Deutsche Bank's 29th Annual Leveraged Finance Conference

Oct 5, 2021

Good afternoon, everyone. Welcome to the afternoon sessions here at Deutsche Bank's twenty ninth Annual Leverage Finance Conference. Again, I'm Anthony Klarman, and I cover the telecommunications, cable and satellite space here for DB. I wanna welcome everyone to the moderated fireside chat for Cogent Communications. Just by way of logistics, for those of you who haven't queued in previously, there is a question and answer dialogue chat box that you should see next to the video screen on your link for the conference. If you have questions that you would like me to address with Dave, you can feel free to type them in there, and I will certainly get to as many questions as I can as time allows. I do wanna introduce from the company, the chief executive officer, Dave Schaefer. Dave, I wanna thank you very much for joining us even even if it's virtual, and and I, like you, certainly hope that we are back in Arizona next year doing this in person. Well, thank you, Anthony, for inviting me. I think I've been to about 20 of these over the years, and it is always a great conference. I wanna thank Deutsche Bank. I'd to also thank investors for taking time out of their busy day, and I am literally here to answer questions. Great. So that's a that's the perfect place to start. And, unfortunately, Dave, I'm gonna have to make you answer the same question again that you probably have been answering for about eighteen months now, which is really that the the the the Cogent business is a tale of two cities that happened during COVID. You you have a corporate business, which is the bigger slice of your business, that saw some significant headwinds as we all went from office work to, you know, working from home. And then, obviously, the NetSetric business, which saw a pretty big surge in demand for bandwidth and capacity. Capacity. You guys have talked at length about the various pulls and pushes that you've seen in each of those businesses. But can you refresh us on where those are today, particularly as we have started to see some signs of reopening and the returning to work. I'm sitting here inside 60 Wall today. I can tell you the floor that I work on has more people on it than it's had since probably March of twenty twenty. What are you feeling on the ground in terms of the dynamics in in your two businesses, and and where does that stand as we sit here today, and maybe how we can think about ending out the year? Yeah. So let me start by saying compared to other wireline telecommunications service providers, we have fared very well through the pandemic. We have grown every quarter sequentially and on a year over year basis. We've seen our margins continue to expand. However, our total growth rate during the past six is about 50% of what it traditionally has been because of the pandemic. In our corporate business, which accounts for 62% of our revenue, we're selling Internet access to end user customers located in skyscrapers in the central business districts of those major metropolitan markets in North America. That business has gone from a average growth rate sequentially of a little better than two and a half percent and an 11% annual growth rate to a negative growth rate of around 4%. And that is direct result of people working remotely. Digging into that corporate segment more granularly, there's a lot of puts and takes as you said. Within the corporate segment, our primary product is selling dedicated Internet access at the primary location. That business has actually fared very well through the pandemic, and we actually saw an acceleration of a trend that was in place prior to the pandemic, and that is companies migrating to larger connections. Traditionally, Cogent had sold a 100 megabit connection as its core product. And twenty two years ago when we got started, many customers and investors laughed saying no business would need a 100 megabits. In fact, today, the vast majority of our sales are one gigabit connections. And that trend of migrating from 100 megabit to gigabit accelerated during the pandemic because many companies had work from home employees, and their firewall in that primary location is functioning as a virtual private network concentrator for those remote employees. So that was a positive for our corporate business. On the negative side, we saw two significant negatives. The first was the sale of Internet access at secondary locations. Because most of these companies had closed those secondary locations, we saw a significant diminution in sale of second and third locations. Now as companies are starting to repopulate offices and return to work, we're seeing two distinct trends. If those secondary offices are in remote MSAs, they are looking to expand or continue to keep those services. If those are within the same MSA, we're seeing hybrid work from home substitute for multiple offices in the same metropolitan market. The second big negative trend that we saw in our corporate business was the almost complete cessation of the sale of virtual private networks to link offices together utilizing either VPLS or SD WAN as an MPLS replacement. Now that trend of MPLS replacement is continuing, but most customers hit the pause button for the past eighteen months. What we have seen over the past four or five months has been a reacceleration in activity in our corporate business, meaning our salespeople are talking to more prospects, prospects are requesting more quotes, and we're actually seeing an uptick in orders. You know, I think there's a lot of geographic dispersion around responses. Typically, in the South and the West, we saw offices repopulate more fully, 70% of employees back in those facilities. Whereas in the Northeast and in Canada, we saw return to office go much more slowly and office populations are probably at 15 to 25% of peak. You know, with the advent of vaccine mandates, most companies are now laying out formal timelines for people to return to office. It sounds like Deutsche Bank's maybe a little ahead of the curve. If we go back and look at our