Cogent Communications Holdings, Inc. (CCOI) Earnings Call Transcript & Summary

Well, good afternoon, and welcome back to Citi's 2023 Communications Media and Entertainment Conference. For those of you I haven't met, I'm Mike Rollins, and I cover the communications services and infrastructure categories within Citi. Before we get started, I'd like to just mention that we do have disclosures available at the registration desk and on the Citi Velocity page, from which we're streaming the audio. And we're going to work to get your questions into today's discussion. Of course, we have the microphones around the room, but we also have a questions box on the website. And so if you have a question, you can enter it in there. And we're also going to continue the tradition of live audience surveys that are completely anonymous. We're just collecting up the responses, and you can access that through the placards and the information here on the table or on a box, it's going to come up on the website. So with all those details out of the way, I'd like to welcome back to the conference Dave Schaeffer, Founder Chairman and CEO of Cogent Communications. Dave, it's great to see you.

Mike, thanks for hosting me. I'd like to thank investors for taking time out of their busy day. And as always, thank Citi for a great venue. And I think rather than give any overview, focusing on Q&A is the best way to use our time, Mike.

Well, great. Well, maybe we'll just start with a high-level question of your strategic and operating priorities for the coming year, and if there are any notable changes from the past year, which given the activity for Cogent over the last few months, looking forward to your update on this.

Yes. So we do obviously have some significant work ahead of us in completing the acquisition of the Sprint GMG network. It's been nearly 17 years since Cogent has done an acquisition. And I was joking with one of my team members from actually the FirstMark acquisition. And I said, if I had told you we were going to acquire Sprint GMG 20 years ago when we acquired FirstMark, you probably would have quit in a heartbeat and thought I was delusional. It was the largest wireline network in the world at one point and a business that had $40 million -- $40 billion of revenue and 70,000 employees at peak. Now the business is much smaller today, but it's also a business that's lived with over 20 years of lack of attention that's attritted in terms of its scale and its profitability. And we have definitive agreement to acquire that business from T-Mobile. We also have kind of core Cogent, and we need to continue to reaccelerate our growth and return to the patterns that we've had in previous years. Cogent has had an organic growth rate for 17 years, averaging 10% a year. The pandemic hit that growth rate decelerated to 3% a year. It has reaccelerated slightly to about 5% last quarter on a year-over-year basis, but still half of our trend line. So we have 2 major strategic initiatives for 2023. The first of those is to complete the acquisition of Sprint GMG acquiring that business from T-Mobile and reducing the cash burn and repurposing the network to generate a positive cash flow by converting the network into a pure optical network that will carry either Internet traffic on one pair of fibers or optical transport or wavelength services and then sell off unused portions of the network. The second challenge for Cogent is to see our organic growth rate continue to improve while our NetCentric business has outperformed historical averages for much longer than we expected throughout the pandemic and continues to grow at substantially above historic rates. We are seeing finally an improvement in our corporate business, which is the larger portion of our revenue stream, 57% of revenues and rather than going from an 11% average grower declining to a negative 8% at the worst of the pandemic and now at least back to being a 1% positive growth business. We feel encouraged, but we need to do much better. And I think the market is improving, and we should be able to achieve those objectives this year.

I was going to introduce our first survey, which actually just follows on your comments on corporate. So we'll see what our audience thinks. So the first question, when will Cogent's Corporate segment revenue return to sustained positive quarterly sequential organic growth: 4Q '22, 1Q '23, 2Q '23, second half of '23 or 2024 and beyond? So we'll let this brew and see the responses come in. But before we get there, and maybe just touching on one of the first priorities you mentioned, which is Sprint acquisition. You've described a lot about how you're approaching this deal and the opportunities to get back to positive cash flow in this business to have this business grow through the projects that you were describing. Are there other underappreciated aspects of this in terms of capital and equipment that you're inheriting with this business or routes that you kind of hinted at earlier, that maybe you could sell off? Like what are some of the underappreciated aspects of this transaction?

