neighborhoods are in because the cost of trenching clay versus gravel or rocks is massively different. For example, the trench or even aerial in West Virginia because of the Appalachian Mountain features with a relatively thin topsoil drives the cost of deployment, even aerial fiber through the roof. Like, one of these dimensions that are absolutely fascinating.

Guest Speaker 13m40s

See, I I'm fascinated by that stuff. And so that gets to the next part of the discussion, which is how do you earn a return on this? Because the nature of convergence where the sort of the frame for all this. I used to describe it to then CEO of, later of Verizon, Ivan Seidenberg, who was my client back all those years ago. And I used to say to Ivan, you know, that customers love a bundle, and they love it to the tune of about a 20% discount, which is to say that if you're gonna get a customer to take a convergence bundle, that is if you take my broadband, you're also gonna take my wireless.

Guest Speaker 14m19s

And if you take my wireless, you're also going to take my broadband. What you're asking the customer to do is to accept restricted choice for the second product in the bundle. People tend to think you're doing the customer some kind of a favor by making it lower friction or what have you. That's not the case at all. What you're really doing is saying, I'm asking you to accept restricted choice.

Guest Speaker 14m41s

And instead of optimizing on each individual product, I'm asking you to take two products from me instead of one. And in compensation for accepting restricted choice, I'm going to pay you for that, and I'm gonna pay you for that in the form of giving you a discount. Now by the way, that may be perfectly fine and sensible to do if the cost structure is such that one plus one on the revenue side is going to equal 1.8 instead of two, that's fine as long as one plus one on the cost side equals 1.7. And if there's real cost synergies associated with offering them together, then it makes sense to do. The problem is there aren't really big cost synergies for a wireless operator building fiber.

Guest Speaker 15m23s

The cost for a wireless operator to build fiber is not really all that different than the cost of anybody else building fiber. There are real cost synergies to offering FWA, by the way, because as a fallow capacity model, its marginal cost is essentially zero as long as you have capacity to do it. The real question is, do you have capacity to do it? But the cost structure of that is quite attractive, and the cost structure of cable operators offering wireless is also quite attractive because of the offload that they can do onto their own Wi Fi and ultimately CBRS networks. So there are real cost advantages for the wireless plus FWA bundle.

Guest Speaker 16m5s

There's real cost advantages to the wireless plus cable bundle, but there aren't really big advantages to the wireless plus fiber bundle.

Roger Entner 16m13s

Here's where it comes in. We clearly see it in the data, and I wrote a couple reports about it. You know, I described it like fiber castles, table forts, FWA camps, and satellite warbands. One of the things that you can see in the data is that where the fiber provider is active, there's about a 15% chairlift, and it's more than just the old ILEG two point zero lift. So, the ILEG, the AT&T where they were the ILEG, Verizon where they were the ILEG, they have about a 5% lift.

Roger Entner 16m52s

But where they have fiber, they have a 15% share list.

Guest Speaker 16m56s

That's about what we've seen in the data from AT&T as well.

Roger Entner 17m0s

That's where they're banking on. Right? It's that 15% share list.

Guest Speaker 17m4s

But, Roger, what's really interesting, we did some modeling of that because, you know, where all this comes together is, okay. So what is the internal rate of return or the IRR of building these things? We'll get to the overall returns in a second. But interestingly, you would think that there's a huge benefit to getting an extra 10% higher market share in wireless. But AT&T's discounts, at least to do it just for AT&T, the discounts that they give to customers are so significant that the IRR of fiber with an extra 15% mobile penetration, but giving away the discount that they do is not really all that different than the IRR of building it without the bundling discount, but also without the uplift in market share.

Roger Entner 17m50s

Yeah. But the problem is that both Verizon and AT&T are losing customers outside their fiber footprint. Right. And they're leaning inside the fiber footprint.

Guest Speaker 18m1s

It's a fascinating point, Roger, and the data bears it out. By the way, kudos for giving them much more colorful names than than I ever have for your fiber castles and so on. Because the data from AT&T suggests that they are gaining accounts at a sufficiently rapid pace within the seven and a half percent of the country that is subscribed to their fiber, then it accounts for more than all of their wireless growth, which means they're losing wireless accounts in the other 92% of the country. That's a real problem if you think about the long term strategies for these companies. Absolutely.

Roger Entner 18m39s

For both of them. What do

Guest Speaker 18m40s

you do in the rest of the country?

Roger Entner 18m42s

But that's why they're both racing towards fiber. That's why T Mobile T Mobile will be fine for the next five years. But when you play this out that if we're moving into a convergence world, just structurally T Mobile will have fiber towers because T Mobile will buy, at least for the time being, only pure play fiber guys because copper is cooties and they don't know how to handle it. The question will become, is there enough space for a non converged player? What's really interesting and what we saw in our data, in the past I always said like, oh, it's one third, one third, one third, that one third wants to bundle, one third can be convinced, and one third doesn't.

Roger Entner 19m29s

The dozen is actually like 5%. Everybody else is persuadable, either by superior product or superior price.

Guest Speaker 19m39s

That's interesting. You know, it's funny. That's consistent with what I've always seen in telecom. Years ago, I did a conjoint study for what percentage of customers would take voice over IP from cable operators. This is twenty years ago.

