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

Hi, everyone. Thank you for joining us today. I'm Sami Badri, senior equity analyst covering the Communications Equipment and Communications Infrastructure sectors at Crédit Suisse. If you have any questions, please forward them to me to my e-mail address, and I will filter and ask Dave on the actual call. We have today David Schaeffer and Sean Wallace, CEO and CFO, of Cogent Communications. Dave, thank you very much for joining us today.

Sami, thank you for hosting us. I'd like to thank Crédit Suisse for a great virtual forum, and I'd like to thank all the investors for the time they have given us to hear a little bit about Cogent.

Absolutely. So Dave, maybe if we could just kick it off, a quick review about 1Q '20 results and the expectations for 2020. And I think if you can just add one more layer into your response to do with sales force productivity and how this year looks a little bit different than last year because last year was noticeably a little bit different than any of the prior years you've really reported that metric. And then we can kind of go from there into some of the other questions.

Yes. So let me start with our general outlook. Cogent gives multiyear guidance, not specific guidance to any given year. But with that, I would say that as a result of the pandemic, our outlook has improved rather than deteriorated. In most businesses, economic activity has been off. But because we have been a beneficiary of 2 structural changes as a result from shelter-in-place orders and limited return to work, those 2 trends being for our corporate customers working at home, requiring more bandwidth in their corporate locations for VPN concentration. And for our NetCentric customers, an acceleration of the transition of linear television to over the top. These 2 trends had been ongoing for a number of years, but have materially accelerated over the last 60 to 90 days and will continue to accelerate going forward. We anticipate our revenues over a multiyear period to grow at a compounded rate of 10% per year. We have averaged 10.3% in the 15.5 years that Cogent has been a public company completely organically. We also, during that period of growth, expect our margins to expand at about 200 basis points a year. And again, our historical track record over 15.5 years is about 220 basis points a year. Now to the question of sales force size and productivity. Our revenue growth is driven by the efficacy of our sales organization, that is the number of people that we have selling as well as the productivity of those individuals. The individual productivity is directly correlated to the average tenure of those individuals. Over the past roughly 15 years as we've been reporting these statistics, we have averaged about 5.1 installed orders per rep per month. As Sami alluded to in his question, our productivity did decline in 2019 and began to rebound, in fact, improved 10%, growing from 4.1 to 4.4 installed orders per rep per month in the first quarter, and we expect continued improvement throughout 2020. In 2019, we grew the sales force at an abnormally fast pace. We had actually underhired in '18, and we compensated for that by overhiring in 2019, with sales force growth being above the 7% to 10% guidance that we gave. And in fact, we grew the sales force approximately 12%. In the first quarter, we saw our sequential growth in the full-time equivalent number of salespeople grow from 502 to 525 or a 4% growth in the sales force, coupled with a 10% improvement in productivity. When a new rep is hired, they go through a fairly extensive training program. At the end of that program, they are measured by a number of soft key performance indicators. At their third month, so after 90 days on the job, they begin to carry a full quota and are accounted as a full-time equivalent. We are seeing a more normalized rate of hiring this year. We expect to grow in that 7% to 10% range, and we expect our sales force productivity to improve back to historical levels from the 4.4% that we reported in the first quarter to something closer to 5% for the full year. Our sales force is divided into 4 tiers. The lowest tier sell to corporate customers that have 3 or less locations. The largest single group of salespeople are the corporate reps that sell to multisite corporations, those companies that have more than 3 physical locations. Our corporate sales force represents 72% of our total sales force head count. The remaining 28% of our sales force focuses on our NetCentric market space. There, that sales force is bifurcated. We have a small group of approximately 10 individuals that focus on the 500 largest named accounts, and then the remainder of the NetCentric sales force focuses globally on smaller service provider customers. So hopefully, that clarity helped you, Sami.

