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

For those of you who don't know me, I'm Matt Niknam, comm infrastructure analyst here at Deutsche Bank, and we are pleased to be joined by a man who really needs no introduction, but I'll give an introduction anyway. Cogent Communications CEO, Dave Schaeffer. Dave welcome back.

I'd like to thank Deutsche Bank for a great venue, and I'd like to thank investors for taking time to hear a little bit about Cogent.

We've got lots to talk about. Maybe just to start, you recently reported fourth quarter results closed out 2022. It sounds like you've got a very busy 2023 ahead. Maybe just to start, we can talk about some of your top strategic goals and priorities for the year?

I think our #1 goal is to complete successfully the closure of our acquisition with Sprint GMG or the T-Mobile Wireline business. To begin some of the cost reductions and increases in profitability in that business as well as to begin the market wavelength services across that network and colocation services within the footprint we're acquiring. I think our second key objective is to see continued reacceleration in Cogent's organic growth rate. The pandemic was tough on us. We went into it with a 10% working on a top line growth rate for the past 17 years. And because of the impact on our corporate business, the growth rate decelerated to as low as a couple of percent. Last quarter, we backed up to 5.7% year-over-year growth. We're more than halfway back. But getting that business to continue to improve, probably our second strategic objective.

Well, we usually talk about M&A at the end of the discussion, but we'll kind of flip it and start an M&A just given the Sprint GMG deal. Maybe just to start, if you can refresh us on the rationale and benefits from the Sprint GMG acquisition and talk about some of the synergies that the asset brings classic Cogent?

What we are buying is the Sprint Global Markets business that is part of 40-year-old business. At peak, that business had $40 billion of revenue, 70,000 employees and $16 billion of EBITDA. Now that's 20 years ago and that's history. That asset has really been ignored for the past 20 years. It has shrunk to $560 million in revenue, negative $300 million in EBITDA and approximately 1,320 employees. There are 1,400 customers buying a total of 28 products. Only 4 of those 28 products are actually gross margin negative. We are acquiring a business that is both declining and burning cash. We're being paid by T-Mobile through the sale of transit services to T-Mobile, which at this point, they may not use $700 million, $350 million in the first 12 months and then the next $350 million spread over the subsequent 42 months. Total cash payments of $700 million with no cost of goods sold and 100% contribution margin. In acquiring that business, we acquired 19,000 route miles of intercity fiber, an additional 1,100 miles of metropolitan fiber and we're acquiring 1.3 million square feet of technical space that is suitable for data center use. There's actually a total of 1.6 million square feet being required. All fee simple owned. Some of the facilities are not suitable for conversion to data centers. Within the acquisition, we actually have 2 very different objectives. The first objective is to jettison the unprofitable revenues, retain the large enterprise customer relationships that have been in place for nearly 30 years and take that business from $560 million to $450 million, modernize the products suite by converting the MPLS customer base to VPLS, increasing and MPLS customer base. And probably shrink the total number of customers from 1,400 to 1,200. In doing that, we should be able to bring traffic on-net. Today, at this business, 93% of services are delivered using third-party networks. Only 7% is on that. Cogent by contrast is 75% on that 25% off debt. We believe we can convert that stable business to approximately a 20% margin business. $450 million of revenues, $90 million of EBITDA and about $30 million of CapEx to run it. We will probably burn a little over half of the $700 million that we're getting from T-Mobile to accomplish that restructure. If that was the only purpose this acquisition would not make sense. The real value and what we're acquiring is the repurposing of the physical assets that were built at a capital cost of nearly $20 billion. Those assets today are basically fallow. There's 1.3 million square feet of data center space, 160 megawatts of power, 37,500 racks empty. There's also a fiber network with average cross section of about 40 fibres that is basically unused today. There's a small amount of test wavelength business and a very de minimis amount of IP services on that network. What we will do is take one pair of fibers and migrate the Cogent IP business onto that network. To give you a sense of scale, their IP network today is transmitting approximately 10 petabytes a day. Our network is closer to 1.1 exabytes a day or about 105x as large as their traffic. We will then free up a pair of fibers to sell optical transport services wave length. This is a market that Cogent has not participated in historically. It is a $2 billion addressable market. We will have more distinct advantages in selling into that market. That $2 billion market is dominated today by Lumen and Zayo. We will have the advantage of having completely unique routes. Over 90% of the Sprint network is long right of way that no one else uses. It was primarily on a railroad right of way. Secondly, we will connect that long-haul network to the existing Cogent metro footprint, and that would give us connectivity to 800 carrier-neutral data centers in North America, where we could sell wavelengths that's over twice as many as any of our competitors. We will have unique routes. We will have more ubiquity. The third advantage is our sales force. We have today a sales force of reps that focus on service providers, whether they be access networks or content generators. And that represents the majority of that wavelength market. We already have existing relationships with over 3/4 of those customers buying transit services from us, and we will be able to go after those customers aggressively. The final advantage that we'll bring is the ability to provision the services quickly. Because of the way our network is architected, we're working to meet the same kind of provisioning windows as we do or transit. But just to remind investors, Cogent 17 years ago had 0% market share in the wholesale transit market globally. Today, we're at 24%, the #2 provider in the world. Because we can leverage that market position, we think we can capture a quarter of the wavelength market within 7 years. Finally, because we do have a negative cost basis in the network, we're being paid to acquire it. We will price as aggressively as necessary to capture market share. When we entered the transit market, our strategy was controversial. We undercut any other provider by 50%. And it allowed us to gain market share, to grow revenues in a flat market and expand margins. That strategy is still continuing, and we are continuing to gain share in that transit market. In the wavelength market, it's not clear to us that we'll have to be as aggressive on price, but we'll prepare to if necessary. The transaction is dramatically derisked for Cogent because of the $700 million cash payment from T-Mobile.

