Prepared remarks
Good morning, and welcome to today's conference call to discuss Uniti's Second Quarter 2026 Earnings Results. My name is Jonathan, and I will be your operator for today. Today's call is being recorded, and a webcast will be available on the company's Investor Relations website, investor.unity.com, beginning today and will remain available for 365 days. It is now my pleasure to introduce Bill DiTullio, Uniti's Senior Vice President of Investor Relations and Treasury. Please begin.
Thanks, Jonathan. Good morning, everyone, and thank you for joining today's conference call to discuss Uniti's second quarter 2026 results. Speaking on the call today will be Kenny Gunderman, our CEO; and Paul Bullington, Uniti's CFO. John Harrobin, President of Kinetic, will also be joining us this morning during Q&A. Before we get started, I'd like to quickly cover our safe harbor statement. Please note that today's remarks may contain forward-looking statements. These statements include, but are not limited to, statements regarding Uniti's fiber build strategy, the business' growth potential, our 2026 outlook and other statements that are not historical facts. Numerous factors could cause actual results to differ materially from those described in the forward-looking statements. For more information on those factors, please see the section titled Safe Harbor Statement in the accompanying presentation and the Risk Factors sections in our filings with the United States Securities and Exchange Commission. With that, I would now like to turn the call over to Kenny.
Thank you, Bill. Good morning, everyone, and thank you for joining. Uniti posted another terrific quarter of results as we continue to execute well on our strategy as the premier insurgent fiber provider. We're consistently growing wholesale, enterprise and consumer fiber revenue at 10% to 20% with an insurgent share taker mentality. Our strategy of being first with fiber to metro and wholesale markets, along with a mission-critical and future-proof technology positions us for many years of predictable execution and value creation for our shareholders. Demand for fiber is continuing to accelerate, and hyperscalers and neo-clouds are leading the way. We posted a record quarter of new bookings at Fiber Infrastructure, exceeding our previous record by almost 30%. Demand was strong across virtually all categories, reinforcing the benefit of our robust diversified customer base, especially at Uniti Wholesale. The use of Agentic AI is now upon us and growing. As has proven typical with AI adoption, both the timeline and TAM appear ahead of our prior expectations. Near-term Agentic AI use cases with measurable ROIs such as customer service agents, cybersecurity threat hunting and mitigation and automation of corporate functions are only a precursor to more bandwidth-intensive functions like physical robotics, autonomous driving, personal AI agents, scientific research and many others. Uniti is building the infrastructure to enable this growth and benefit our shareholders. This growth started and continues with dark fiber networks in Tier 2 and 3 markets with fiber strand counts of 50 to 100 times prior levels and is now evolving to lit Waves packages at the same 50 to 100 times prior levels. These wholesale investments by our hyperscaler and neo-cloud customers are a precursor to enabling greater Agentic AI usage by our enterprise and residential customers, and we've seen estimates of dramatic multiplier effects on broadband usage as a result. Distribution of inference to the edge requires low latency and high-bandwidth connectivity, highlighting the significant benefit of our 5 million future-proof connected endpoints at Uniti, including fiber-to-the-home, on-net buildings and data centers, towers and small cells. Fixed wireless, LEO and even cable remain somewhat competitive today at the edge, but over time, as usage, reliability and latency become more mission-critical, customers large and small will demand fiber at the edge, presenting us an opportunity to win back share from these other technologies. In short, Uniti is well positioned strategically. We have the right assets, plan and team in place going forward. Now we have to execute. Our results indicate we're doing just that. Total fiber revenue grew 10% year-over-year and fiber revenue at Fiber Infrastructure grew 6%, right in line with our expectations and demonstrating the robust demand we're seeing. In addition to the record bookings at Fiber Infrastructure, we had the strongest quarter ever of net adds at Kinetic and, just as importantly, the number of new fiber homes constructed. As we've been foreshadowing for some time, we really hit our stride for the fiber build in the second quarter, and we fully expect that momentum to continue. Our footprint at Kinetic has substantial opportunities to build first, and there are increasingly fewer skilled footprints like ours in the country. We're moving in haste to take advantage of this benefit. To that end, as highlighted on Slide 5, our priorities have not changed for the full year. Building off the momentum in the Kinetic build engine, we're now raising our expectations on new fiber homes constructed this year to 475,000 to 525,000. We also continue to lean into this generational opportunity at Fiber Infrastructure, which I'll comment further on in a moment. At Uniti Solutions, we're seeing growing success in cross-selling products into our on-net fiber base at Uniti Fiber and Kinetic. Today, we estimate our managed services attachment rate to be only around 6%, excluding voice, but we're growing that materially as during the quarter, about 16% of new bookings came with the managed services product. We believe selling more products to on-net customers will not only lead to greater sales, but enhanced margins and more loyal customers. 