corporate business for the past sixty two quarters as a public company, prior to the pandemic, we only had two quarters of negative growth in the great financial crisis. The pandemic has been the financial crisis on steroids. We're going to end up with what looks like six or seven, maybe even eight quarters of negative corporate growth. But that rate of decline is continuing to moderate. We have been fortunate that our corporate customer base is typically the most creditworthy customers because the landlords in the buildings we serve have vetted those clients for us. So we've actually seen our churn rates remain consistently low. Our DSOs are at an all time low, and the company's bad debt expense has actually declined throughout the pandemic. I think that's very different than many other service providers who end up selling to a more credit challenged customer base. So in our corporate business, we are seeing improvements, we're seeing increased activity, but we're probably still a couple of quarters away from resuming the type of growth we've historically demonstrated. Now I'm gonna pivot over to our NetCentric business where it really has been good news. Our NetCentric business, which is 38% of revenues, is a global business. We deliver services in approximately 1,300 carrier neutral data centers in 49 countries around the world. The customer base for NetCentric services is further bifurcated into two groups. We sell to approximately 7,500 regional access networks who are providing upstream connectivity for their retail customer bases. These are companies as large as China Telecom and China Mobile, Jio in India, to very small cable operators and competitive carriers. On the other side of that bit travel, we are selling to about 5,000 content generating companies. These are companies that offer streaming services, applications, search, social networking. They're pushing the applications out that those residential users are consuming. We have seen an acceleration in our NetCentric revenue growth. So our NetCentric revenue growth has averaged over the past fifteen and a half years approximately 9% year over year. That growth tends to be lumpy for three reasons. Roughly 55% of those revenues come from outside of The US, so there's typically some foreign exchange volatility. Secondly, there is seasonality in our NetCentric business. Because most Internet traffic demand is in the Northern Hemisphere, when it's nice outside in the summer, people consume less bandwidth than they do in colder months when they're inside more. We saw that seasonality actually decline throughout the pandemic and have seen pretty consistent growth in traffic throughout the year. The third factor is business concentration. Particularly on the content side, the Internet has consolidated around a handful of large names. What we have actually seen during the pandemic are multiple positive trends. One, an acceleration in traffic growth. Two, continued price declines at historic averages. Three, a significant broadening of our customer base, going away from those handful of names to many smaller players like Disney. You say Disney is huge, but compared to Netflix, it's relatively an early adopter. And then we've also seen a significant increase in the percentage of traffic that remains on our network, I we are getting paid by both sides. So when we sell NetCentric services, it is a meter service and we charge by the megabit. We have two ways of exchanging traffic. Roughly 70% of our traffic goes from one paying customer to another paying customer. So an example would be a Facebook application sending traffic to an access network operator, or a search engine sending you know, Google sending traffic to a different access network operator. In 30% of the cases, we get paid by one side of that transaction, and we exchange traffic with a settlement free peer. This is another global tier one backbone that we are not collecting from. So that percentage of one-sided traffic over the past two years has declined from 50% of our traffic to 30% of our traffic. As a result, we've seen our effective yield on traffic go up and the combination of these factors resulted in our NetCentric revenue growing at 30.5% year over year in the second quarter. Now some of that was FX with 55% of the business outside of The US, but even netting out the FX tailwind, the business grew at over 23%, which is about two and a half times faster than trend line. As we are starting to see lockdowns be eased, vaccination rates go up in the developed world, and people returning to a more normal life, we're starting to see a deceleration in that NetCentric business. As a result of these two different business segments, we think we're still a couple of quarters away from returning to our normalized total revenue growth rate. But we do think that there is plenty of continued demand in NetCentric as we still see only about 40% of video consumed in the developed world being delivered over streaming with 60% of video still being delivered linearly. Does that help, Laina? So so, Dave, yeah, that is. That and that's a very helpful context to set for the rest. I guess it's gonna still take some time for those two lines to intersect again and and to to overcome some of those headwinds. If if I had taken a poll in the early middle part of twenty twenty one, it seemed like the corporate business was sort of setting up for post Labor Day. What what do your street on the feet in terms of your sales force tell you now in terms of the discussions they're having on when they think when they think that pace can really start coming back? And does this present a new opportunity for your sales force to go in and have a new discussion with the customer, I guess, on both sides of your product offering, to evaluate the types of products customers are taking, whether the last eighteen months has created a new need for additional services and products? And how is the sales force sort of reacting to the potential opportunities that come from perhaps, you know, taking another look at some of the some of the services that you're offering to