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

Well, thanks, and good afternoon, and welcome back to Citi's Global TMT West conference. For those of you I haven't had an opportunity to meet, I'm Mike Rollins, and I cover the communications services and infrastructure sectors for Citi Research. Just by way of reference, we do have disclosures available on the conference registration site. And I'd like to welcome back to our conference, Dave Schaeffer, CEO and Founder of Cogent Communications. We also have joining Dave, CFO Sean Wallace. Dave, Sean, thank you both for joining us today. Happy New Year.

Happy new year to you, Mike. Congratulations on the first virtual conference. I'm always glad to be back. Although I would tell your art department to correct your Citi scape to make sure the Citicorp tower is front and center, not World Trade Center.

I think that was another option for me to select. I'll have to try to bring it in later. Well, thanks so much for joining. We're just trying to do our best to support internet data traffic growth with the virtual format.

And so, Dave, it's been a tradition, not only for you to be at the conference, but for the first question of our -- for you at the conference, is just to think about your strategic and operating priorities for the year ahead.

Yes. So 2020 was not the year that we'd expected when you asked that question last year. And I think we have been both a beneficiary and a victim of COVID-19 and the changes it has brought. As we're looking into 2021, we, I believe, will have 2 years. Probably the first 6 or 7 months will look much like the latter part of 2020. And then hopefully, we're beginning to emerge from the pandemic and resuming a more normal life and more normal business life. I think for Cogent, our priorities remain focused on growing our top line revenue in a difficult economic environment. It means adding additional salespeople, training those salespeople, making them effective. Secondly, because of the surge in traffic that the pandemic has created, it's really forced us to accelerate some of our network augments. We do that on a routine basis, but we're seeing unprecedented aggregate volume growth and still very robust sequential and year-over-year growth in traffic. So we're constantly growing our network. We're continuing to expand our footprint by adding additional carrier-neutral data centers as well as multi-tenant office rollings and finally, controlling costs and making sure that we achieve our margin expansion targets as well as our growth in free cash.

That's really helpful. And I think we'll hopefully have time to unpack all of those. Maybe just to start, you mentioned that this is not the year -- last year was not the year that you expected, and you talked about that there were pluses and minuses from the pandemic. Can you unpack some of those ongoing impacts of the pandemic? And has the second wave in a number of places caused any changes in the operating conditions for Cogent?

So first of all, I want to thank the entire Cogent team. We continue to work remote. I am a bit unique. I'm here in Cogent's office and have not had to work from home. But I guess that's one of the prerogatives of a CEO. But virtually, all of our staff has been working from home. We have been hiring people remotely. We've been training them, and we've been monitoring their productivity. I think in terms of our business, as I said, there have been pluses and minuses. Our corporate revenues, which represent approximately 70% of total revenues, in our primary multi-tenant office buildings, we continue to see strong demand and an increase in the uptake of 1-gigabit services over our 100-megabit product. So the aggregate deceleration in corporate on-net growth has been very minor, maybe a couple of percent, considering the economy has fared much worse. We've seen our corporate on-net growth rate decelerate from, call it, 11.5% in primary locations, down to about 9.5%, which is really the secondary locations. Virtually all of our off-net corporate business and a portion of our on-net business is in those secondary [indiscernible] offices. There, we've seen our growth rate materially decelerate from, again, about an 11% total growth rate, down to low single digits, kind of 1%, 2%. I think the continuation of the pandemic and further restrictions in many markets will probably keep that growth rate muted for the next 6 or 7 months. As employees begin to return to their offices, we will see those secondary locations begin to accelerate. In our NetCentric business, which is a volume-based business, where we're selling to either content producers or access networks, we've actually seen an acceleration. We've seen that business improve in terms of growing back more at historic growth rates of roughly 9% on a revenue basis. We've seen our unit volume growth accelerate, and we've seen a broadening of the customer base, which has helped our effective price per megabit. So that 30% of our revenue base is actually doing better since the pandemic, and we continue to see strong increases in streaming content such as this conference or professionally produced streaming content for entertainment.

That's really helpful. And I'd like to throw out a survey question first on the corporate opportunity. And one question that were asked by investors is, can Cogent -- this is the first survey question for our operating staff on the virtual stream. Can Cogent meaningfully improve its market share in corporate multi-tenant buildings? And it's a choice of yes; no, competition is likely to pick up; or no, Cogent is already close to its fair share. Now we'll get to your thoughts on this, too, Dave, but before we do, just give a little teaser. What would you say -- as people evaluate this question, what would you say your market share is of unique tenants in these buildings today?