2026 is an important inflection year for Uniti and the terrific industry tailwinds we just discussed are fueling that pivot. We previously committed to some key milestones as highlighted on Slide 6, and we're demonstrating progress towards our goals. Convergence offerings and new initiatives from big cable are having an expected impact on ARPU across the industry, and we've seen some near-term pressure. However, we expect ARPU to stabilize industry-wide and at Kinetic, especially given the favorable competitive dynamics in our markets and the upsell opportunities that exist in our base. The current industry dynamics suggest it has never been more important to build fiber first or early, and that's exactly what we're doing now at an accelerated pace. Slide 7 shows that we're well on our way to 3.5 million homes passed with fiber and 1.25 million fiber subs, and we're closer to 90% of our revenue coming from our core business. We remain focused on operational excellence, customer obsession and together with executing on our strategy of building fiber into unique locations, progress on our key KPIs will remain up and to the right. Driving lower churn is critically important to driving higher customer lifetime value, and we are very pleased with our plan and progress there as highlighted on Slide 8. Our fiber churn at Kinetic this quarter is the best second quarter churn we've ever seen, and we believe that with the various actions we've taken to date plus future planned actions will bring Kinetic fiber churn down to industry-leading levels. Managing churn effectively is a team effort, and we've actually made it a company-wide metric for our incentive compensation plan as a result. Turning to Fiber Infrastructure on Slide 9. Having a previously built robust fiber network in Tier 2 and Tier 3 markets is presenting us an opportunity to capture meaningful share of the generational demand as evidenced by a record quarter of new bookings. Importantly, over 50% of the new bookings this quarter were Waves or lit capacity as opposed to dark fiber, reinforcing the pivot from the build cycle to more lease-up and inference. As we previously pointed out, we expect approximately $1.5 billion of revenue in the next few years as we build profitably to enable AI learning. However, the truly exciting opportunity for Uniti is the approximately $500 million of recurring revenue that will sit on top of those new builds, and we're starting to realize that recurring revenue now. We've also said that we expect that broadband usage to come from a diversified base of customers. And during the quarter, we demonstrated a healthy mix of demand across our key customer segments, including about 20% from neo-clouds, 18% from superscalers, 10% from hyperscalers and 6% from fiber-to-the-home providers. A special thank you to Mike Friloux, Greg Ortel and the terrific sales leaders at Uniti Wholesale and our world-class network and service delivery teams under the leadership of Eric Daniels and Cathy DeLaGarza for following through for our customers. With that said, our funnel of opportunities is even more exciting. So we expect continued positive momentum. As a reminder, while quarterly bookings in wholesale will always be lumpy given the nature of the customer base, the general trend over the next few years should be up and to the right. Slide 10 illustrates we're building our fiber network profitably and strategically. Although we're building some attractive new greenfield routes for hyperscalers, close to 80% of our hyperscaler business actually includes selling all or at least partial existing infrastructure, leveraging heavily the previously built network. Make no mistake, this is a big advantage for Uniti from a cost and time to deploy perspective as evidenced by our blended anchor lease-up cash yields of 37%, the highest we've ever seen. Turning to Slide 11. During the quarter, we sold more Waves than ever. And as a result, Waves were the single biggest product contributing to our record bookings. As a reminder, we're not enabling Waves capability all across the country. We're focusing on routes that are unique to Uniti that give us a competitive advantage. Our pre-existing routes are enhanced by the unique build cycle that we're currently undertaking for the hyperscalers as we strategically expand our connectivity among unique markets. Highlights in the quarter were a 20-terabit Waves package connecting a neo-cloud from a data center in a Tier 2 market back to a large metro and an 18-terabit Waves package sold to a superscaler out of another major data center in a Tier 2 market. To put this in perspective, that's ninety-six 400-gig Waves sold as lease-up in one quarter to just two customers over fiber recently built into new data centers. And there's more to come as our current Waves funnel represents approximately 1.3 petabytes of traffic and the vast majority of this funnel consists of relatively new customers. As we see it today, the demand appears staggering. With that, I'll turn the call to Paul.