Guest Speaker 19m53s

And the upshot of that study when we really step back and try to figure out, okay. What is it really saying? What it was really saying was 25% of the customers in those days hated their cable company so much there was no price which they would switch to cable. 25% of the customers hated their phone company so much there was no price that they wouldn't switch. But 50% of the customers said, I don't care whichever is cheaper.

Guest Speaker 20m16s

And, ultimately, it said that the market share for the cheapest player was gonna be 75%. And at a

Roger Entner 20m22s

time when people And you

Guest Speaker 20m23s

work can never get past 15 or 20% of the landline business. It could never happen. And the data said, oh, yeah. It can.

Roger Entner 20m31s

Oh, yeah. No. And and I remember when Cablevision became the first non ILEC, a non copper company to be the ILEC on Long Island.

Guest Speaker 20m44s

Yes. Yeah. Dominant provider. That's right. Exactly.

Roger Entner 20m48s

The cable provider became the biggest market share company.

Guest Speaker 20m51s

So, Roger, I wanna go back to where we were a minute ago where we were talking about the returns. Because I think, to me, here's the interesting part. Because it gets to on the one hand, you wanna make sure your fiber footprint is as big as it can be. But on the other hand and we've published all this work for our clients, so I won't go too far into the math. But the challenge is as you get out to that six seventh decile that I was describing, even if you believe that the ARPU that you're going to get, the revenue per month that you're gonna get from fiber is $70 a month and growing at a compound annual growth rate of 3% per year.

Guest Speaker 21m32s

You still can't make money even with the uplift in market share from adding mobile in the seven step up. But if you assume that there's going to be price compression and that it's unreasonable

Roger Entner 21m44s

And it gets all the worst. Right?

Guest Speaker 21m46s

Yeah. It goes to the point where you can't even earn an acceptable rate of return in the places where you're building today. And I think that's to me the fascinating question because it sure looks like we're starting to see real price compression in broadband when Comcast offers a $50 five year rate lock and Optimum offers a $50 five year rate lock. And they're sort of signaling, put $50 in your model, not $70 in your model. And at $50 in your model, I'm not suggesting that the big phone companies are gonna go bankrupt.

Guest Speaker 22m22s

But the smaller operators that are trying to build into a world where it's gonna be $50 ARPUs are facing bankruptcy. And all of their business models are predicated on the assumption that there's going to be a buyer out there that's gonna take them out at 12 times EBITDA. But what if there isn't? If there isn't, the free cash flow in perpetuity that you get from those systems doesn't come close to earning a return on the capital employed. And, eventually, that means you can't service the debt.

Guest Speaker 22m53s

So you're gonna have lenders that are struggling unless you just say, well but AT&T or T Mobile or Verizon will buy them anyway even if it doesn't make any economic sense. And I'm skeptical of that. It's not obvious that you're gonna have those buyouts. And if you don't, then there's going to be a real crisis coming in the fiber industry because everybody is committed to keep on going even though we've passed the point where the density is high enough and the pricing is high enough to earn a return on that capital.

Roger Entner 23m26s

Well, they might buy them out of bankruptcy. Right? Yeah. The thing is, in my conversations with the smaller rural telcos, we're building reports on the 500 largest privately held telcos to look exactly through these things. And like, the financing, like, I just looked yesterday at the three zero six ECAM recipients and how much money they get, which is a very interesting figure.

Roger Entner 23m53s

Happy to share with you. The smaller ones are realizing that they can't push out too far for exactly what you described. It becomes not profitable. At the same time, what we are seeing is, over the next three years, the same rural telco consolidation that we saw in wireless twenty years ago. It's the same trend.

Roger Entner 24m19s

It's the same mechanics with the same outcomes. Interesting. There is still a ton of money sloshing around that wants to consolidate. There are, like, three kinds of buyers. Right?

Roger Entner 24m32s

The ultimate buyer is the are the five large telcos that will be around in the end. AT&T, Verizon, T Mobile, Charter, and Comcast. Those are the likely right end consumers of all of this. And then you have the PE funds that are intermediate holders to ultimately sell to somebody else. And then you still have some smaller rural telcos who will buy up their neighbors for efficiency gains.

Roger Entner 25m1s

But then it stabilizes, but it's fascinating what's going on in that market.

Guest Speaker 25m6s

It really is. You and I have been doing this long enough that we've seen these cycles come and go before when the cycle turns, and it usually is driven by the debt markets way before the equity markets where the debt markets figure out that the smaller guys can't service the debt. Then everybody starts to take notice of, wait a second. We've got a problem.

Roger Entner 25m27s

But as of right now, there's still money to be had at very reasonable price.

Guest Speaker 25m33s

The problem is that that makes the problem worse instead of better.

Roger Entner 25m36s

Yes. I know. But in the end, we're here to help them all clean it up and and That's great.

Don Kellogg 25m44s

In the long run, we're all dead. Right? And in the long run, we're all dead. Right?

Roger Entner 25m48s

Absolutely. But, no, this is this was awesome. Thanks so much for coming. I hope that in in not too distant future, you come back and we can talk about satellite.

Guest Speaker 26m0s

I'm looking forward to doing that. And as always, I love my conversations with you. I always come away smarter in all that I learned from you. So thank you for that, Roger. I appreciate it.

Roger Entner 26m9s

Well, that makes two of us. That's why we like each other and why we spend time with each other because we both learn and appreciate each other greatly.

Don Kellogg 26m18s

Alright. Thanks, gentlemen. Roger, we'll talk to you next week.

Roger Entner 26m22s

Talk to you next week.