customers? Yes. Well, I'm actually gonna take those questions in reverse order. So anytime a customer has to change their IT infrastructure, that becomes a catalyst for our ability to sell. We close between 4050% of the proposals that we issue. The challenge is getting the customer engaged and wanting to focus on their Internet. The Internet is utility, and if it's not broken, most customers don't spend any real resources on looking to modify it. So there are usually three windows of opportunity to sell. When a customer moves, they need to make a new decision, when a current provider fails, or when their IT infrastructure is being reconfigured. Well, the pandemic created a fourth opportunity, which is companies supporting work from home and hybrid work models. So we engage with the customer at the primary location with one of those four catalysts. What we are seeing is virtually all customers are now willing to engage. Where say six to twelve months ago, they throw up their hands and said, I have no idea what the future is going to bring, and I'm not ready to look at a new network architecture. Now they're in the process of planning and deploying those new networks, So that's a big positive. Much higher engagement level, most customers willing to engage. The challenge has been that with the Delta variant, companies are uncertain around their timeline. As you had said in your question, and as what the market had told us in May and June, virtually all companies were expecting at least a partial return to office after Labor Day. With the surge in cases, we saw most customers push those dates out. Some in October, some at the end of the year, some still indefinitely. Secondly, companies are still formulating what percentage of their employees will work in the office and for how many days they will work in the office. So the hybrid work model is still evolving. The good news for Cogent is our service is kind of a binary service. If the customer has even one employee working in the office or remotely, they need Cogent services. So that's the good news. The bad news is they're still struggling with their exact timelines and exact configurations. If we look at office occupancy data, in the best parts of the country, we're seeing about 60% return to office as measured by elevator rides or security card swipes. And probably in the most cautious markets, we're still at between 1520% of pre pandemic levels. So I think when you look at a market like Texas, it's pretty much wide open. Even though Florida saw a significant surge in cases, they're running at 60 to 70% of normal occupancy. Whereas if you go into Boston or DC, Toronto, and New York, we're probably closer to that 25%. So I think companies are figuring it out. There are three things that will drive our growth. One, companies needing to upgrade to support this new environment. Two, their decision on which offices they'll retain and which offices they will shutter. And then third, that replacement cycle for MPLS will not only reaccelerate, I think it will go much more quickly than it did pre pandemic because I think many customers realize that that expensive legacy architecture is just not conducive to the new more dynamic work environment. I'm gonna pivot over slightly now to how our NetCentric customers interact with our sales force. There, every customer needs to buy Internet connectivity as an input to the final good or service they sell. They are reacting to increases in demand from their customer base. So whether it's Hulu or it's Spectrum, they're reacting to what their customers need. And there, Cogent is in front of them with a price discount that's 50% of what other providers charge, which has allowed us to gain a disproportionate share of the market. We're the second biggest carrier of traffic in the world. We actually have more networks connected to us than any other provider. And prior to September, we were doing this all with a remote Cogent workforce. So to remind investors, Cogent's about 1,100 employees, roughly 600 quota bearing salespeople, 150 sales management support people, and about three fifty people in operations. When the pandemic hit, we immediately pivoted to a 100% remote workforce. Only after Labor Day did we begin a partial mandatory return in North America, US actually. Canada and most of Europe will go to voluntary returns over this month of October and mandatory in November, and Singapore is probably pushed out till the end of the year. That impacted our ability to sell. It impacted our ability to hire new salespeople to train them and monitor their success. So we develop tools to manage remote sales force, but we've always acknowledged our sales force is more productive when in an office. We have 70 teams that physically sit in 35 offices literally around the world. And, you know, we are in the process of getting our employees back into the offices. There are Salesforce turnover rate increased during the pandemic, and our sales force productivity declined. We're seeing both of those trends reverse as we're bringing people back to the office. Dave, may maybe it would be helpful here to to remind people what the penetration rate is of the buildings you serve, how now you think about the catalyst to upsell that penetration rate into some of those buildings. And maybe as a third piece to that, sorry to create too much of a compound question, but one massive theme of the last several months has been a tremendous investment in fiber by the Lex and cable companies on bringing more fiber to the premise, including SMB and mid and larger tier enterprise customers. Do you think that dynamic presents an ability for you to sell greater capacity throughout the distributed network that you have inside some of those buildings? And how should we think about the potential for that to lead to some increased penetration gains and and productivity among the sales force? Yes. Let me start with some penetration statistics. The typical building we serve is 550,000 square feet. It's 41 stories tall and has 51 unique businesses. We have sold approximately 15 of those businesses service, meaning