And as we think about just extended TAM, you touched on wavelength and the $2 billion opportunity, 25% market share you're targeting. You also mentioned a lot of data center space you're getting access to. Anything in terms of additional colocation revenue potential, just in terms of other maybe areas Cogent may not be as active in today that the deal may present to you?

Cogent today prior to the acquisition, has 54 data centers, approximately 8,000 racks, 69 megawatts of power, and that business generates roughly $20 million a year for Cogent. We will end up taking that rack count to over 45,000. We will almost quadruple the amount of power that we have. And these facilities that we're acquiring are actually owned. There's no rental payments associated with them. We think we can add pretty quickly about $30 million of run rate in the colocation product. The only direct cost there is the incremental power, which is typically built directly to the customer, and this will be additive to the wavelength opportunity.

What's the latest on deal-close timing?

We were initially expecting the deal not to close until the latter half of the year. About 10 days ago, we received our final regulatory approval from the California Public Service Commission. We've received regulatory approval in 23 international markets, all 50 states and the FCC. The 2 items that are left for the deal to close are the completion of the carve-out audit by T-Mobile. They are busily working on it. And I know Peter is going to be here tomorrow, and you can ask him. But we anticipate that being done over the next month. The second thing will be the completion of the transition services agreement while we negotiated an initial framework along with the transaction negotiations, we really didn't fill in the details. We're now beginning that process. There's over 220 different areas that need to be covered. And they're not all just for Cogent's benefit. Many of them actually benefit T-Mobile long term. There's going to be wireless equipment that are located in buildings that we're acquiring. There's going to be certain network assets that are shared, network control systems that have to be shared. From our side, we've made access to some of their centralized data center and processing capability because those applications were moved off of stand-alone Sprint into a combined T-Mobile environment. It is our expectation that it will take another 6 to 8 weeks to fully negotiate that transition service agreement, coupled with the completion of the audit. We're anticipating closing some time in the second quarter.

And so if we tie this all together, you've obviously in the past provided sort of a multiyear growth to profitability sort of algorithm for classic Cogent. But if we think about the pro forma company, what does annual sort of target growth margin expansion leverage look like for the pro forma entity?