Yes. I'm going to give you kind of 3 answers to that question, Michael. The first one is the physical fiber itself. There's 19,000 route miles of intercity fiber and 1,300 miles of metropolitan fiber. The cross-section of fiber strand count ranges from 24 to 144 fibers, depending on the city pairs in [ question ]. The fact that the fiber was buried 6 feet deep rather than a more modern network that would typically be 24 inches below grade and the fiber was encased in an armor jacket means that our network, while nearly 40 years old, has had fewer cuts than networks that are half its age. Most of the fiber in the ground today intercity was deployed in the late '90s, early 2000s. This network was deployed in the mid-1980s, but it has better loss characteristics and fewer splices than those more modern networks. The second thing that I think is underappreciated is the fact that the network was built with single-mode fiber 28 or SMF-28. So common fiber in the early '80s, it was argued that that fiber would become obsolete as transmission speeds increase. So the first commercial networks were async 565 or 0.5 gigabit a second networks. The SMF fiber actually performed very well up to about 40 gigabits per second. And when many companies were anticipating migrating to 100 gig, the fiber manufacturers Corning Lucent at the time, Sumitomo were all pushing non-zero dispersion-shifted fiber, and that was what was deployed in the late '90s and early 2000s. Well, as it turns out no equipment vendor was able to make a noncoherent 100-gig transmission system work. So all of the vendors, whether it be Ciena, Infinera, Cisco, Lucent, now Nokia, Fujitsu, NEC, all migrated to a coherent transmission technology. Coherent did not work on the non-zero dispersion-shifted fiber and actually performed better on the single-mode 28 fiber that was deployed. So what is old is new again. New networks are all being deployed with SMF-28, and that will continue to be the standard, I think, for the foreseeable future, as we're moving to 400, 800 gig and 1.2 terabit per wavelength networks. The second thing that I think is underappreciated in the Sprint network is the fact that the routes are unique to Sprint. Over 90% of the routes, no other provider shares the right of way while the endpoints are all of the major North American endpoints because this network was designed to connect all 168 ladders together to allow for tandem switching in each of the Sprint facilities. And because of that ubiquity, but the uniqueness of routing, for many of the purchasers of wavelength services, that diversity has tremendous value. I think the third asset that we are acquiring that is underappreciated are the 1,300 pieces of fee simple real estate that was owned by Sprint and now T-Mobile. Now these can range from shelters along railroad tracks that are just amplifier sites all the way up to 100,000-square-foot data centers. There are 47 significant facilities that comprise about a little over 400,000 square feet of raised floor space and have over 150 megawatts of conditioned power. These facilities are suitable for data center conversion. And today, Sprint is generating less than $1 million a year on colocation and power within their facilities. So we are going to repurpose those facilities increased Cogent's data center footprint from 54 facilities that we own today and 604,000 square feet to just under 1.1 million square feet and increase total power from 69 megawatts to over 220 megawatts. So I think it gives us a very large data center footprint, probably makes us probably the third largest operator of data centers in the country.

And are those facilities, are they really just connected by the Sprint network? Or are they -- do they have some degree of network density that creates a network effect?

So the answer is yes and no.

Okay.

Because the network was built primarily for voice initially, it always has connectivity to the Sprint backbone and typically, the 2 competitive carriers that would be in the facility are AT&T and Verizon because those 2 companies do not serve the wholesale market in any significant scale. We are in the process of building fiber from those Sprint facilities to Cogent facilities in each market. We have outlined the need to spend approximately $50 million of onetime CapEx in order to tie the networks together. By doing so, we will increase the number of endpoints where services can be sold. So today, Sprint is only interconnected to [ 24 ] carrier-neutral data centers in North America. Cogent is connected to over 800 carrier neutrals. Probably by the time the transaction closes late summer, early fall, we'll have virtually all of those 800 facilities available for the sale of services off of the Sprint backbone. That's really important because that allows us the greatest footprint for the sale of wavelengths of any carrier in North America. It's a market today that's dominated by Lumen and Zayo. And because we have a larger third-party data center footprint than any other provider, this should help us gain market share quickly.