Thank you, Kenny. Starting on Slide 13, I'd like to review the key second quarter highlights for both Kinetic and our Fiber Infrastructure segment. Results for the quarter were once again strong, and we continue to make significant progress across several fronts. Starting with Kinetic, we expanded our fiber network to pass an additional 141,000 homes with fiber, our highest level of new passings on record, ending the quarter with approximately 2.1 million homes passed with fiber. Kinetic also added 38,000 net new fiber subscribers during the second quarter, ending the quarter with 603,000 total fiber subscribers. As Kenny mentioned earlier, Fiber Net Adds for the quarter were the highest on record, and total Kinetic fiber subscribers grew 25% from the prior year period. Kinetic Consumer Fiber revenue grew 19% year-over-year during the quarter. This growth is being driven by strong adoption of our fiber-to-the-home product, bolstered by the performance of the various marketing initiatives at Kinetic that target both our newer and more seasoned cohorts. At Fiber Infrastructure, we recorded consolidated bookings MRR of approximately $2.2 million, our highest level on record and an almost 30% increase from the previous record level. These record levels continue to be driven by the robust demand we are seeing for both dark and lit fiber solutions from hyperscaler and neo-cloud superscaler customers. Slide 14 highlights the sustained momentum we are seeing within Kinetic Fiber. We achieved fiber penetration of 29% during the quarter, which is up 90 basis points year-over-year. We also now passed 46% of our Kinetic consumer footprint with fiber. Consumer Fiber ARPU during the quarter was impacted by a number of factors, including new customer volume, the timing of rate plan adjustments and customer retention initiatives. For the remainder of the year, we expect Consumer Fiber ARPU to decrease low single digits year-over-year in the third quarter, but should stabilize in the fourth quarter with an expected increase of low single digits year-over-year. Turning to Slide 15. The continued strong improvement in our cohort fiber penetration is being driven by highly targeted marketing, customer experience and customer retention initiatives being deployed by the Kinetic team. Penetration levels in our year 1 2025 cohort are now exceeding year 2 penetration rates in the prior year cohort and year 3 penetration rates in our older cohorts. We expect to maintain or improve this trajectory going forward. And given our current trajectory, we remain confident that our 40% terminal penetration target is very achievable and perhaps conservative. Slide 16 lays out our key targets for Kinetic in 2026. Given the accelerated pace of our fiber build, we are increasing our incremental homes passed with fiber target by 25,000 and now expect to reach 2.33 million to 2.38 million homes passed with fiber by the end of this year. This would bring fiber coverage within the Kinetic footprint to over 50%, a significant milestone in our goal to reach 3.5 million homes by the end of 2029. We also expect to end the year with between 675,000 and 700,000 Fiber subs and realized $635 million to $655 million of Consumer Fiber revenue in 2026, an increase of roughly 25% to 30% from the prior year. Slide 17 provides a pro forma view of Uniti's consolidated results for the second quarter. Consolidated pro forma revenue and adjusted EBITDA were down 5% and 10% year-over-year, respectively, during the quarter, primarily driven by the continued declines at Uniti Solutions and in legacy copper and TDM services. However, we continue to see strong growth in the strategic parts of our business. Kinetic Fiber-based revenue, inclusive of consumer business and wholesale services, grew 12% year-over-year. As we continue to execute on and accelerate our Fiber overbuild plan, Fiber services at Kinetic will deliver consistent strong growth quarter-over-quarter. At Fiber Infrastructure, revenue and adjusted EBITDA grew 10% and 20%, respectively, year-over-year. In addition to the information provided in our earnings materials, we have also included supplemental pro forma financial information on our investor website. Slide 18 demonstrates that the growth in each of our core Fiber lines of business has been very strong, and we expect that growth to continue given the superior nature of Fiber as a Service. With this pace of growth, we continue to expect Fiber to overtake legacy services as the majority of our revenue by the end of 2026. Please turn to Slide 19, and I'll now cover our updated full year 2026 outlook for the combined company. Beginning with Kinetic, we continue to expect revenues and contribution margin to be $2.145 billion and $905 million, respectively, at the midpoint. Given the high number of Fiber premises now expected to be constructed in 2026 and the incremental capital expected to be