Absolutely. Thank you very much for that summary. I think the sales productivity metric is one of the key things that someone like me looks at, and I think it's in center focus for a lot of investors. Shifting over more to network traffic growth and some of the dynamics that we're actually seeing. Ever since the pandemic really broke out, there's no question that we actually saw traffic really surge and increase, driven by OTT, driven by actual corporate customers consuming more network bandwidth and traffic. How would you characterize that dynamic today? Has that growth actually normalized? Or is it a bit more -- is it still very strong?

So Cogent has traditionally grown approximately twice as fast as the industry. The Internet is approximately 30 years old and has grown at a compounded growth rate of about 23%, 24% over that period and will continue to grow as new applications are developed, as Internet usage becomes a more global phenomenon, and finally, the amount of bandwidth consumed per minute increases. And the key driver of growth for the past several years and probably for the next 5, going forward, is over-the-top video. In the developed world, the average person consumes about 300 minutes a day of video. And today, probably only about 1/4 of that video is delivered over the top. Now that's materially up from even a year ago when it was about 18%. So we are seeing a rapid acceleration in the adoption of over the top, and we've seen a proliferation of new products being offered. Virtually all of those new service providers utilize Cogent. We also see that the approximately 7,000 access networks, that we sell upstream connectivity to, are seeing an increase in bit unit volumes due to the fact that their consumers are using more over-the-top video. So what we saw with a very rapid change in consumer behavior and the shelter-in-place orders was a step function up in that rate of growth for about 6 to 8 weeks. After that initial surge, and that data has been reported not only by Cogent but by dozens of access networks around the world, from cable operators and telcos globally, we've seen now a more normalized growth rate, reverting back to that long-term 20%, 22%, 23% traffic growth for the Internet, albeit off of an elevated base. Cogent will continue to grow substantially faster than that and capture market share. Roughly 95% of our traffic comes from our NetCentric customers, even though they only account for 31% of revenues. The primary application in that market segment is over the top. And because Cogent offers its services in over 1,240 carrier-neutral data centers around the world in 46 countries at a 50% discount to any other provider, we continue to gain market share in that market segment.

Got it. You already alluded and kind of answered my question related to OTT and some of the other kind of workloads. So we're going to skip over to some of the other questions that I have prepared for you. Could you walk us through some of the actual corporate customer contract changes that you worked through during the very busy periods of COVID? I know there was a lot of bandwidth increases, mainly tied to the increases in traffic that everyone all of a sudden had to execute, tie it to working from home. Could you just walk us through the contract dynamics and the changes that you worked with your corporate customers on increasing bandwidth for everyone?

Yes. So let me start with a fact that at first blush seems counterintuitive. If an office had, on average, 30 employees, that's a typical Cogent office, occupying about 8,000 square feet, and that office is now empty. You would say, "Why do they need bandwidth for an empty office?" Well, what happened was those 30 employees now are all working from home, utilizing their broadband residential connection. But what they are doing is launching a VPN, or virtual private network, client on their personal computing device, whether it be their phone, their tablet or their PC, and they're going to a VPN concentrator. For a very large company, like Crédit Suisse, you will do that VPN concentration in a carrier-neutral data center. But the vast majority of midsized businesses do that at their corporate office through their firewall. So as a result, all 30 employees needed to go in and out of that firewall at the exact same instance. That was not the usage pattern when they were in the office. So the majority of our corporate customers -- as a matter of fact at the end of Q1, 87% of our corporate on-net customers had a 100-megabit connection. Roughly 11% had a connection that was 1 gig and only about 3% -- 2.5% had connections that were between 100 meg and 1 gig. It actually costs Cogent less money to deliver a gigabit than 100 meg connection because we do not need to install a media converter in the customer's premise in order to convert the signal from optical to electrical. We can do a direct optical handoff to the customer. As a result, we have been pushing customers towards gigabit products. And for the past couple of quarters, Q4 of last year and Q1 of this year, we saw more new sales of gigabit than 100 meg, but we still have 87% of the base. So we launched a program to go back to the existing base to upsell them by charging them $200 more per month, no matter what their current pricing was, for the remainder of their contract term. And as a result of that, we saw a huge uptick in the number of customers converting from 100 meg to gig. That had been a very slow process. It accelerated dramatically, and it continues at an accelerating level, which is going to result in our average corporate on-net ARPUs actually increasing, offsetting the trend of longer duration contracts, which typically tend to lower ARPUs.