Great. So let's see what the survey results were. So in terms of when the corporate segment revenue will return to sustained positive quarterly sequential organic growth, there's a lot of words in there, 0 for 4Q; 25% for 1Q '23; roughly 50% for the second half of '23; and 25% in 2024 and beyond. So you mentioned earlier some of your observations about how the corporate market may be improving. If you could share more thoughts of what's happening in that segment and what you're seeing in terms of pipeline sales and the types of engagement with these corporate buildings that you connect to?

So Cogent today connects to about 1,850 skyscrapers in North America. That represents over 1 billion square feet of net rentable office space. There is a total North American inventory of about 9 billion square feet. So we're in just under 12% of the market. The vacancy rate and our footprint pre-pandemic was approximately 6%. That vacancy rate skyrocketed to 18%. It improved slightly at the end of third quarter to 17.7%. We anticipate continued but slow improvement in that vacancy footprint. But I think it's probably a 4- to 5-year process before we get back to pre-pandemic levels. In most recessions, it's usually new business formation that mops up excess supply. I do believe this time it's different. We have seen studies and have specific examples of buildings that both we're in and we've looked at where buildings are being repurposed for either residential or hospitality. It's estimated that approximately 10% of the market or about 900 million square feet will be converted. That will dramatically tighten the occupancy levels. Secondly, we've seen rents fall, and what is normally the pattern is tenants take advantage of those lower rents and more desirable buildings and move up market away from suburban B and C buildings to CBD A buildings. All of the third-party studies, Newmark, CBRE, Cushman & Wakefield, JLL, all support that, and we're seeing that as well. So I think the underlying occupancy is improving. The second factor is customers desire to modernize their networks. Many customers have procrastinated or postponed network modernization for 3 or 4 years because of the uncertainty of the pandemic. Pre-pandemic, companies would design their networks where 97% of employee work days were anticipated to be in the office, 3% were anticipated to be remote. The new design standard that companies are migrating to are 60% of work days in office, 40% in a remote environment. That has profound implications. It means you need a bigger connection at your primary aggregation point; two, it means more of your compute and storage has moved off-site; three, it's more likely that you will need a second hardened aggregation point, oftentimes in a data center, which is a brand-new opportunity for Cogent. And then on the negative side, many companies are reducing the number of offices that they are occupying. We just saw the headline today from Salesforce.com where they're cutting salesforce by 10%, but they are also reducing the number of offices. So rather than shrink the offices they remain in, they're just eliminating noncore locations. That, I think, is the more common pattern. And I think when companies 5 years ago, thought about office-to-office connectivity, they needed to have a highly secured VPN usually based on MPLS. Today, because they've now lived with an ad-hoc VPN for remote employees for the past 2 years using the Internet, I think most companies are now willing to accept an Internet-based VPN or just naked Internet connectivity for their office-to-office communications. So there's both pluses and minuses for Cogent, but when we package these things together, what we have seen is growth in our sales force. It was actually the fastest growth in a single quarter in the company's history in the third quarter. We've regained about half of the net number of salespeople we lost during the pandemic. We expect to continue to see improvement in fourth quarter and first quarter. And hopefully, by midyear, we'll be back at pre-pandemic sales force levels. We'll continue to grow from there in the latter part of the year, returning to that annualized 7% to 10% sales force growth number. Secondly, we have seen the number of proposals issued per rep accelerating. I think that means customers are now ready to make decisions. And then the third thing is we've seen the gestation period of those proposals shrink, so we're seeing a more rapid decision cycle. All in all, we're seeing good recovery. I'll comment now on the survey. We did return to positive growth in the fourth quarter. We have been very clear to investors, and I'm glad they listened to what we said, that we're uncertain if that's permanent, and we're just straight up from here. I think things feel pretty good. I think we are going to continue to see improvements. But after 2 plus -- 2.5 years of pandemic, I think we're realistic that the path is going to take longer and be bumpier than we expected. I think kind of the modal response at midyear for sustained growth is both realistic and probably a little bit conservative.

So just to make sure we got all the details correct, you mentioned that you returned to positive growth in the fourth quarter?

We did.

So...

Fourth quarter of last...

'22, you were positive.

Well. We were positive in third quarter.