spent later this year to support our accelerated build plan for 2027, we now expect to deploy approximately $1.27 billion of net CapEx at the midpoint of our guidance, a $100 million increase from our previous range. At Fiber Infrastructure, we expect revenues and contribution margin to be $1 billion and $575 million, respectively, at the midpoint for full year 2026. The increase from our prior guidance range reflects the strong hyperscale and AI activity we saw in the second quarter. As a reminder, we expect the revenue from large sales-type lease Dark Fiber deals to be lumpy and to come in unevenly during 2026. As such, we have further adjusted our third quarter and fourth quarter guidance to account for the timing of these deals in the second half of the year. While we still expect to see a significant portion of those deals hit later this year in the fourth quarter, as we previously communicated, there is a possibility that some of those deals could slip into early 2027. Our outlook for net CapEx at Fiber Infrastructure this year remains $140 million at the midpoint of our guidance and represents a capital intensity of approximately 14%. Please also note that as has always been our practice, our net CapEx reporting offsets our gross CapEx by upfront payments received in an IRU arrangement as the cash received will offset a significant portion of the CapEx related to those deals. Turning to Uniti Solutions. We expect revenues and contribution margin of $700 million and $320 million at the midpoint. The increase in our contribution margin guidance is due to higher-margin strategic services sold during the quarter. As we have mentioned several times before, while Uniti Solutions is not core to our go-forward Fiber Infrastructure strategy, it does generate meaningful predictable cash flow. While we expect revenue and EBITDA to continue to decline at a mid-teens pace year-over-year over the next few years, a crucial part of our strategy is to retain the most profitable portion of this business while winding down low-value legacy and TDM services. Altogether, we expect consolidated revenue and adjusted EBITDA of approximately $3.655 billion and $1.475 billion at the midpoint of our 2026 outlook with consolidated net CapEx of about $1.525 billion. Finally, I'd like to provide some brief comments on our capital structure. Since announcing our agreement to merge with Windstream, we have successfully executed on a series of planned actions that were systematically implemented to extend our debt maturities, lower our overall cost of debt, establish access to new debt markets, optimize our mix of secured and unsecured debt and drive meaningful interest expense savings. As Slide 20 highlights, partially as a result of these actions, the blended yields on our debt have improved significantly, falling an impressive 600 basis points over the past three years from around 12.5% in February of 2023 to around 6.5% today on a blended basis. In recent quarters, we have discussed the attractive nature of the ABS debt market and the likelihood that ABS would play a significant and growing role in our capital structure going forward. To that end, we recently completed our second ABS transaction at Kinetic that will not only substantially help fund our Fiber build for the next year, but also allows us to potentially pay down up to $500 million of secured debt through the asset sale offers that are currently ongoing. As I've said many times previously, we intend to be balanced in our approach to raising ABS financing and these associated offers to pay down secured debt with ABS proceeds is in keeping with that balanced approach. While ABS will be an important part of our strategy to fund the strategic investments we are making in our business, it's not the only source of capital we have at our disposal. For example, as has been our practice at Uniti, we are constantly evaluating our portfolio of assets for optimization. Optimization opportunities could include assets that are underutilized or fallow, assets that are outside of our prioritized footprint or assets for which we can receive premium valuation multiples. As we mentioned last quarter, we believe there are $500 million to $1 billion of noncore assets that we could monetize. It's important to also note that the monetization of these assets would have a negligible effect on our adjusted EBITDA as many of them are underutilized today and currently produce minimal to no cash flow for the business. As Slide 21 shows between Excess Fiber, noncore and nonclustered assets in operations, such as select non-clustered Kinetic and non-Southeast Fiber Infrastructure markets as well as spectrum and other real estate assets, we believe the opportunity exists to generate material proceeds over the next 12 to 36 months, and we are making progress on several potential transactions. With that, we'd be happy to take your questions.
