Thanks, everybody, for joining us. This is our First Comm. Infrastructure session of the day. We're really pleased to have with us Dan Schlanger, EVP and CFO of Crown Castle. Thank you very much for joining us, Dan.
Thanks for having us.
Do you have to make any safe harbor statements or anything like that?
Just check the damn 10-Q.
Okay. So let's start off. So look, towers have had some tough sledding from a stock performance standpoint this year. I think it kind of began with the realization at second quarter results that, among other reasons, I've actually got three reasons. I think number one is that people gave up on the idea of recession, so owning defensive stocks was kind of fell out of favor. Second was that as people decided that we're not gonna have a recession, higher rates for longer meant people had to continue to make adjustments for refinancing expenses and those sorts of things. Both of those things are kind of out of your control and not your really fundamental wheelhouse.
But the third thing was, I think, the realization that the anticipated carrier slowdown in capital spending and activity levels was happening more abruptly, I think was the word that got thrown around starting in the kind of at the end of the second quarter into the second half. So maybe we start there and talk about why do you believe that carrier activity levels kind of dropped off in ways that were kind of unanticipated at the beginning of the year?
It's a good question, and what I thought we anticipated, as your question implies, that activity levels were going to slow coming into 2023. When we gave guidance at the end of 2022 for this year, we were very clear about that. We had come off a year growing more than 6% in our tower business, and we thought that we were gonna get to about 5% growth. At the time, I think a lot of investors had not anticipated that level of drop-off and took it as some negative that was affecting us and not our peers.
That was from the third quarter of 2022 guide.
Yeah, the guide that we gave in October of 2022 for the 2023 year.
Mm-hmm.
At the time, even at that point, we said, "We still think we're gonna lead the industry in growth." I think we got a lot of questions about that. "Well, how can you lead the industry if you're slowing down?" I think we were more right than wrong, that we anticipated a slowdown, and a slowdown has occurred. It's very similar to other generational upgrades that have happened in wireless technologies for decades. When we've gone from 2G to 3G or 3G to 4G, we've seen a surge in activity that occurs as our customers push out that new generational upgrade, and then they settle into more of a run rate type of capital environment that continues to expand the utilization and deployment of that new technology for a decade.
We saw that going from 2G to 3G and 3G to 4G. I think what happened this time was, as you pointed out, the slowdown was more abrupt than what we expected. It occurred more abruptly is that our customers acted more at the same time than they had historically in the generational upgrades. So you were around for the 4G upgrade, which at the time was called LTE.
Thanks for pointing that out, yeah.
I think it's a good thing in your business.
I was actually giving you credit.
Oh, I'll take it positively.
Thank you.
But Verizon led that push and pushed really hard into LTE and wanted to claim, and did claim, the brand of being the best network. T hey were kind of the first phase of the surge, and then other companies followed them.
Mm-hmm.
Therefore, when Verizon came down, other companies were still ramping or on the higher level, and so it looked like less of an abrupt change.
Mm.
This time around, our customers generally acted at the same time.
In unison, yeah.
Then in addition to that, in unison activity was DISH-
Mm.
which was a new nationwide network that had not been built in previous generational upgrades. Because the nationwide networks had generally been built prior to 4G, and actually prior to 3G. Like, it had been the nationwide network was out there. So we had all of that activity hit at the same time over the course of 2020 to 2022, which was reflected in our business. Our tower growth was really high, greater than 6%. We grew our dividend 9%-11% a year. Like, it was. We were delivering those results to our shareholders. I think what we didn't recognize quickly enough or when we were thinking about it, was, well, that also means there's a chance that everybody comes down at the same time, and we knew it was going to be a reduction in activity.
But if you go from a higher peak than historical to a lower than historical, it looks very abrupt.
Mm.