So we have, on average, 51 unique tenants in these multi-tenant buildings, and approximately 15 of those are Cogent customers. So that gives our penetration at about a 27%, 28% penetration rate today.

Okay. Okay. So we'll see what our audience comes back with. And as they're going through that, in the corporate organization, I guess, the whole sales organization, you had a new CRM. So as you're thinking about sales opportunities and productivity, are you through the learning process and into the productivity phase now of this new system or there's still some time and training needed to get to full speed?

No, I think we have pretty much not only transitioned and regained the ground we've lost by deploying the new system, but we're actually seeing the net benefits of making our salespeople more efficient. We are a strong believer in commercially off-the-shelf software. Cogent had operated for 14 years on one platform. When that platform was reaching end of life, we pivoted to another standout CRM for 6 years. What we found, however, is it was almost impossible to integrate that software-as-a-service platform into all the other parts of Cogent. And the productivity of our sales force was deteriorating because they were spending time going back and forth between different systems. A little over a year ago, we decided to build our own CRM. We diverted development resources to do that. And in July, the slowest month of the year, we rolled out that new platform. In July and August, we did see some teething pains associated with deploying a new platform. But since then, we've seen a consistent improvement in our sales efficacy and productivity as measured by number of minutes on the phone, number of phone calls, number of emails, number of leads converted to opportunities, but maybe most importantly, number of orders sold and converted into revenue. So even in a difficult macro environment, we are continuing to see improvements in our sales force efficacy, and we're doing that off of the largest sales base we have ever had. So we continue to grow the sales force, and we're watching sales force productivity rebound back to historic levels.

And that's around -- is it 5 units per month per rep?

That is correct, Michael.

So let's see if we have those results from our first survey question about your share in multi-tenant buildings. Let's see what our audience thinks. So 100% said no, competition is likely to pick up; 0 yes; 0 no, close to fair share. So, Dave, can you unpack the competitive landscape in these buildings and talk through how you see the opportunities to improve your share?

Yes. So our primary competitor remains the incumbent. We occasionally will see alternate providers encaving, but there are some significant architectural differences that give Cogent a competitive advantage, and there is the sales distribution difference that gives us an advantage. Now let me take each of those separately. When we identify a building to bring on to our network, we prewire that building with fiber, bringing fiber to every -- a breakout box on every third floor. That allows us to turn up customers both cost effectively and very quickly. We can typically turn up an on-net customer in 9 days, whereas the incumbents average about 80 days for similar service. Secondly, the incumbents have not prewired the building, so they need to do a home run fiber from the suite to their minimum point of entry in the basement. That additional step is expensive. On average, you spend about $1,000 per story making that home run, making those installations more expensive. The other big architectural advantage is our network is designed as a series of rings. We have approximately 950 rings in 209 markets around the world. That gives us the ability to protect every building. The incumbents for the cable plan will not have that level of redundancy and therefore, will, on average, fail 3x more likely than our network will. So quicker install, more reliability. The third architectural advantage comes from the fact that we are not selling a heterogenous product set based on TDM but rather use Ethernet, and our network is optimized for internet. That gives us the ability to sell quickly, simply and at better prices. I'm now going to pivot to our distribution model. Our distribution model is based on outbound telesales and direct marketing. Our competition usually relies on advertising, over-the-air marketing. It does not have the direct sales force that Cogent has. In fact, most of the incumbents and competitive carriers have pivoted to an indirect channel model. At Cogent, we made the decision to have a direct sales force. 99% of our sales come from our direct sales force. We also do have channel which is an indirect way to sell through trusted advisers, but it's less than 1% of our revenue base. So our direct sales method gives us both a broader cost of revenue acquisition and privity to a direct relationship with the customer. That lowers our cost of revenue acquisition. And we remain focused on the only growing part of the telecommunications bundle, whereas our competitors remain focused on a more legacy-based product set that I think is increasingly becoming antiquated. So I'll actually disagree with the survey respondent. I would suspect with 100%, that was a single individual. And while competition will always speed up when companies are in distress and most telecom companies are undergoing distress, what we see is those companies pivoting to other market segments and other products, particularly focusing much more on wireless than they are on a direct internet connection. Remember, Cogent's tagline says it all, smart people buy dumb pipes. None of our competitors are willing to focus on selling just dumb pipes.

