Conference Transcript - CTGG 2026-03-02

Frank Loudon, Analyst — Raymond James

All right. Great. Thank you very much. Really appreciate everybody being here. My name is Frank Loudon. I'm a senior telecom analyst here at Raymond James. We have Chris back here from Lumen again this year. Chris, I'll turn it over to you. I think we have a quick video.

Chris Stansbury, CFO

Yeah, why don't we run the video and then we'll get into it.

Operator

Everyone wants to talk about AI. You just want it to work. Here's the simple truth. If your network can't move massive workloads quickly, securely, and effortlessly across multiple clouds, none of the AI in the world matters. That's why we built the Lumen Network the way we did. A high-capacity backbone, real-time control, and multi-cloud routing without the headache. We're not here to add more noise to the AI conversation. We're here to simplify the complex infrastructure that makes AI work. And that starts with Reach, a network connected to nearly every major U.S. data center, expanding fast and built for AI. Because performance, reliability, and control aren't hype.

Chris Stansbury, CFO

They're your job.

Operator

AI is a network story. Start with a network built for what's next. Lumen, the trusted network for AI.

Frank Loudon, Analyst — Raymond James

All right, great. So maybe, Chris, sort of walk us through sort of the Lumen story. Talk to us a little bit about, you know, where you fit into the space, including your major business line. There's been a lot of change in the last 12 months with the company, even recently here. So talk to us about that and kind of where you fit in the space, and we'll jump into some specifics.

Chris Stansbury, CFO

Okay, great. Yeah, no, we, the transformation is largely complete. We are an enterprise-focused networking company. And the advantages that we have in that space have really been visible over the last couple of years as we've signed almost $13 billion in what we call private connectivity fabric deals, deals with hyperscalers for their AI training needs. But those contracts are really evidence of the differentiation we have in our physical network. We have capabilities that others do not, which is why we're winning in that space. But as we go forward, the real opportunity here is as we digitize that and make the network programmable, on demand, easy to consume. Because the network that has evolved over time does not support AI. And I think the world is finally starting to recognize that. The math of a GPU is significant. $2,000 an hour to run a GPU. you. And if you're not keeping those things spinning, you're losing a lot of money. And as data gets more distributed, you need the ability to move data from anywhere to anywhere. And we're the only company that's building that capability as we go forward. So we're super excited about the focus. We're excited about where that vision takes us as we go forward. And it's starting to

Frank Loudon, Analyst — Raymond James

get fun. So maybe give us a little more color on that. You said the networks don't support AI. I get this question all the time because I've been doing this way too long. I was around when all the networks, you know, when were built. And they said, don't we have enough fiber? So what is it about the networks that, what do you mean when you say that?

Chris Stansbury, CFO

So there's really two issues. There's is there enough fiber? Is there enough connectivity? And then there's the anywhere to anywhere piece. And so if you're talking to network engineers, they talk about two things. They talk about north-south traffic and east-west traffic. And they don't mean geography. What they mean by north-south traffic is prim to cloud, prim to DC, prim to prim. Typical connectivity, fixed lines between A and B. And what's really interesting about that space is it's about $12 billion at TAM, but it's only growing at 1%. Yet the amount of data that's being consumed every year is growing by multiples more than that. Why? Price compression. That's the commodity piece of networking. And that's where a lot of our physical competitors compete. East-West traffic is the new layer. It's TAM today is about $11 billion. It's growing at 13%. And East-West traffic is where you're talking about, I need to move data right now from this cloud environment to this cloud environment or from these two cloud environments to this DC or from this DC and this cloud environment to the edge so that I can do my API development and then push it back. And I need to do it right now. I can't wait for truck rolls. I can't wait for technicians to come and do this because back to GPU math, $2,000 an hour to run a GPU. So simple math, 10-gig fixed circuit moving a petabyte of data, 222 hours. 400-gig circuit, six hours. $420,000 in cost savings. So it's not about the cost of the wave anymore. It's about the total cost of ownership. And as data gets more and more distributed, you need the ability to go anywhere to anywhere on demand. And that's what we're creating. We are the supply chain for data.

Frank Loudon, Analyst — Raymond James

All right, great. That's very helpful. So you guys had an analyst day last week. I was there, you know, some others. Maybe walk just maybe to the top two or three takeaways from that and what you were trying to deliver that message and what you want folks to know.