That's a very good point what you just mentioned. And I think given your 2Q results are upcoming and given that these full effects of these contract changes haven't necessarily really been flowing through into the actual model and reported results, do you think that investors understand this mechanical change that's occurring in the way you're building customers and in your revenue stream?

I'm not sure, Sami. I mean in many ways, Cogent is a boring business. It doesn't change very much. We've been public for 15.5 years. And we don't do a lot of mergers and acquisitions. We don't have a lot of capital markets activity. And we slowly and methodically distribute more and more capital to shareholders. We've raised our dividend for 31 consecutive quarters. And as a result, investors tend to get kind of complacent about the Cogent business model and think it's just going to be more of the same. And the types of trends that we've been talking about on today's call are examples of structural changes in the market that are helping Cogent in providing tailwinds to our business and should accelerate the growth of free cash flow even faster than we have been delivering. And I think investors finally are a bit joyous on the telecommunications sector which we get classified as a member of because virtually all of the other companies that we get comp to have negative organic growth rates. And Cogent stands unique over, again, 15 years of having organic growth rates. So it really takes an investor that extra amount of effort to look under the hood and understand why we're doing better than everybody else.

Well, Dave, for the record, I don't think your company is boring at all. I think it's extremely fascinating with the structural things you've been able to affect. And for most investors on this call, I don't think people are actually crediting the company with the structural changes that we've actually seen play out over the last 6 months specifically. So kind of tied back to pricing. We covered corporate. We covered sales force productivity. The other variable I really want to talk about is NetCentric and elements of NetCentric are derived from OTT. But maybe we could talk about NetCentric and NetCentric pricing and how that is actually trending. Could you give us a little bit of an overview of how that's going and what the expectation is for 2020?

So the average price per megabit declines for nearly 30 years at a compounded rate of 23% per year. It's much like Moore's Law, in that we benefit from 2 key technology improvements. The cost to move a bit a mile on fiber continues to decline as optronics continue to improve at about 80% a year, much faster than Moore's Law. And then secondly, the cost to route a bit also continues to decline, albeit at a slightly smaller pace of about 40% per year. Our network architecture allows us to capture those improvements much more effectively than any other network. So for that reason, Cogent's capital intensity has declined, yet our margins have expanded, and we are able to undercut our competitors. Again, going back to what investors believe, at least have preconceived notions about, their natural tendency is to dislike commodity-based businesses. Our NetCentric business is a commodity business, but we are the low-cost producer and we have a substantial moat around our business that gives us a competitive advantage. Now our rate of price decline on average is similar to that of the market, but from a starting point that's 50% below the market, trying to maintain that price differential to gain market share. Now what can impact that on a volume weighted basis is who is buying in a given quarter. If all of our purchases are coming from our various biggest customers, that will tend to depress the weighted price per megabit decline further. What we saw in the last quarter and beginning in this quarter as well is a continued broadening of our customer base. I'm not saying the big guys aren't buying more, they absolutely are. But we're seeing a number of new OTT players enter the market with new products that broaden the base. We've also seen a continued internationalization of the OTT phenomenon. And roughly half of our NetCentric traffic and revenues come from outside of the United States. It's a small market, big continent. We are the #1 carrier in Africa. We have 80% market share on the African continent. Now it's only 2%, 2.5% of global traffic, but it's growing 2 to 3x faster than the developed world. So we've announced actually we're going to be putting points of presence in Africa. And next month, we'll be opening a POP in Johannesburg, South Africa. Today, our South African customers and our African customers from almost 40 countries meet us in Southern Europe. So it is this continued globalization that's also helping us achieve a better volume-weighted price decline than our reported price decline, resulting in our NetCentric revenues continuing to accelerate. They're still not back at historic norms, but they're definitely up materially from where they were in the past year.