there's another 36 that are potentially ripe for us to sell to. For the average business that bought from Cogent, they buy 1.5 connections at a location. Basically, 50% of our customers buy only Internet, 50% buy Internet and VPN services. We think we can easily double our penetration in those buildings. If I pivot over now and look at penetration in the NetCentric market, we have 22% of the world's traffic in the roughly 1,300 data centers that we connect to. We have just under 30 connections out of about 200 potential customers in those facilities. And we will continue to gain market share and have grown faster than the market. Now to the point of additional fiber availability. Cogent's network is truly unique. It is built on dark fiber that we have purchased from over two eighty different fiber suppliers who constructed the underlying infrastructure. We typically buy that fiber on long term prepaid IRUs. We have 59,000 miles of intercity fiber, route miles. We have roughly 16,000 route miles and about 38,000 fiber miles of metropolitan fiber and about two fifteen markets around the world. As more fiber gets built, we have more opportunity to buy from a larger and more diverse base of suppliers driving down costs. The second benefit to fiber ubiquity is it's accelerating the migration of legacy applications, whether it be video, voice, or private networking from copper, HFC, or legacy fiber networks to the Internet. That is a positive For Cogent, anything that brings more traffic to the Internet is ultimately good for us. We do not fear the Internet. We embrace it. Whereas most legacy providers are looking to maintain legacy higher revenue per bit traffic and are reluctant to invest in the Internet. Now on the downside, does that mean we have more competition? So most of the capital that's been going into fiber overbuilds is local capital, not long haul or international markets. In the backbone business, we have seen that market consolidate from 200 players to less than a dozen. There are really only four major players in the world. We're number two out of those four, and the top four carry 75% of the world's traffic. So we don't see any of the new capital creating new competitors in our an eccentric space. In our corporate space, we have cream skimmed the market. We lay out the buildings that have enough demand to justify the capital expenditure and have cream skinned to market by going after those buildings. We've bought fiber from providers to get right in front of the building then built a lateral into the building and up the riser. In most cases, the new fiber overbuilds are going elsewhere. They're going into smaller, more suburban type buildings. We don't compete in that market except when we sell off net services. Some of our corporate customers want a holistic solution. In those cases, Cogent goes out and buys fiber or buys what services, not fiber, because we cannot cost justify building into those facilities. So, you know, we get the benefit of more competition in that off net footprint, driving down our input costs and which then result in our ability to sell at lower price points. Net net, the deployment of fiber is great for the consumer. It's good for us. The jury is still out on whether or not it's good for the investors. To date, most fiber overbills have yet to show a return on cap. Very true. And I and thank you for those stats on the network. You know, in a in a in a moment of jest, one of the problems with these fireside chats is that I missed that old classic slide that you have about how you've assembled the network buying up lots of assets over the years, many of which I worked on the financings of many moons ago. My question is about opportunities for footprint growth. As you are able to sit back now and kind of study the markets you're in and potentially new markets that you would go into, you know, one, have the underlying economics changed at all with COVID in terms of new buildings that you would wanna be connecting to based upon changes that may have happened in occupancy and tenancy? And two, from a geographic perspective, if we were to think about Cogent five years from now, would you think the geographic makeup of your business would be materially different than it was today? So two very different questions. First of all, our footprint expands three ways. We go into brand new markets. We extend our fiber in the markets that we are in, and then we add buildings to our footprint. In terms of data centers, we are pretty indiscriminate with 1,300. That's probably 400 more data centers than our next closest competitor globally. We serve 49 countries. We recently just expanded into Chile. We'll be announcing Argentina within the next month. We continue to evaluate more international markets around the world for NetCentric. On the corporate side, we have chosen to focus only on The US and Canada for two reasons. One, the buildings have the correct demographics, large, tightly clustered, multi tenant buildings where we can get good efficiency out of the fiber that we have purchased, and two, a regulatory environment that promotes competition. As we look around the world, we have not seen enough critical mass to justify expanding our corporate business into any international markets other than a very limited amount of off net to serve our domestic on net customers. You know, as we look at specific buildings in North America for connectivity, we don't get the economics of going down market. You know, the average corporate building in America is only 11,000 square feet with 2.1 businesses in that building. You know, compare that to our buildings at 550,000 feet and 51 discrete businesses. So we have cheated and made our economics better than others and gotten a better return on capital for three reasons. One, the purchase of distressed assets that you mentioned, Anthony. We bought 13 companies. They had raised 14,000,000,000 of invested capital. Six of those companies were public and had a combined market cap of $61,000,000,000. We paid $60,000,000, acquired a 115,000,000 of cash. We also acquired 400,000,000 