Questions and answers
Our first question comes from Gregory Williams of TD Cowen.
It sounds like business is booming on the Fiber Infrastructure side. So my questions are more on the Kinetic side, actually. One is just on the consumer Fiber ARPU. As you guys noted, it was down 2.6% and you gave us some guidance for the third quarter and fourth quarter, saying that you have new customers, timing of rate plan adjustments and retention. Can you just help provide more color to the ARPU trajectory? We're seeing ARPU pressures everywhere, whether it's cable repricing, fixed wireless and now Starlink in the mix. How do you think about the ARPU curve longer term as you think about maybe 2% or 3% growth from previous messaging? Second question is on Kinetic CapEx. You did raise it by $100 million, and you did raise the Fiber home deployment by $25,000. So it implies cost per home pass might be going up? Or is this just because you're front-end loading some of the costs or warehousing equipment? Or is it actually indeed cost per Home Passed going up from rising equipment costs?
Yes. Greg, this is John. I'll take that. And I'll start with the second one. On equipment costs, we've seen all the stories that you've seen. We believe that there's going to be a slight increase in our cost of fiber in the outer years, 2027 and beyond. And as a result, I think when you look at the range that we've guided to in terms of cost per passing, we're going to come in at the upper end of that range. So I would plan for that, and that's what we are planning for as well. We're not going to be impacted by the CPE memory charge issues for a variety of reasons, including our long-term contracts and volume and the fact that we can balance different SKUs in order to hit the targeted CPE cost number that we planned for. So I think we're good on the CPE side and expect an uptick a little bit in the fiber material side beginning in mid-2027 and beyond. Relative to the ARPU trends, as Paul said, we expect a slight decrease year-over-year next quarter and then return to positive in the fourth quarter. And on an annual basis, our forecast still shows 2% to 3% ARPU accretion in 2027 and beyond. All the reasons Paul mentioned are accurate. When he talks about cost of retention, that is in response to the competitive promotions that cable started with and the telcos matched on. And we've seen that. We're fortunate in a sense that we only compete with big cable in less than 60% of our fiber territory. That compares to mid-80s to low 90s for many peers. So we're not as impacted. But make no mistake, 60% is still a number, and we are in a growth business. And like Kenny said, we're not going to chase unprofitable growth, but we're not going to cede the market either. So we're being smart about it. We have regional pricing tiers and cohorts that we adjust based on our performance, the strength of that market and competitor pricing. We track it and monitor it, and we adjust to be rational yet maintain our growth. As you can see, we haven't been slowed down by growth on the top line or churn improvements either. So I think we'll see it play out. We said churn would come in where it did at an improved guidance last quarter, and we expect that to improve. And like Paul said, we're going to return back to that 2% to 3% ARPU growth beginning in the fourth quarter.
Greg, let me just add a couple of things to what John said to reemphasize on the CapEx, to make sure we're clear there. You can't really take the increase in capital that we're guiding to and apply it directly to just those 25,000 additional homes we're guiding to. As mentioned in my comments, a lot of that CapEx is a pull forward of CapEx to accelerate 2027 growth. So those households that we're investing in at the end of 2026 will be households that are passing in 2027. As we've talked about before, these builds tend to be more front-loaded from a CapEx standpoint with design and permitting and then construction starts on homes to be passed in future periods. So we're confident in our ability to hit the cost-to-pass range that we've put out prior, even though, as John said, there might be forces that are driving us a little bit more towards the higher end of that range, but that range still holds.