Got it. Got it. Thank you for that color on the NetCentric side. One thing that, as an analyst, I look at and I think what investors look at and specifically your model tied mainly to adjusted EBITDA margins is, when you grow between, say, 2% and 5% top line, what is the adjusted EBITDA margin expansion? And then if you were to hit, say, between 8% and 10% top line growth, what is also the expectation for adjusted EBITDA margin expansion? And the only reason why I ask you this is because last year was almost like a puzzling year for a lot of people, where you saw pretty healthy adjusted EBITDA margin expansion with almost like a dislocation in the guided -- in the guidance algorithm that we generally have been trained to hear. So kind of understanding these guideposts suggest the significant operating leverage in the company would be very helpful for us.

Yes. So again, I want to be clear that both of our guidance targets are meant to be looked at over a multiyear period. We underperformed on revenue growth last year at about 6%, yet we delivered on our margin expansion. Quite honestly, there's probably a little bit of cushion in that 200 basis points because it was an average over the past 15 years, and much of that 15-year history, Cogent was rapidly expanding its footprint. So that means we spent a greater percentage of revenues on CapEx, but we also pick up the OpEx of those new markets before they become cash flow positive. Our rate of physical expansion has slipped. We've been very clear with investors. We're going to be disciplined about where we deploy capital. And for that reason, our on-net corporate footprint is about 960 million square feet. It's going to grow but at a slower rate. And then secondly, our NetCentric footprint at 1,240 data centers is the most extensive in the world, and we'll continue to add to that footprint, but only as new data centers that meet our return on capital criteria. So our contribution margins have actually improved. Our historical contribution margins over 15 years have been 44%. Last year and probably this year, they'll be in the low 60s. So that means we can get more aggregate margin expansion. If our growth rate returns to our historic average, which it's trending towards, our margin expansion at this lower growth rate will be greater than 200 basis points.

Yes. That's just tying back to one of your responses is quite a cushion. There -- a lot of variability, and I would almost say upside surprise embedded in that if you guys cross the 5% top line growth, especially with the way you've built the company in the network. Shifting gears a little bit more to the cash flow statement and balance sheet. You're clearly generating a lot of cash. Your leverage levels are coming down, almost testing the bottom end or even almost below your guidance range at some point in the next couple of quarters. One of the big things that a lot of people point out at your stock is the consistent $0.02 increase to your dividend per quarter, which you have, I would say, robustly maintained and communicated to investors. How should we think about this changing as you look at dividends, buybacks, your capital allocation and your leverage levels? Is there a scope to potentially see a dividend increase or a more proactive or aggressive buyback mandate or ASR, I guess you could say, for the company? And just walking us through the capital allocation philosophy from here would be very helpful.

So I'm going to start with having a good business that's producing more cash than it consumes. That's different than many telecom companies. Secondly, we have been very rational about our balance sheet. And we have used leverage in a low cost-of-capital environment to accelerate returns of capital. We have shrunk our float by 20%, buying back 10.3 million shares over the years. We also have this consistent history of growing our dividend. We've returned over $800 million to our equity holders, and we've done it in a tax-efficient manner with roughly half of our dividend being treated as return of capital. Now we can't predict going forward exactly how well we'll do on return of capital, but tax efficiency does weigh into our thinking. We have a commitment not to go below 2.5x net leverage. We actually just increased our gross leverage by borrowing some additional capital in Europe and repaying without a penalty, some of our North American unsecured debt, lowering our cost of capital by about 130 basis points. That means we have more cash to give to shareholders and less to give to bondholders. We're going to be rational about that. We -- if we see huge market dislocations, we'll be aggressive about using buybacks. If not, we will methodically grow our dividend and stay within our leverage range. The arithmetic that you alluded to, Sami, says that at some point, we have to raise the rate at which we're growing the dividend. The Board debates that issue vigorously at every Board meeting, and we are committed to being good stewards of capital. I think management and the Board view that as their highest priority. We're running a great business. I'm not going to say it runs itself. It definitely has lots of operational challenges, but we also have the additional responsibility of returning capital in an efficient way, and we're committed to that.