Chris Stansbury, CFO

Yeah, so again, the company that exists today post the sale of our consumer fiber to the home business to AT&T is now a company whose leverage is much healthier. We're at 3.8 times on our pathway to between three and three and a half. We are inflecting EBITDA this year. We will inflect revenue on the business segment in 2028 because there is still a sizable legacy business inside of the portfolio. But as we go forward, every year over the next five years, we see margin expansion because as we're delivering routes to hyperscalers, as we're growing our digital business, that is more margin accretive. We're also seeing capital intensity fall. So the combination between inflection of EBITDA, margins going from mid to high 20s to mid 30s, capital intensity on the non-hyperscale business, because they pay for that up front, it isn't capital intensive, we're going from mid 20s to mid to high teens. So free cash flow conversion is significant. Our aspirations are fully funded. We don't need to raise any capital to execute against what it is we plan to do. And, in fact, there's excess cash. So the company's in great position. We have optionality, and we're really focused on driving customer adoption.

Frank Loudon, Analyst — Raymond James

All right, great. So you've guided to EBITDA growth this year. Talk to us about what's given you confidence to hit that range. And if we look back on this a year from now, what are some of the key factors to that success?

Chris Stansbury, CFO

Yeah, a few things. So first of all, if you look at our enterprise revenue rates of decline, we're about half of what the industry is. And it's because we're already differentiated. We group our products into strategic growth items and legacy declining items. And over half of the mix is now in strategic. Over the next five years, that will grow to over 70%. And so as those rates of decline continue to improve, and they will improve again this year and then more notably a year from now, that helps EBITDA. The other thing driving EBITDA is we are, like a lot of legacy networking companies, burdened with technical debt because the only way scale was driven in the past was through consolidation. And the way consolidation was managed was by ripping out management layers. IT was never integrated. And so we've got a lot of work going on that's allowing us to do that and get rid of complexity, take out cost, and focus the company. And so last year, we targeted $250 million exiting the year. We exited at $400 million. This year, we're going to exit at $700 million in savings and $1 billion by 2027. So that's what's allowing, those two things are really what's allowing us to inflect EBITDA this year. As we look back, I think the most critical thing is obviously getting the top line, you know, back to growth. And the tell will be the adoption of our network as a service platform. That's the digital consumption layer. And what I mean by that is, historically, every service required its own infrastructure layer, its own set of ports, people in trucks driving around installing things, taking months to deploy, and hundreds of millions of dollars. The ports that we are deploying today, we call it a fabric port. They will support as many services as you can throw at it. And those services are now delivered digitally. So as soon as those ports are at a customer location, we're seeing more and more consumption of services. We're seeing less churn. It's obviously higher margin because there's no truck rolls. And it also is allowing us to pull in technology partners like Commvault who see the value of our network improving their customer experience. And so they're starting to pull through, and we will pull through their services. So it's a very scalable model in an industry that otherwise has been static. And if you look at the materials on our website from Investor Day, you'll see the scalability of those ports and how you can go from $1,000 a month to many thousands of dollars a month, depending on which services a customer buys. So those adoption metrics, number of customers, number of ports, number of services, is what we're maniacally focused on. And that'll be the tell as to how well we do.

Frank Loudon, Analyst — Raymond James

Yeah, so, you know, the networks of service and the fabric ports, you know, talk to us. You mentioned these about the analyst egg. It's a significant part of replacing the legacy revenue that's attriting. Maybe, you know, break it down for us. You know, what exactly are those products? How do they work in a simplistic manner so we can kind of conceptualize what they are?

Chris Stansbury, CFO

So at the most basic level, when that port is in place, you'll be able to procure Internet on demand, Ethernet on demand, VPN, what I would call more traditional north-south type of transactions. As you move upstream, you will be able to consume things like direct cloud on-ramps. What do I mean by that? Well, today, again, the network doesn't support AI. The way you get to cloud today is basically by taking unpaved county roads. You may be, you know, a large investment firm in Manhattan and have a 400 gig connection from Lumen into our facility at Hudson Yards. and then a 400-gig connection into a third-party carrier where you then convert that signal to Ethernet, you pay cross-connect fees, you slow down the connection, you're on the public Internet rather than a private connection, so you have not just latency, you have security issues. That's how you get to the cloud. Well, AI, think about where AI is going. By 2030, we'll have 10 times the number of data centers we have today. I already gave you the GPU math, and energy is a real issue, and so this stuff, the data is getting sent everywhere. And so you can't have latency. If you think about an AI application that already exists today, we've all already engaged with a customer avatar that is an AI agent at XYZ company. If it's taking 10 seconds for that agent to respond to what you're telling it, it's not going to be a very good experience. That's a really simple example. Go to a really complex example, an automated factory or production system where things will start crashing into each other. Latency matters. And so that's what we're building.