Let me start with what Cogent has actually done since going public, with Deutsche's Bank help we have been able to generate 10% organic top line growth, 200 basis points a year of margin expansion to the point where we're about $600 million of revenue and just under 40% EBITDA margins. We are acquiring a business that will shrink to $450 million. Most of that shrinkage is occurring right now while T-Mobile still owns the asset because there's unprofitable products that have been end-of-life by T-Mobile. If we had closed in the fall, they would have been further along in their retirement. We'll probably end up with a little more than the $450 million that we anticipated and probably a little worse margin profile just because we're earlier in the process. But those revenues and products will burn all. The $450 million large enterprise business that is comprised of selling VPN services and DIA will be flat to 2% growth. Call it a 1% growing business with probably a 20% EBITDA margin. You've got $600 million of Cogent with a 40% margin. You've got 450 of T-Mobile with a roughly 20% margin. Now you have the wavelength business on top of that. The wavelength business is very embryonic. It's really just a test bed at this point with about an $8 million run rate. It was severely constricted by the fact that T-Mobile did not have a sales force to sell it and it only terminated in 24 carrier neutral data centers as opposed to the 800 that we're bringing online. That business will grow very rapidly for a few years. I think once we kind of reach maturity in that business, it should grow at about 5% to 7%. When you blend these 3 different components of growth together, the total Cogent business within 5 years should be about a $1.5 billion business. It should have roughly low 30s EBITDA margins, with margins expanding about 100 basis points a year and combined revenue growth, including the low-growth enterprise business of between 5% and 7%. And looking at the blended business, the investor can think about top line of 5% to 7% growth.

I'm going to hit on leverage a little bit later in the discussion. But maybe just as we're sort of talking about the deal itself, where would your leverage, I think you've been a little bit maybe north of fourth of last quarter, but I'm just wondering where you sort of envision the pro forma business fee?

We have a net leverage target of between 2.5 and 3.5x EBITDA. We have been above that for 2 reasons. Part of it is spending associated with the integration of the networks that actually began in advance of the regulatory approvals through a series of commercial agreements. The second was our continuance of growing the dividend at our historical rate. We have 42 sequential consecutive quarters of growing our dividend. Prior to the pandemic, EBITDA was growing at mid-teens, of 15%, 16%. And we were growing our dividend at about 13% or 14%, $0.025 per share sequentially. We continued that during the first part of the pandemic for 2.5 years, believing that our corporate business would recover more quickly. Now our NetCentric business, which is 44% of revenues actually accelerated and continues to grow about 60% faster than its historic average. 9% historic average growth, about 16% today on a constant currency basis. But the 56% of our revenues that come from corporate users that have traditionally grown at 11% are now growing at about 2%. That actually costed a negative 9% during the pandemic. As a result, our leverage checked up because we're going to be able to count the vast majority of the T-Mobile payments as revenue. We are making available to them for support. Whether they use them or not is still to be determined. We can recognize that as revenue. Now the payments are asymmetrically front-end loaded. The revenue will have to be recognized ratably over the 54-month contract. In the first year, we'll be getting $29 million a month for a year. That payment steps down to $9 million a month for the next 42 months. From an income statement perspective, we will recognize approximately $13 million of revenue per month. That because it has a 100% margin contribution, that will rapidly delevel us.

I'm going to pause. If anybody has any questions in the audience? Let's talk about the core classic business. And you hit on corporate. It's a good place to start. Obviously, as you noted, we've seen revenue from the start of the pandemic, I think we declined for about 8 straight quarters, but now we've actually gotten back to a little bit of improvement for last 2 quarters. Can you help us just think about the puts and takes here, the outlook for '23 and maybe we can weave in some of the commentary around vacancy trends in some of your central business districts?

Our corporate business is focused on one billion square feet of large multi-tenant office buildings in the central business districts of major cities. That's the exact footprint that was hurt the most by the pandemic. The vacancy rates in the buildings we serve went from 6% to 18%. They've retreated somewhat here at about 17.5%, but still far above historical averages. If you look at employee entry into those buildings, as measured by security badge swipes, we went from 100% pre-pandemic down to literally zero at the worst of the pandemic and the climb back has been slow. We're at about 55% a day. The corporate recovery has been slower and more gradual than we had originally anticipated. The other compounding factor has been the fact that many businesses don't know what the future is going to look like. They went from a world where they designed their networks around 97% of workdays in the office, 3% remote and then 100% remote, no workdays, the standard has been kind of solidifying around 60% in office, 40% hybrid. There's a lot of companies who are ready to move on. They need to modernize their networks and they're making decisions. In fact, our corporate growth rate has turned positive for the last couple of quarters, it's back up to a couple of percent annual growth rate. And again, this is against an enterprise market measured by all the other providers that are shrinking at about 8% to 10% a year. We'll continue to grow at an accelerated rate. I do think it's going to be at least a year before we're back to that historic 10-plus percent corporate growth. I think for companies, many of them are still grooming secondary locations, some of them just don't know what the future is. They're a bit paralyzed. For other companies, they're now ready to move on. They've been living with antiquated technologies like MPLS through the pandemic that they wanted to jettison before the pandemic started, and they've just kind of been procrastinated, and it's not uniform. What we are seeing is a number of good waiting indicators. Number of proposals being issued are up. Time for proposal, the contract signings are shortened. We're seeing vacancy rates decline, albeit slowly, and we're seeing office occupancy rates as measured by employee entrances go up. There are a number of leading indicators that we're tracking that says corporate growth should continue to improve, albeit at a gradual pace.