Even though the general driver of the activity is the same dynamic. Push out really hard to get your new next-generational upgrade out in the world, utilized by customers, and then pull back a little bit and settle into a more run rate. That dynamic is still in play. What is interesting, though, is that despite this reduction in activity that we are seeing, we believe that the run rate for 5G will be significantly above the run rate for 4G as an ongoing expense or capital expense for our customers, for two reasons. The first is just the quantum of data is so much higher now than it was at the beginning of 4G.
So the base, if you're growing 20%-30% mobile data demand per year, the base being higher matters because then the increment's higher, and therefore, the amount of money you have to spend is higher. Secondly, we believe that because the density of the demand has already been established, at the beginning of 4G, towers were the only thing being deployed. By the end of 4G, we had a bunch of small cells being deployed, because the density of the demand was already so high that small cells were required. That's true right now for 5G, which is why you saw both Verizon and T-Mobile give us large orders, even during this period of time, they were significantly focused on towers.
They gave us large orders for small cells because they knew the density of the demand was already there, and they needed to be delivering some additional capacity to the areas where that demand density required. So we believe, because small cells will come in as a more, an earlier version of what the network architecture upgrades will look like, that amount of money will be spent as well. So we have two things going on. Like I said, the quantum, overall quantum of data is higher, and we have more small cells, more density of the demand. Those two things will require our customers to spend more money in 5G than 4G.
So a couple follow-up questions on that. I guess I'll come back to the kind of core fundamentals on that, but I guess within the sector, we kind of have two bookends at this moment in time with this slowdown, kind of, or lull, or whatever you want to call it. One end is American Tower, who kind of, you know, unapologetically went after a lot of holistic MLAs, you know, presumably made trade-offs, but now they've got 90% of their business contracted out through, you know, 2025, and, like, through 2027, I think they're saying, like, 70%. And the portion that's not contracted is really just the maturity of existing leases, 99% of which always get renewed anyway, so there's not a lot of variability in that.
On the other end is SBA, who has very unapologetically not really valued these holistic relationships and has been very activity-driven, and has been guiding to more than a third slowdown, percentage slowdown in their new business revenue expectations for next year. So there's kind of these heads where higher highs and lower lows in activity levels, whereas you kind of more smooth it out with these holistic MLAs. Where does Crown sit in that spectrum, both philosophically and actually, contractually today?
Let me answer the question on the contractual first, and then I'll get to the philosophical. In our earnings materials from last quarter, we had a chart that we put in that showed through 2027, we believe our tower growth will be 5% per year on average, and that 75% of that has been contracted already. So just on a math basis, we are much closer to AMT than not having anything contracted. You said they said 70% for 2027, we're at 75%. I would say that's the same thing.
Let me ask a follow-up on that. So 70%-75% of 5% is 3.5%. We know that the escalators are contractual, so that's 3%. So that does that mean that you have a 0.5% contractual volume growth and, or am I misunderstanding what you're saying with that?
So I would say that the 3.75% is the minimum that we are going to get, according to our contractual agreements. That does include escalators, so you can interpret it however you want. Philosophically, what we think about is our job is to try to deliver the highest risk-adjusted return we possibly can to our shareholders. We have managed the risk side of that equation in many ways, one of which is through the agreements we have signed with our customers. When we look back, and you can see this in the earnings materials we put together, in the 4G upgrade cycle, we had some swings and volatility in the growth rates of our business.
So as we renegotiated our agreements, we took that into consideration and tried to get a floor that we thought was reasonable without cutting off too much of the upside. We have sophisticated counterparties. There's no way we're going to get something for nothing, so we do know that there are trade-offs. So the floor that we're at in 3.75% now is actually higher than what some of the years were in our 4G growth cycle. So we did effectively raise the floor. O ur growth over the course of the last two years at 6%-6.5% is still higher than in most of the years that we had in 4G. So we didn't cut off the upside. So we've reduced the risk without taking off too much upside or much upside at all.