Before we get to the NetCentric side, I want to throw out our survey and let's continue a little bit more on the corporate. So the survey for NetCentric is -- we throw up survey #2, please. What are your expectations for average annual NetCentric revenue growth over the next 3 years? And I think the historical perspective, and Dave, correct me if I'm wrong, it's about 9% average long-term growth in the NetCentric business. Is that right?

That is correct.

And so the choices are 0% or negative; 0 to 3%; 4% to 6%; 7% to 9%; or 10% or more. And while we get the responses to the survey, Dave, as you're thinking more about the corporate opportunity, one of the things that you mentioned is the opportunity to have your customers buy up on the ARPU with higher speed connections. Is there some way to quantify the potential benefit of that, whether it's in the near term or over the next 12, 24 months?

So there's a benefit to Cogent and a benefit to the customer. The benefit to Cogent is very quantifiable as we typically sell that gigabit connection at a $200 per month ARPU premium to our 100-megabit product. The benefit to the customer is that as they have employees working remotely, they have ample symmetric bandwidth to support their work-from-home network with no network congestion. Now our corporate customers buy services on a flat subscription model. It is not usage-based. Most of our customers do not use their networks 24/7, and they do not use all of the bandwidth that they purchased. So the quantification of that $200 premium is hard to be able to assess for each customer. However, it's like an insurance policy. The more you have the bandwidth, the more that surge application. And as you have remote employees, typically, professionals that have a fairly high cost, the last thing we want to do is be penny wise and pound foolish, save $200 a month only to find employees having congestion. Now there are other places in the packet path that could cause congestion. The most common problem actually occurs in those remote employees' local networks, their own networks, which are typically delivered over an HFC plant. They're asymmetric, so they're not today equipped with the amount of upstream, and they are a shared medium. Remember, Cogent's network is a symmetric network and one that has no contention. When you purchase the bandwidth, it is reserved 24 hours a day, 7 days a week at full line range.

And so if we take this all together, you mentioned the ARPU opportunity just now. You talked about the share opportunity. You talked about some of the headwinds of secondary locations in the off-net right now. If you take all those things together and the sales productivity discussion, are we at the trough or past the trough for corporate revenue growth? Or is the trough still out there somewhere and we just haven't gotten there yet?

So the future is a little hard to predict and particularly in this volatile of an environment. This is the worst corporate performance we've had in the company's 20-year history. The closest analogy was what happened in the great financial recession, where we saw corporate growth dip to about 4% year-over-year for a couple of quarters and then begin to rebound. In this case, the trough has been deeper, more like 1% to 2%, and it appears the shock through the system is longer than the financial recession. With that said, the overlying need for more bandwidth is greater today than it's ever been. So that creates a bit of a positive dynamic. I'm going to further give you a granular answer in looking at both the locations and the product set. So in the primary locations, I think we have reached trough and will probably even do better. I think in secondary locations, I think there's 2 answers. If these secondary locations are in different MSAs, I think we will see a rebound, and companies will add bandwidth to those remote offices in distant markets. For those offices that exist in the same MSA as the primary location, I think it's likely that many of those offices will permanently be shuttered as companies go to a hybrid work from home. So I think we will see a recovery, but it may not be exactly like the world worked prior to coming into the pandemic. The third way to look at this is to look at it by product. For internet access that has become more important than ever. Remember, 25% of Cogent's corporate revenues come from selling VPN services, either VPLS or SD-WAN. While there is a huge market of MPLS that is working to migrate to these more efficient technologies, that migration has been on hold in the pandemic, and I would say there has been an aggregate attrition in the market. Our VPN business has probably been the most severely impacted business with growth in that business going from high teens to virtually 0 because of the pandemic. I think post pandemic, we will return to mid-double digit, call it 15-type percent VPN growth because I think there is a pent-up demand to transition off of MPLS, which is more expensive and more difficult to implement. So a more complicated answer than just saying we've hit the trough, but it does appear we're bouncing along and things will get better as vaccines are distributed and employees return to the office, even on a hybrid model.

That's really helpful. And I had 2 follow-ups to that. The first was you mentioned the differentiation of a remote location in the same MSA versus one that's in a more remote area. As you look at their off-net connectivity, what's your sense of the percent of those off-net locations that are in the same MSA as the primary versus outside of that MSA and maybe better protected?