And so in this sort of new normal where maybe vacancy rates may never go back to 6%. And maybe I'm wrong, we'll see how it plays out. You're still confident in sort of the 10% plus growth, admittedly on a year plus?

I actually think a few things are going to happen. Average office forepoints have shrunk about 20% from pre-pandemic levels. The company typically took 11,000 feet. Now it's taking 9,000 feet. The good news for Cogent is because we sell one connection per location to a business, we actually have a bigger TAM if the building returns to its historical 94% occupancy rate. The second thing that's happened is a significant amount of office inventory that's come out of the market being converted to residential use. That trend is continuing. It typically is not in the buildings that we're in, but it is mopping up the excess supply at the lower end of the market in the BRC building. We feel pretty comfortable that in our footprint, while the equity ownership of those buildings may change, they are still going to be occupied and will return to historical average occupancy rates, which were about 94%. And as a result, we'll actually have a bigger number of customers we can sell to and actually a larger opportunity than we had pre-pandemic.

Let's maybe pivot to the NetCentric business. That's actually seen some pretty outsized revenue growth. I think you were in the mid-teens in terms of growth last year on a constant currency basis despite some tougher comps. If we can maybe talk about what's driving the strength? And then what's the outlook for that business for 23?

If I have to admit why I was too optimistic on the rebound in the corporate business. I was too pessimistic of the outlook for the NetCentric business. I thought that there was a significant pull forward of the migration from linear to streaming and it was a onetime event. In fact, what we're seeing is that trend continuing through multiple streaming providers and more importantly, the globalization of streaming as a phenomenon, moving away from being just a developed world product and now really being widely accepted in the less developed world. We saw our NetCentric business go from 3% growth going into the pandemic, actually below long-term average growth of 9%. These are revenue numbers that shut all the way up to over 25%, 25.5% at peak. And today, it's about 16%. It is coming down, but it's coming down very slowly. What is driving that are 4 factors. One, the globalization. More of the growth is coming from less developed markets where pricing is better. Second, the market is more fragmented. We charge more for smaller customers than we do for our largest customers. Telkom Kenya pays more than Google per megabit for their service. Third, we have been able to get a higher percentage of our business where we get 2 sided payment, meaning the traffic is going from one Cogent paying customer to another, from an access network to a content network or a content player to an access network. And then finally, on the content side, the market has also become more fragmented. It was dominated by one single player a couple of years ago. While that player remains significant in this Cogent customer, there are many other alternate streaming services. And it's not really our place to opine on winners and losers, but rather just to supply the bandwidth to facilitate all of their business models, whether they be ad-based or subscription-based, whether they be English language or some other language.

And this is primarily, it sounds like video or are there other use cases that are generating this ramp in traffic?

Video is by far and away the dominant use case today. We provide the majority of the bandwidth for TikTok, for example, is that traditional video? Not exactly. It's probably more like the way YouTube was when YouTube first got started and YouTube morphed more into a Netflix type model. We support all of those models. We've also seen gaming be a significant driver of bandwidth in certain windows. It's not as consistent as video. And then finally, I think there will be a leg of growth that comes from virtual and augmented reality. I think that's more hype than kind of a real traffic generator today. But I think there's a lot of work being done by a number of our customers on various products that will meaningfully move the needle over the next few years. Virtual reality is just a different form of video, but it's still one that kind of drives off to the same type of user experience.

And so as we think about sort of the multiyear, you've been maybe punching above your weight and growing a little bit faster, but is 9% still sort of a reasonable assumption to think about for NetCentric on a multiyear?