We believe that having an MLA that has a tremendous amount of locked-in growth into it is beneficial because our shareholders want us to have relatively consistent growth over a long period of time, in our belief, as opposed to being able to grow 10% one year and 0% the next year and average out to 5%. I don't think that would be good for our shareholders. That's why we have not gone to CPI-related escalators. We want more fixed escalators because we think that stability is as important as the growth itself and why we have locked in the growth through MLAs, a lot of them.
And that risk-adjusted return, we believe, has played out because we—I look at what SBA has returned in terms of just growth in their business, and their net growth has not exceeded our net growth over the last five years. When you think a la carte, in the highest growth period, we just went through the last two or three years have been one of the highest growth periods we've seen. You would think that a la carte would allow you to price up and therefore get more growth, and they weren't able to. So they have more risk because they do not have long-term contracts, and they have not shown more return. That doesn't, to us, make sense from a long-term value creation standpoint. So we are always looking to try to figure out how to drive the highest risk-adjusted return we can.
That carries over to a lot of other aspects to our business. It carries over as to why we focus only on the U.S., because we believe it is a lower-risk market. We believe it is the best market in the world to own communications infrastructure because the growth is the highest, the investment from our customer is the highest, and the risk is the lowest. As long as we're paying a U.S. dividend, the risk is the lowest. You can't pay dividends in REITs. You gotta pay them in dollars. So it matters. We have lowered the risk in our balance sheet because we have gone to a longer tenor. We've almost doubled the average maturity of our debt over the last several years. We've gone to very low floating rate debt because we want less risk in our balance sheet.
We have extended those maturities, so we don't have any one period where we have a huge stack that needs to be refinanced, and we have a very large revolver that allows us flexibility to choose when we go to the market. All of that was a choice. If we had, during the course of zero interest rate environment, chosen to go shorter term and floating rate only, we would have printed more AFFO. We would have had higher cash flows. But we chose not to because we wanted to reduce the risk in our balance sheet and try to drive the best risk-adjusted return we can. That is our philosophy. It will continue to be our philosophy. I don't believe we have been recognized in the market for having done so.
I can't find where we would've been recognized in the market for having done so, but I believe that we have delivered on what we think is the best way to drive long-term value in a business like ours, which is lower the risk and don't cut off your upside.
Let me just make sure I'm getting it right. So 3.75% growth, is that a CAGR?
Yes.
Does that include the Sprint churn and-
No, it excludes-
Excludes the Sprint churn.
Sorry, I should have said that. Thanks for reminding me.
So with respect to the 3.75 kind of locked in, what has to happen to create that 1.25 premium?
Additional activity that wouldn't be caught up in the agreement itself, which would be both activity with the big four customers, where they have some rights to some portions of what we have, but they don't have rights to everything all the time, and then other customers that we could lease our towers to.
And so, obviously, you know, looking back over the many years, you know, I've been looking at the space, you know, there's always been a premium above the escalator. There's always activity. There's always densification that's happening. I think to your point, there's been this big spend. Now there's a lull again, I would call it a lull. But the difference between, I think, where we were in 4G, where we are in 5G, is that I remember, unfortunately, the 3G, and I had a Sprint phone, and I could download ringtones, and that was super cool.
That's cool.
We'd all share, "Oh, my gosh, I got, like, you know,"...
Baby Got Back?
Prince, Purple Rain on my phone. But then when we got to 4G, there was something we could do that we could never have done before, which was, you know, the iOS, Apple Store, the Google Play Store, applications became a thing. All of a sudden, you know, Google and, you know, Seamless Web and Uber just became parts of our lives, and it really was a step function change in utility for the consumer, and it just fueled massive investment because it was so obvious that there was a monetization opportunity. I think that right now, you know, the industry's struggling with the lack of some new incremental opportunity to put new money to work.
That is informing a pullback and a lack of need or necessity to put real incremental capacity to work because they don't really see a reason to put new money to work there. Do you agree with that, or do you. If you do, when do you see it changing? If you don't, why?