Frank Loudon, Analyst — Raymond James

All right, great. So I thought one of the most interesting things that I took away from the analyst day was kind of a different approach to kind of selling to sort of transform the business. In the past, the telecom industry was chasing big contracts, sort of dollar-based, whereas you guys are looking at it more as kind of selling units and so forth and then letting the revenue come later. Talk to us about that and how you guys think that's going to really transform the business that you're in.

Chris Stansbury, CFO

Yeah, again, it really is going from a very kind of static upfront sale to a dynamic sale. And it's exactly, I mean, the good news is we're not inventing anything new. This is exactly what the hyperscalers did with compute and storage, right? 20 years ago, if you were a large enterprise and you had growing data needs, you'd say, okay, I need to buy some land. I need to build a building. I need to go buy servers and storage arrays and connectivity. And all of that will be under one roof. I may have two or three locations, so I have redundancy. And boy, is this complicated, right? Now we're in a world where enterprise still has some of that on-prem capability, but they've also got stuff sitting in private cloud, public cloud. And so that's a real thing. We're going from an upfront selling motion. There will still be some upfront. There will still be base demands that enterprise has for, you know, their daily needs, but there's also going to be a need for this on-demand access where they're doing API development. They effectively need to program the network to meet their business need, where needs will go up, they'll come down, they'll go up, they'll come down. And that's what we're building, because that's what AI demands. As enterprise starts to adopt AI, they need that flexibility.

Frank Loudon, Analyst — Raymond James

All right, great. And tied in with that, you mentioned this earlier, but in the analyst day, you had a slide kind of laid out, you know, the ports and so forth and the range. So maybe talk to us a little bit about that. You know, what has to happen to kind of see that vision come through? What is incumbent on you guys? And what are the assumptions that you're making to hit those guys?

Chris Stansbury, CFO

It's a great question. And part of, I think, what got us to where we are today is our ability to be nimble. And that's exactly what we're doing as we go forward. We can't tell you today which outcome, which path enterprise is going to take, but we are ready to follow wherever that demand goes. So let me give you an example. We showed four different scenarios last week on how we can get to our aspiration of $500 to $600 million in digital revenue by 2028. uh the first two scenarios had very different numbers of ports and customers and it was really sorry it had the same number of ports and customers but it varied on the the mrr based off the amount of north south versus east west traffic okay so that that was the variable the last two examples said well how much of that is kind of what i would call much higher value-added services like a direct cloud on-ramp versus more ecosystem partners all of those get us to the digital outcome that we're targeting what's interesting is is that the range of number of customers and ports was such that if you compare that to the non copper ports we have today and just say okay how much do you have to do incrementally we're talking like a low single-digit percentage of the footprint that exists in our network today so our assumptions that we laid out for this revenue shift was really that everything we sell today growers are declines with market we deliver on schedule with the hyperscalers these thirteen billion dollars of deals we've signed we're not assuming anymore there There is more for the taking, but we're not assuming more because they're chunky. And then $500 to $600 in digital. The digital growth we exited last year. Last year we did $117 million in digital. We're assuming linear growth. That's not how it's going to play. Ultimately, there's a J-curve moment in there. We don't know where it is. Where you go from steady growth to exponential growth, our focus is on driving adoption so that we pull that J-curve in. who knows when that'll be, but you'll see a lot of conversation from us around those adoption metrics. Number of customers, number of ports, number of services. All right, great. And one

Frank Loudon, Analyst — Raymond James

other aspect of the story that's gotten a lot of attention are the PCF bills for the AI guys. You announced about $13 billion, maybe a little north of that. Where are we on those being constructed? How much of those are in play, have been built today, and what's the timeline?