I think that's right. I think we can grow over the next couple of decades at about the rate we've grown over the past couple of decades in NetCentric. Our competition has shrunk, it has become defocused and those that are left have a significant number of other challenges. And it's important to remember that the transit market, while critical to the global economy is only a $1.5 billion TAM. And we're almost any vertically integrated telco, it's not big enough to solve the problems that their balance sheets are facing. Logically, they focus on other areas in their business. That's given us the opportunity to grow and continue to capture market share.

I want to talk a little bit about the sales force. I know you've been pretty active rehiring in terms of quota bearing graphs trying to get back to where you were pre pandemic. It's driven a little bit of a decline near term in terms of productivity per sales rep just in terms of the headcount actually growing. Where are you now in terms of hiring? And then how does the sales force evolve pro forma for the Sprint GMG deal?

I'll start with the Sprint GMG transaction. There are 60 salespeople. They are completely relationship driven. They do not go out and look for new customers. In fact, they don't even have a commission plan, they're paid on salary. That's not what I think of as a sales force. Pre pandemic, Cogent had approximately 600 salespeople. We actually continued to hire through the pandemic at record rates. But because we sent our sales force home and they were working remotely, the turnover rate spiked. We have an outbound tele sales model. Over 90% of our sales occur either through e-mail or through the phone, never actually physically meeting the customer. Only 1% of our sales come through third parties like VARs and channel. Our turnover rate went from an average of 5.7% of the sales force a month, spiked at 8.7% during the pandemic. We saw our total sales force decline from 600 to 480. We have rebuilt that sales force by continuing to hire, but more importantly, by reducing the turnover rate down to 4.7%. We got our employees back in the office. We intensified our training efforts, lower that turnover rate. Maybe somebody in the audience is thinking, well, that's still an awful number. You're still turning over 75% of your sales force a year. Well, anyone's ever actually cold-call telemarketed and lasted a year. They're a very unusual person. It is a very, very challenging job. And I wish all of my salespeople succeeded. They all come to Cogent with the best intentions. We do the best we can in training them, but there are a few that can actually do that day in and day out. Last quarter, on a year-over-year basis, grew the sales force 11.8%. As a result, our average sales force tenure in the past year declined from 34 months down to 29 months. That resulted in lower productivity numbers. We anticipate another quarter or so of outpaced hiring. We want to get back to that 600 number that we were at pre pandemic. At that point, we want to revert to a more normalized rate of sales force growth of about 7% a year. Down from 12% back to 7%. That should allow productivity to increase. We then will layer on the 60 acquired sales professionals from Sprint. They will be given a quota plan. They will be expected to help grow revenue as well as maintain the relationships they have. Hopefully, they'll all do a good job at that. And then we'll just grow off that base. I think investors should expect for '24 and '25, more like a 7% growth rate in sales force, and therefore, a reversion in productivity back to the roughly 5.2 orders, which has been our historic average per rep per month.

We tie this all together and the time we have left, what do you think is most misunderstood by the street about the Cogent story? And where do you see the potential for upside surprise over the next 12 months?

We get lumped with other enterprise telecom companies. And that's an awful neighborhood to be in because the Internet is deflating the industry and cannibalizing legacy products. We buck that trend because we focus very efficiently on the fastest growing but also most commoditized part of the market. I think in the core Cogent business, our product differentiation is still misunderstood by investors. Our ability to grow top line when everyone else is shrinking, expand margins. On the Sprint acquisition, I think investors scratch their heads and say, this is a 40-year old network, how could it be worth anything? And sometimes, it's better to be lucky than smart. The original fiber that was deployed actually turns out to be better, who were coherent transmission than the fibers that were deployed in the late '90s and early 2000. It's a long unique right of way, and it was buried in armored cable. The downsize is, it's very hard to pull additional fiber but with the advances of optronics, you don't need that. I think the other part of the Cogent story that's maybe not fully appreciated is our ability to repurpose those assets. And it's not like we've never done this before. Cogent started out in '99 to build their network organically. When the dot-com no-down occurred, we did 13 acquisitions and built Cogent out of those repurpose assets. We've been waiting for 16 years until the next opportunity came along. And I think investors will be rewarded because of our patience.

A great place to end it. Dave, thank you again.

Matt, thank you for hosting me and thank you, everyone.