Yeah. It's never been our business to try to figure out what the use case is. It's been our business to provide infrastructure to allow that use case to happen. All of the things you mentioned, though, were not the original use case that was envisioned-
Right
- for 4G. Not one of them was. I believe that we all have a hard time as humans just predicting what the future is going to be and what that use case looks like until it comes, and there's always been a chicken and the egg problem. You either need the network to work so well that the use case can be utilized, or you need the use case there to drive the network to get there. I think we're in the kind of cycle of which one's gonna go first.
Mm.
But our customers spent more than $100 billion to increase the utilization of their network through the purchase of spectrum and the deployment of that spectrum. That is not no investment, and that is not a slowdown of investment. A s I said before, I fully believe that the overall cycle of 5G will require significantly more investment than the cycle of 4G.
Mm.
The use cases are going to come. I believe they will be different than the ones that we all think of right now. I don't think that they're going to be, that we, as a consumer, somehow are going to have a use case that is akin to Uber. That would be the repeating of the past. That isn't what happens. There are step function changes-
Mm
that happen as technology changes. I believe the most likely use cases will be industrial ones, where companies use 5G to try to reduce their costs, as opposed to consumers use 5G-
Mm
To try to figure out ways to better live their lives. I f customers become industrial use cases, I believe they will pay for the incremental benefit that they get for 5G and will allow for a significant investment of both technology and infrastructure. W e, as Crown Castle, are uniquely positioned for that investment because I don't see that running on towers.
Mm.
I see that more running on small cells. You're talking about industrial use cases, it seems more likely they're gonna be small cells than towers. More capacity and denser markets, that's what we're looking for. More capacity and denser markets points more to small cells than it does to towers.
Mm.
But towers are gonna be required and are still the most efficient way to deploy spectrum over large areas of both population and geography. So we will continue to grow as a tower industry because that continual just eating of demand that we do, 20%-30% a year, will require towers to continue to add capacity and coverage to the market. But incremental use cases very well may be on the small cell side of the business that we think we are the only company that's positioned to take advantage of.
Mm.
We believe we are the cheapest way that anybody can take advantage of that type of activity, because we can share the infrastructure, and it therefore lowers the cost for any one customer or one person to use that infrastructure. That feels like a very good market for us and a very good setup for why we have done what we've done and position us to the future.
So you make a good point about the lack of predictability about these evolutional changes. One of the things that was the unanticipated outcome in terms of use case for 5G to this point in time, has been fixed wireless access. The networks were not built assuming the fixed wireless access was gonna be the killer app, at least in the early days of the evolution of the network. So, is fixed wireless access, 'cause just to make the point, T-Mobile wants to double the number of fixed wireless access customers they have in the next, I think it's two years. Verizon, roughly the same. And Verizon's gonna get twice as much spectrum across twice as much footprint in the country, and they've talked about having to roll that out over the next two years.
Is fixed wireless access a meaningful contributor to that missing 1.25% in getting to 5% growth rate today, next year, or the year after, or no?
I agree with the premise of everything you said. I think fixed wireless access was never the thing that, that our customers, the carriers, were looking at to try to build the network when they built the network. I think in the early days, T-Mobile has, and it's still been very clear, that they limit the number of consumers per sector-
Mm-hmm.
so that they don't overwhelm their network with fixed wireless access. Because I, I don't know, you may know this number better than I do, what does the average home consume in terms of gigabits per month?
Charter was telling us yesterday that they think it's about 700 GB a month.
What's the mobile user?
Ten.
Okay, so you have 70 users per home. That's a big allocation of spectrum, big allocation of capacity. So if they want to double, they're talking about 70 users per home. That's a huge increase for us as an industry.
Mm.
That would absolutely be a positive for Crown Castle and our peers. That everybody's gonna benefit if the spectrum that T-Mobile acquired with Sprint or that AT&T and Verizon acquired with the C-Band auction is being utilized to provide a single location with 700 GB a month. That's a great outcome for us, because what it's doing is it's consuming all of that capacity faster than we would have otherwise consume it as an industry. Which means we need more and more infrastructure because we need more and more of those sites that allow for that spectrum to be deployed.