And our next question comes from the line of Richard Choe from JPMorgan.
I wanted to ask about the demand funnel or pipeline you see from the three different categories you called out: neo-cloud, superscalers and hyperscalers. Is there any differentiation in what type of projects they're looking at or you're potentially doing for them?
Richard, I think you trailed off a little bit at the end there, but I think I got the gist of your question, which is the distribution of our funnel across the key customer segments and the types of products. Fundamentally, the funnel is very, very strong across all of those customer segments. I think our distribution of bookings in the quarter is actually a pretty good representation of the funnel. Of course, it ebbs and flows because any time you've got a funnel that just grows and grows, that means you're not selling enough. So we need the funnel to grow, then retract and grow again. Fundamentally, each of those segments is strong. As we've been saying now for really two years, and we're really just parroting what the hyperscalers and neo-clouds have been saying, there is a constraint on compute supply in the industry and therefore on the infrastructure to enable that compute. That has been a consistent message. It's been a consistent message among that customer segment that they need to invest to stay ahead of the demand, and they're consistently surprised by the amount of demand that's chewing up the capacity. They're struggling to stay ahead of it. We're at the tip of the spear of that, and we see the same amount of demand across all those segments. I think the hyperscalers are much more in the vein of building new fiber and whether it's greenfield or interconnecting routes or overpulling routes and more in the vein of dark fiber as a product. Think about them as more focused on enabling AI and building the learning models. That's a large percentage of the $1.5 billion of revenue that we've talked about over the next several years as being the build cycle. They're generally the anchor customers in that build cycle. We've talked about the emergence of the neo-clouds and the superscalers and other customer categories that will use those learning models and that fiber for inference and for reselling compute capacity. We're starting to see that demand sooner than we expected. We're now using words like petabytes to capture the amount of demand. I'd say the majority of the funnel is really coming from the neo-cloud group. These multi-terabit Waves packages are more of the norm than the exception. We talked about two of them this quarter. They're right down the fairway of lease-up on greenfield builds that we've had over the past couple of years, driving really terrific margins in a capital-efficient way, and we've got a lot more of those coming. So the short answer is demand continues to accelerate, and we're starting to see the neo-clouds and the superscalers as the higher growth component of that, and they're taking more Waves and lit capacity versus dark fiber than the hyperscalers.
Yes. Sorry for fading out there. In terms of the hyperscaler lease-up IRR, the 15% for spade, is that just timing? Or is that a mixture of timing, volume and maybe pricing?
It's a combination of all of the above. As you know, Richard, many of these hyperscaler deals come with very large NRCs. We used to measure our anchor deals on a cash flow yield basis, but we can't do that with hyperscalers because it's not measurable in many cases since the cash upfront largely offsets the initial capital. When you think about both the anchor and the lease-up, in many cases it's the same hyperscaler taking the original anchor plus the lease-up. So it's definitely a function of timing and the amount of fiber they're taking. But in all cases, the returns are very attractive, and we think they're only going to get more attractive because, as I said earlier, the build cycle is the precursor to where the real demand and the real upside for us is, and that's on the recurring revenue.
Our next question comes from the line of Frank Louthan from Raymond James.
Great. On the lease-ups going forward, how do those work? Are there any restrictions on what you can charge new tenants or any most-favored-nation or degradation clauses for the anchor tenants? And then on the wavelengths, do you offer waves at any point to any endpoint a customer wants? Or are your Waves between certain POPs or data centers?