Got it. Got it. So we have about 5 minutes, and I did get one question, but I'm going to phrase it a little bit differently, which is there is clearly a dislocation or discount embedded in the stock when you look at it from a dividend yield perspective, which is actually the way I value your company. I think it -- that's the correct way to do it. It's my opinion, obviously. One thing that we would -- that would be very, I guess, interesting is, based on the conversations you're having with investors, what seems to be the key hurdle to getting them very comfortable with the business about -- with Cogent's business, its model or even its capital return profile? And to kind of tie to a point when I first got to know you, one of my first meetings revolved around don't compare me to every other fiber or telco company, and you should probably look at comps as like a data point, but not the actual absolute like final view. Maybe kind of just tying into that same frame of thought is what seems to be coming up with investors that's almost like something that investors aren't really getting over with, a key hurdle to help you understand your company, your business and your model.

So we sell Internet services, so therefore, we get lumped in with telecom companies. In many ways, our business has many of the attributes of a tower business. Those businesses are valued differently on AFFO per share. We, if measured by that metric, would look inexpensive, particularly with a lower rate of capital intensity and a higher rate of growth organically. We have the added benefit of having a diverse customer base, not a concentrated base, and we have the structural tailwinds to demand. Now every business always needs to worry about future demand. But because we are predicated on only selling Internet, we think that we're pretty isolated compared to most other business. Listen, we're in an environment now where GDP on a year-over-year basis can be down 35%, 40%. We're going to be up 8%, 10%. That's a great place to be. And going forward, our products are in more demand than virtually anything else in the universe. Ultimately, the purpose of a business is to produce cash for its stakeholders, and the majority of our value resides with our equity. And I think investors need to get comfortable that free cash flow per share and the growth in that free cash flow per share is the metric that they should use to weigh Cogent against other investments. Listen, we're in a competitive market operationally. We have big incumbents we can speak to sell corporate customers. We have big international carriers for NetCentric. When we compete for investors' mind share, they can invest in 4,000, 4,500 other public companies. Why they would choose Cogent is we produce a consistent growth of free cash flow per share based on our structural advantages of demand and our advantages that we are protected from competition. The dirty word is we're a commodity business, but there are great commodity businesses, and we have one that just has a lower cost of production. The Saudi Aramco produces oil cheaper than Lukoil produces oil. Well, there's a reason. Their cost of production is just much lower. We have structural advantages that no other competitor has, and I think investors need to understand that's a sustainable barrier going forward.

Yes, you're absolutely right. You're absolutely right. We've got one quick question that came in. It was what would compel you to buy back a significant amount of stock with the bandwidth you have on leverage on your balance sheet?

I think it would be a view that the capital markets, the broader stock market, had sold off, and no matter how well Cogent does as an absolute company, it will be taken down with that market. We will capture that volatility as a gain to shareholders, and we would be extremely aggressive, maybe even so aggressive is to suspend the dividend for a period of time to focus entirely on buybacks. My job is to allocate capital most efficiently. And if I conclude it as a 10% owner, it was in the best interest of myself and my other 90% of my partners to use that money to concentrate future earnings per share with buybacks, I would at least recommend to the Board that they take that policy. We have not had that type of capital markets dislocation, which is very surprising considering the level of economic turmoil that we have, and it's clearly a monetary policy that's driving that. Listen, I can't fight the Fed, so I'll just give shareholders cash and let them figure out what to do with it.

Absolutely. Dave, thank you very much for spending time with us, and thank you for joining us in the Communications Conference. Always a pleasure catching up with you. Next up for everyone on the fireside chats, we have Keith Olsen, the CEO of vXchange. Dave, once again, and Sean, thank you for joining us.

Thanks, Sami. Take care.

Absolutely. Take care.