I believe that the ability to provide fixed wireless will likely start as a tower deployment, because it's the cheapest and easiest and best way to deploy spectrum over a large area, like I talked about, but over time, likely gets to more small cells.
Mm-hmm.
Because the way towers and small cells work together is towers generally provide the coverage and small cells provide the capacity. If you get enough capacity in a neighborhood, you build a small cell system there. That feels very positive to me when I think about where we've positioned ourselves. I don't. To your point, though, I don't think that anybody would have guessed that fixed wireless was going to be the first. What's interesting is, I don't think fixed wireless is a new concept. Everybody has always thought about it. It just was never available. The capacity and latency was never available, and 5G has enabled it to be available.
Mm.
That's exactly what I mean, is that you never know what's gonna happen until you have it. T hen people are like: You know what? Maybe we can use it this way, and if we do, maybe I can solve this problem. T hat's what happens in the world, and especially in the U.S., because the innovation typically happens here, where the next thing is developed by people who work and live in the U.S. I t's great because it drives a tremendous amount of activity for us. I think at any point. I'll go back to 3G. I believe at 3G, the upgrade to 3G, the idea then was, as long as, as soon as everybody has a phone and is using their phone, there will be no more growth in the wireless industry.
And then in 2006, the iPhone came out, and everybody's like: "That's not gonna work. It's web-based. We don't need web-based." We were using our BlackBerrys at the time, because the BlackBerry was the, the pinnacle of, of technological advancement. T hen when the iPhone came out and everybody was skeptical: "We're not gonna use that thing. It's web-based. It's not secure. It doesn't have a keyboard. It's just not gonna work out." I t turned out to totally change how we all live our lives. That's. You can't predict that.
Mm-hmm.
So every prediction I've ever seen about incremental wireless data usage has 20%-30% for some period of time, and then it falls to 10%-15%, every time. And if you look over the history, that's never occurred because there's always something that you can't think about that happens, that drives it back up, that data demand back up. I believe that's gonna happen again. I just, I think that we demand it. We don't know we demand it, but we demand it.
So that being said, so I'm assuming your plan is to give guidance in October?
Yes, it is.
Okay. So let me get as much of that as I can now. I guess the tower revenue growth guidance for 2023 implies, I think, $60 million-ish for the back half of 2023.
Mm-hmm, it does.
Would it be wise or unwise to assume that that represents a good jumping off point run rate for 2024?
Yeah, uh-
Well-
Our business is a relatively stable business, and I would say every time that we give guidance, it's always reasonable to take the fourth quarter and annualize it as your baseline. But that doesn't mean that that's the right number for the next year. Things can change.
Mm-hmm.
Of course, it's the good baseline. Our business just doesn't change that much year-to-year. You can look at our leasing over a long period of time, and if you were to annualize the fourth quarter, you'd be pretty close almost every year. O n a business the size of ours, we're talking about $5 million and $10 million differences that people get focused on.
Mm.
Which I understand, because we're such a long-term growth business, $5 million or $10 million now matters over a long period of time. But it's that $5 million or $10 million that I think is what people really care about, not-
Mm
whether you should annualize the fourth quarter.
Got it.
That's not what I care about.
So maybe-
Mm. Yeah, it is.
Well, okay, if you want to give it to me, go ahead.
Yeah, it is.
So maybe shifting gears a little bit to fiber services to start. The two kind of questions there, kind of the mirror of towers. The fiber kind of revenue growth guidance implies a big step up, you know, the $30 million range a quarter in the second half of the year. Why is there kind of this assumed step function, second half versus first half in 2023, and is that a good jumping off point for 2024?
I think part of what happened was we have a weird comparative period in 2022 in the first quarter versus 2023 in the first quarter. So we had non-recurring items that hit in 2022, that we had called out at the time-
Mm-hmm
about $10 million. If you normalize, you take, just take that out of 2022, and it looks like the first quarter of 2023 is more in line with the numbers you're talking about.