of preferences and 500,000,000 of debt, which we then restructured. And I know this is a debt conference, so I don't wanna, you know, say anything that upsets your clients, but most of them lost money in the companies that we acquired and dissembled to then rebuild into Cogent. But we cheat another way in that we ignore the residential market. So there are a 131,000,000 residential structures in The US and Canada housing just around 400,000,000 people. We cannot make those economics work to build and then distribute into those residential users. I know there's a lot of new fiber overbuilt business models that are attempting to do that. The jury is out on whether or not they're ultimately gonna be successful because the Internet, while not optimally, can continue to run on top of those legacy coax networks. And then finally, we have a very different revenue acquisition model. At first blush it appears to be the most expensive because it's an outbound telesales organization. Over 99% of our sales come through direct sales by Cogent, less than 1% from channel. That's very different than most other business service wireline telcos, which generally run about 50% agent or channel. That means we have a lower cost of revenue acquisition coupled with a longer duration of customer life. Our average corporate customer lasts about sixty five months. So, you know, very durable revenue stream with a low subscriber acquisition cost and high margin has allowed Cogent to be profitable and generate free cash while growing. So if you look at telecom, the world's kind of bifurcated into two groups. Those companies that are declining but maybe throwing off cash from your legacy businesses that eventually go to zero, but those companies that are attempting to grow and generate no cash. Cogent's really the best of both worlds. We're growing our cash flow and at the same time, we are able to return capital to our investors. That's a good pivot because with the last couple of minutes we have left, I did wanna talk about the capital structure and and the capital allocation policies of the company. You you have been busy in the capital markets earlier in 2021 taking care of some near term maturities. It it does strike me that when you guys come to market, you often go for shorter duration debt instruments than we see many companies of your size go for. You have some debt coming up that I think is callable and will present some other opportunities for financing. How do you think about the ideal makeup of the debt stack here from a duration and maturity perspective? And would we expect you to start thinking about some longer duration refinancing options in the future? And then as that plays into the capital allocation strategy, how should we think about the priorities around cash flow as it relates to returning capital to shareholders versus focusing on deleveraging versus reinvestment in the business? So first and foremost, we make sure that we can reinvest in our business. That's our primary objective, to invest in the sales force, to invest in the network. We have had surplus cash from operations beyond our ability to reinvest for the past fifteen years. That's a great place to be. We have bought back about 20% of our float, and we have issued a dividend and grown that dividend now for 37 consecutive sequential quarters. We currently have a dividend yield that's actually above our debt cost of capital, which is a bit odd, and it sends us to return even more capital. We did not build the company on debt. We were actually debt free when the company went public and decided to add leverage to our balance sheet as we saw the cost of debt capital come down. We believe that we're in a long period of low interest rates. While interest rates may rise modestly by historical standards, they're gonna remain low for the foreseeable future. With that, we are comfortable in getting shorter duration debt. Our debt decisions are kind of one dimensional. They're just based on the absolute lowest cost of capital. Our debt is completely elective debt as opposed to required debt. If we saw interest rates rise, we could quickly delever by reducing our dividend and, you know, just generating free cash flow to eliminate our debt. We have that flexibility. But with low interest rates, we think it's prudent to continue to add leverage for our balance sheet, and we today have about 3.4 times net leverage. We have about 400,000,000 of cash on the balance sheet. Investors should expect us to continue to return more than a 100% of free cash from operations for the foreseeable future because the business naturally de levers because of our growth. Even with our pandemically reduced growth rate of about 5% top line growth versus our long term average of 10%. We were able to grow margins at a 130 basis points year over year versus our long term average of about 210 basis points. But with that, we're growing free cash flow in line with our dividend. But because our borrowing power is magnified by the amount of leverage we wanna take on, we constantly have excess cash on the balance sheet. So what we're doing is trying to optimize what is the mechanism of return, buybacks versus dividends, and what is the pacing of that return. I think as investors measure us over the next several years, they should expect us to continue to return capital in both the most tax efficient and, you know, the most, you know, long term value creating way to shareholders. So, again, we have a great problem. We don't need to delever. We actually need to lever up more, and it's totally reactive to market conditions. Well, Dave, I want to thank you very much for all of those comments. We have come to the end of our time for this presentation, but I wanna thank you, Dave, for your long running and continued sponsorship in attending this conference. And thank you very much for being with us here this afternoon, and certainly the the hope that next year we will be in person together again in Arizona. We will. Take care, and stay safe, Anthony. Stay safe, Anthony. Thanks for hosting us. Bye bye. Thanks very much, everyone. Bye now.