Chris Stansbury, CFO

We're really just at the beginning of the delivery schedule because while we're bringing enormous value to the hyperscalers because we're shortening their time to getting these networks built, it's still going to take years. These are still three to four year kind of builds. So the first routes are starting to be delivered. You're starting to see revenue come into the model from that. There will be more this year. It's more material next year. So our business revenue declines in 25% are about 4.5%. We think that'll improve to declines of high threes this year. Next year, we're in the ones as more and more of those routes get delivered.

Frank Loudon, Analyst — Raymond James

All right, great. I'm switching gears a little bit. So a year ago, kind of the light went on that getting out of the consumer business was very possible. You're out of that business. Talk to us about the rest of the legacy business that you have there. How do you manage that for cash? You know, retiring copper, a lot of cost savings opportunities and a long cash tail on that. Talk to us about how we should think about that legacy part of the business you're going to have for a while.

Chris Stansbury, CFO

Yeah, and think of it in two buckets. So we still have between a billion and a billion and a half dollars of revenue in consumer copper. So that's voice and DSL. It's largely in rural areas. It's customers who are very sticky. They've already got a better choice in cable, and they haven't switched. You know, they aren't huge bandwidth consumers. And so that's declining kind of low double digits every year. So it's got a long runway. There's not really any value in trying to sell it because it's low single-digit multiple, and it throws off an enormous amount of cash, which is allowing us to invest in our enterprise aspirations. We are aggressively going to manage that even further for cash. There's opportunities to decommission old assets and combine things or, at some level, convert customers to an alternative like fixed wireless. And we've got multiple pathways there. So there's a great cash opportunity there. On the enterprise side, our legacy products are in decline, our largely voice, private line, as well as traditional VPN and Ethernet. Those we will also continue to manage for cash. And I think the big shift you will see from us, which is very different than the rest of the industry, is we are going to aggressively try to cannibalize ourselves. I mean, one of the things that I got asked when I came to Lumen four years ago is, you know, what's different in enterprise networking? And my comment was that the mindset was playing to not lose versus playing to win. And playing to not lose means that once you connected a customer, you never talked to them again because they might disconnect, right? So it was just madness. The reality is, if we can more aggressively convert our existing customers who are on legacy services to their digital equivalent, churn goes down, margins go up, they start buying more. It's a tremendous playbook. So we haven't built that into the model that we showed last week. The $600 million is really incremental of that. But if we can drive more consumption into those buckets, that's the best way to manage those declines and those customers. So we're going to be very focused on that.

Frank Loudon, Analyst — Raymond James

All right, great. So Wavelinks are another product you guys have that are tied in with the whole AI infrastructure, GPUs and so forth. You know, moving large data sets around, much better suited to that type of traffic. You know, what sort of the TAM do you see for Waves, and what kind of market share do you think you can get as you're a player in that market?

Chris Stansbury, CFO

It's a good question, but it's also what I would call kind of more of a traditional networking question. It's really not how we think about it. I know our competitors talk about Waves growth and price per wave and everything else. That's not how enterprise is starting to buy. they're buying based off of a solution which is I need to get a I'm in my business as fast as possible I need to move all these workloads and waves is one component of it so we see waves growing in our assumptions in the model with market but there's a but okay as we're building these these big networks for the hyperscalers we're also pulling fiber for we call rapid routes 400 gig on-demand ability to connect that's waves right we're building out our metro expansion so as the enterprise demand comes we're ready for it the reality is is that the that that waves business if you just talk about waves you're selling to the procurement department and okay this is what this guy is gonna sell it to me for what are you gonna sell it to me for when you're talking about east West traffic you're selling to the CIO because you're solving a solution at a much higher level our belief is as we do more of this east-west connectivity the cloud-to-cloud the ability to move data from anywhere to anywhere we're gonna take share in the north-south because it's just going to be easier to do business with lumen across the board so that's how we think about it and our competition is very focused on selling waves at a discount so have at it that's We're not interested in that.

Frank Loudon, Analyst — Raymond James

So it's the solution sale process versus a product kind of mindset.

Chris Stansbury, CFO

And AI is going to drive that and drive it very quickly.

Frank Loudon, Analyst — Raymond James

So even though you're kind of out of the fiber of the home business, you have pretty sizable construction projects and commitments you're taking on for the next few years. Talk to us about how you feel about labor and material availability for the next few years to meet those goals. What kind of assurances do you have from the vendors that that will be there for you when you need it?