Mm-hmm.
So I don't think that there is a step function, and I believe that it is reasonable for us to assume that we will reach our guidance, or we wouldn't have given our guidance.
Your longer-term guidance for that business is 3% growth. You know, just as we've kind of been listening to some of the other, you know, companies in the fiber services business, 3% growth is, is pretty aspirational, for them, whether it's, you know, lack of return to work or slowness in decision-making or, you know, budget cuts or whatever it happens to be. Where does the 3% growth guidance comfort level come from?
Well, it comes from our ability to understand the market that we play in and the history we've had. If you look at the history of our business, we've had a pretty consistent level of leasing activity over the course of a very long time, and we feel like we have, again, a reasonable expectation that the leasing activity will continue. We have purposely focused our business on delivering engineered solutions to larger enterprises. So whether that's a government agency or education or medical facility or financial services firm, those types of customers plan for longer periods of time and require a higher level of engineering and capacity. Because we have built our fiber network based on the idea that we want to have small cells, we have more capacity than most fiber networks.
So we build significantly more strands of fiber throughout the markets than most have done. When we have purchased fiber, we were always focused on making sure that the systems we're purchasing had lots of strands of fiber. So we have the ability to dedicate more capacity and more engineering capacity to singular customers, and we believe that allows us to have more visibility, more stickiness, and a higher growth rate, partly because we are, we don't have a lot of smaller businesses that are more apt to churn and more price sensitive. When you add all that together, we've shown we can grow at 3%, and we believe we can continue to do so.
And then kind of shifting to the small cell business. You know, you've kind of... You had a pause there while the C-Band build was happening. You were kind of run rating more like 5,000-6,000 nodes a year. You've got it to kind of that coming back up into the 10,000 level this year. But there's obviously some churn in there, which is kind of mitigating that. But the 10,000-11,000 kind of nodes added per year seems to be the comfort level that you have and the expectation you have.
When you talk about, you know, I don't know if you've kind of really even given this outlook into the 2027 zip code, but should I think about the next three, four, or five years, that 10,000-11,000 nodes per year is kind of the cadence because that's just what's doable in a, in a real-world scenario?
We haven't given anything in 2027, but I can tell you that going into 2024, we believe we will be at 10,000 nodes or more going into 2024, just based on the backlog we have.
Mm-hmm.
So we have 60,000 nodes in backlog, and we believe we can put those on air and, and get to 10,000 nodes or more and believe that we can drive double-digit revenue growth going into 2024 over 2023. Because you do take some of that churn out you were speaking of, and we get more growth. That's a, that's a great place for us to be. T o your point, the activity that caused 2024 to be where it is started two or three years ago, because it takes a long time to get small cells on air. So even in the middle of our customers focused on towers, they were planning for small cells, and therefore, now we're seeing the benefits of that planning. Our... As you pointed out, the small cell business, during that tower focus, went down.
Makes sense that, okay, we don't wanna do those right now, but we know we need them. So can we plan for over a period of time to put them on air? W e are doing so right now. I think that's very positive for our business. I think it shows that all the dynamics we've been hoping for are coming to fruition, that our customers understand the need for small cells, understand the need or the economic viability and benefit of utilizing a third-party shared infrastructure model, and we get to do new builds and colocations. That all plays very well with what we had expected. T herefore, we see lots of growth coming forward, coming into the future. To answer your question about the longer term, 2027, or the implication in your question, we have 60,000 nodes in backlog.
There is no real-world constraint as to how many nodes we can put on air.
In a given year?
In a given year. There's no, there's no constraint that says we couldn't do more. The constraint more is the demand from our customers.
Mm-hmm.
When do they want those small cells on air to provide what network coverage at what point? Because if we got a single order, this is hypothetical, so if we got a single order today for 50,000 small cells that they want to build by 2027, we could put them all on air in 2027 if we knew about that right now.