Frank, we have very few limitations, restrictions or unique terms with respect to what we can or can't do with the fiber that we're building. That's by conscious design. Many of our customers know that we're building fiber as a shared infrastructure asset, which means in order to get acceptable returns for ourselves, we've got to be able to lease that fiber up to other customers, and that's an understanding from the very beginning. I could list on one hand the number of unique deal terms that we might have. So it's absolutely more of the exception than the rule. As we look forward, while we're currently building fiber and we've built fiber, there are very few limitations on what we're going to be able to do with that fiber going forward. There's probably more limitations on what our customers can do with what we sell to them than the other way around. That's something we've covered in the past and has always been a part of our business model and will be on a go-forward basis. With respect to Waves, we couldn't be happier with our Waves strategy and the progress we're making. We've talked for several quarters about how the Waves market is growing roughly 10% a year, which I think is conservative. We have probably less than 5% market share in that market, so we don't have a big embedded base of Waves that are susceptible to competition. We've also talked about how we're selectively lighting routes around the country where we've got a right to win on network quality and uniqueness and customer service capabilities as opposed to on price, where many Tier 1 routes have multiple competitors and they compete largely on price. We don't want to be in that game. We want to be in the Waves business where we've got unique routes and can compete based on network quality. This hyperscaler build has helped us accelerate that strategy because we're building a lot of new fiber in Tier 2 and Tier 3 markets, in some cases connecting back to Tier 1 markets. You have an embedded Waves opportunity that we're now taking advantage of. The two big Waves packages we highlighted this quarter are exactly that. This is dark fiber that was recently built and connects data centers. These are clearly not single-use data centers, as evidenced by the fact we had an anchor customer and now we're leasing up to Waves customers. That's the strategy. It always has been, but I think this AI-fueled hyperscaler build is accelerating the Waves product for us, and we're very excited about it.
And our next question comes from the line of Matthew Griffiths from BofA.
So you mentioned in the prepared remarks that some Fiber Infrastructure revenue could slip maybe from 4Q into 2027. Is the driver of that just data center construction completion timing? Or what other factors could be behind that? And then I'm curious on the terms of the Waves and Dark Fiber deals. Can you comment on the number of years you're signing up for and if that's moving in a certain direction?
Matthew, yes. On the large hyperscaler deals for the year, we've tried to be transparent not only coming into this year, but foreshadowing next year and beyond that these big deals are ones that will move the needle from a quarter-to-quarter basis and even from a year-to-year basis because they're big deals. In the first half of the year, we've actually either been able to pull in deals sooner than expected or they're on schedule, a pleasant surprise to the market. That's part of what's helped us outperform expectations for the first couple of quarters. We're foreshadowing that the third quarter will be down because we don't actually think there's going to be a material amount of larger deals in the third quarter, and therefore most of them will be in the fourth quarter. That's the math and the expectation on current timing. We flag it because we don't want people to be surprised if something does slip from 2026 into 2027 and move the numbers based on paper, but we're really talking about a matter of weeks or maybe a month or so as opposed to quarters or years. So it's not a reflection of the quality of the book of business. It's just a reality that when you're building hundreds of miles of new fiber and building 864-strand count fiber with multiple conduits or even 1,728 fiber with multiple conduits, and you're building through multiple permitting authorities and using multiple contractors or subcontractors in addition to our internal crews, there's always a risk of slippage even when you're largely in control of your destiny. I wouldn't flag any particular issue around the builds or deployments that we're concerned about. We're flagging it because of the materiality of the numbers and the reality of building large amounts of fiber. With respect to your second question, we've talked a lot about the greenfield builds and the overpulls for the hyperscalers as being largely 10- to 20-year deals, and that has continued to be the case. We love those deals. We love dark fiber. Locking in that long-term revenue and anchor customers is terrific. But when you start to get into the lease-up phase, you're generally going to be into shorter-term contracts by definition. And with this new class of customers—the neo-clouds and superscalers—we're very focused on the credit quality of the customer and making sure that when we paper these opportunities or do our analysis to assess customers, we're focused on their credit quality and put ourselves in their shoes about what they're using the fiber for. We're pushing for longer-term deals; they're pushing for shorter-term deals; and we're landing at a place where we're very comfortable from a credit quality perspective. Generally, lit deals and Waves deals are in the three- to five-year range. I'd say we're closer to the three-year range in many cases; some are shorter, some are longer. But we're very comfortable with the terms we're signing and with the credit quality of the customers so far.
This does conclude the question-and-answer session of today's program as well as today's program. Thank you, ladies and gentlemen, for your participation. You may now disconnect. Good day.