Chris Stansbury, CFO

Yeah, we have three construction partners for these massive network builds that we're doing, kind of east, central, west. And we were actually really fortunate in timing because what happened is a number of these companies had bought equipment, had staffed up in anticipation of benefits from Bede that didn't come. And so when we went and approached these companies, two of them are public, DICOM and Maztec. They're very public about our relationship with them. We treated them as extensions of our business. They are partners. They are part of our culture. When there's a problem, we're not screaming at them. We're working together to resolve it, and it's going very, very well. So I would argue we've got kind of first seat at the table with those construction companies, given the quantum of what we're building, as well as the way we engage with them. And then on the material side, we've got a really special relationship with corning. So we got the first off-the-line capacity of their new modern fiber. You know, it wasn't that long ago that the most strands we were pulling through a conduit were a little over 400 count. We're now pulling over 800 and 1,600 count through the same pipes. And we've been able to expand that as the contract growth has come from the hyperscaler. So we're very proud and gracious that we've got that relationship with Corning, so we've been able to keep pace. So, so far, so good. We're not overly concerned about that.

Frank Loudon, Analyst — Raymond James

All right, great. Well, maybe we'll see if we have any questions from the audience. If not, I've got a couple more. But you think no? Yep, go ahead.

Chris Stansbury, CFO

Yeah, so great, great question. and I'll just repeat it, what are the metrics that we would look to that would be an indication of a J-curve moment? It's all about adoption. So if I think back, in August, we had 1,000 network-as-a-service customers. We've now got over 2,000. So we're starting to see this thing pick up pace, but it's going to be number of customers, number of ports, and number of services per port. Because the math, I mean, again, what we're bringing, to at least from a financial modeling standpoint we're bringing to enterprise networking is really what the fiber to the home companies have had for a long time which is the price times quantity math and the quantity side of that is customers ports services so those are those are the key metrics and we have been reporting those every quarter for about a year we will continue to do that

Frank Loudon, Analyst — Raymond James

all right great any others all right why don't we talk a little bit about the balance sheet chris you've done a lifetime of uh balance sheet and restructuring work in the last three years um you know talk to us about where you are on leverage how should we think about that going forward you you've got some new new debt ratings all that kind of fun stuff that are the benefits of all that work but talk to us about where where the balance sheet is and how we should think about

Chris Stansbury, CFO

going for. I'm only 28. I just look like this because of the balance sheet. No. When I came to Lumen almost four years ago, when you adjust for a divestiture that we were in the middle of, we had about $20 billion of debt, and 10 of it, almost 10 of it, was due in 2027. Our CEO, when she joined, and I shared it with her, said, that looks like a middle finger chart, and that's exactly what it looked like, staring us in the face. And so we spent a lot of time restructuring that, bought ourselves some time, and we did that because we saw the AI boom coming and we wanted to capture that market. We did not want to be in court. And so because of the $13 billion and because of the sale of the consumer fiber to the home business, which in the near term was a big user of cash, generated very little EBITDA and was burning about a billion dollars a year in CapEx, and, candidly, was a space where we didn't have a right to win because the convergence between fiber to the home and wireless is real. So all of those things combined have allowed us to do six refinancing transactions over the last year, pay down $4.8 billion of debt so that our debt started last year just under $19 billion. It's now just under $12 billion. The maturity curve has been pushed out and smoothed out, and we have reduced our interest expense by half a billion dollars a year over the last 12 months. So we have resources. We've got optionality. I'm super proud of what the team has done, and now we're starting to get into that more regular way financing where those benefits have resulted in upgrades, and I think as the business continues to perform, we'll see more upgrades, but our cost of capital has come down dramatically and we'll continue to look for opportunities in

Frank Loudon, Analyst — Raymond James

normal way financing to continue to optimize all right great so if you had to kind of identify sort of the biggest one or two misconceptions about the company today that you'd like to maybe

Chris Stansbury, CFO

set the record straight on what would those be the biggest thing is is that this isn't we are not playing by old traditional telecom rules right again I said earlier you know playing playing to win versus playing to not lose and we disrupted ourselves this company is a totally different company than it was four years ago and now we're gonna go disrupt the market and and we have a significant lead over anybody else in the space and this is the time for lumen to shine we're

Frank Loudon, Analyst — Raymond James

focused on execution great all right thank you very much Chris for being here thanks everybody we have a breakout session after this if anyone wants to

Chris Stansbury, CFO

join thank you very much