Mm-hmm.
Because we could just put all of it in place. Yes, it takes, it takes 18-36 months to build them, but it takes 18-36 months, whether there's one or 50,000 of them. We can get it done. We've shown that we can increase the... I, I don't, I don't know about 50,000. It was a hypothetical question, but we can figure it out. We've shown that we can go up and down in the amount of small cells we build in a year and, and manage that process really efficiently. So I don't think we have been the bottleneck, nor do I think that we would cause a bottleneck. But the one thing that's running through my mind as I'm saying this is maybe we might run into some issues with permitting in specific municipalities that don't have the capacity.
But the reason that we're so good at what we do is that we have those relationships in municipalities all over the country, and our customers want a lot of small cells built all at once. So they don't come and say, "Hey, build me two here." They say, "Build me a thousand." So we try to do that, that whole system with the municipality so that we can get it done all at once. So making it 2,000, I don't think, would add a tremendous amount of time. So I believe that the constraint more is: what's the demand, specific demand? Where do our customers want the small cells? When do they want them on air?
That's the constraint on our growth, more so than is there an industry dynamic that would allow us to put a constraint on our business.
Mm-hmm. So I guess, I guess where I was going with that was because it kind of feels like the law of large numbers is gonna, you know, put downward pressure on the revenue rate of growth. But, I think Jay, last week, mentioned that, you know, you might have a 30%, 30-70 kind of colo anchor build, small cell profile, but at the margin, it's 50/50.
Right.
But, you know, of the nodes going on air, let's just say in 2024, half of those will be colos, which means that the margin in that business should be creeping up, and so we should be seeing accelerating EBITDA growth, I would imagine, even if we see somewhat decelerating revenue growth because of that colo mix shift. Is that right or wrong?
It depends on the colo mix shift and when it happens. So we believe what he was talking about was 2023.
Mm-hmm.
That we believe that about 50% of our nodes were gonna be colocation and 50% anchor build. If, for example, we were in 2024 to have the same colocation to anchor build as 50/50, which was very good, I don't think you would see the dynamic you're talking about. We would have to take an acceleration in the amount of colocation-
Mm-hmm
... in order for us to see that dynamic. But I don't believe that there is a revenue pressure at this point. You take all the things I just said, we believe we'll be at least at 10,000 nodes. We're gonna grow our revenue double digits next year. That's an acceleration of our revenue growth.
Mm-hmm.
We believe that there's going to be significant colocation. We'll talk about that more in October, which we believe will drive really good returns in our business. I think the issue that we have is that the returns don't come through because we are also still building a huge amount of anchor builds. So you don't see the overall yields that we produce every quarter increasing because you have some anchor build activity, some colocation activity, and it evens out to about 7.5% yields, and that number doesn't move around very much.
So I wanna ask one more question. I know we're gonna run out of time, but for the last couple years, you know, you've been able to kind of say to the market, "Hey, you know, we've got $3.4 billion operating cash flow, $1.4 billion CapEx, $2 billion free cash flow. We've got a $2.7 billion dividend, but we're gonna be able to cover that dividend because our EBITDA growth at a 5x multiple will allow us to, you know, not have to do anything other than what we typically do in the debt markets." Obviously, debt markets are more expensive now.
You know, that math, you know, implies that, you know, we need at least kind of $150 million of sequential EBITDA growth in 2024 over 2023 to keep that cycle going. Are we gonna keep that cycle going?
There are actually a few things in there that I think you missed, but I don't think we have time to get into it.
I guess.
So I can't confirm or deny that, and I wouldn't because we're gonna give guidance in October.
I tried. Well, Dan, thank you for coming. I really appreciate it. It was great.
It was great.
Good conversation.
Thank you.
Thanks for coming, guys.
Appreciate it.
Thank you so much. I think right now we're gonna have a short coffee break, and then back in here with Rogers. Thank you so much.
Thanks.