The vast majority of our secondary locations are in different MSAs. Probably about 80% of secondary locations are in separate MSAs. About 20% would be an example in New York City, where some of the remote locations may be in Connecticut or New Jersey. But that is a minority of our secondary locations.

And the second question is a follow-up to this, is that as you've been bouncing on the bottom, there's different cities experiencing very different conditions in terms of the pandemic. Are you seeing clear differences in performance in the cities that might be better positioned or have more interactions, participation versus those cities and locations that are more restricted at the moment?

Yes. So I think the answer is dependent more on the human perception of the problem than the reality of the problem. I would say the impact on business of the pandemic has been most significant in the Northeast. I think it's been the least impacted in the South and Southwest, so Florida, Texas. I think the impact of the pandemic may actually be worst in some of those markets. But the attitude towards dealing with the pandemic are more laissez-faire, and therefore, there's been less impact. I do think as we continue to see unprecedented hospitalizations and death, pretty much everywhere in the country, people are being cautious in their corporate purchases. I do think as we start to see vaccinations increase and some companies beginning to allow at least voluntary return to office, our corporate growth rate will pick up, and I would suspect that will be within the next 6 months.

And moving over to the NetCentric side, if we can reveal the results of our next survey. And so what are your expectations for average annual NetCentric revenue growth over the next 3 years? So 75% said 4% to 6%; 25% indicated 7% to 9%. And so within this context, maybe you can give us an update. You talked about some of the strength earlier in the conversation. On your earnings call, you talked about the strength continuing of volume in the fourth quarter. And the question I get is in terms of the monetization of that growth and -- is there an opportunity to even better monetize this data traffic growth into more revenue growth for Cogent's NetCentric business?

Sure. The answer is yes. Just to remind investors, there have been periods where Cogent's NetCentric business has run substantially faster than 9%. Yes. That's based both on traffic and volume-weighted price declines. Because we are seeing this proliferation of video services, our customer base has become broader. Our effective price per megabit is declining more in line with our nominal rate of price decline. As a result of that and the increased acceleration of the migration from linear video to streaming video, I think that the 9-plus percent, at least for the next year, is very reasonable. Because this is a usage-based service and highly dependent on the nature of the content that is being streamed, I think 2 to 3 years is much more difficult to predict. But I think in the intermediate window, we will see improved NetCentric performance.

Given the importance to the changing video consumption patterns for the NetCentric growth, is there some kind of kicker from 4K as those TVs get more broadly adopted and streaming services increasingly embrace that? Is that a multiplier effect? Or is that part of the growth that you're already seeing in the numbers?

I think it's embedded in the numbers. So any streaming service has to modify their content to work on the most common user devices. For a company like Netflix that's supporting about 200 different user devices, that seems to be kind of the sweet spot. I think that will be probably true with the other providers as well. We've seen an acceleration of 4K. We've even seen some deployment of 8K in very limited situations. And we've seen some early experiments with augmented reality embedded in video. I think to continue the acceleration of video moving from linear to over the top, we will need a better product, and that will have higher resolution and more augmented content. So I think embedded in this transition and driving market volume growth in the 20% to 25% range, Cogent growth in the 40% range, will be the combination of more content, more minutes a day and higher bit load per minute. And remember, the carriers are constantly under pressure with their linear products that those networks can't be modified the way an IP network can handle these more robust streams.

So our last survey question is on the current opportunity for Cogent to sustain the dividend. And so the question is, can Cogent sustain the current dividend strategy? Yes, expected continued growth, which is currently $0.025 per quarter sequentially; yes, but expect slower growth for a flat payout to the current level; no, expect to cut within 3 years, given slower revenue growth; and no, expect to cut within 3 years given a change in preference for buybacks. And just a reminder, all the responses are anonymous. As people queue in their preferences for that question, Dave, maybe taking a step back, when you look at the portfolio effect of what you just described in NetCentric, what you described in corporate, what's your confidence level that revenue growth for Cogent continues? And how does that growth compare to the current long-term annual goal that the company has?

Yes. So again, let's go back and look at 18 years of revenue history. Our total revenue growth over that 18-year period, organically, has been 10.3% for both segments of our business. Our guidance is to grow at about 10%, in line with that growth rate. We have also been able to deliver, on average, 200 basis points of EBITDA margin expansion per year for the past 18 years. We also have seen our capital intensity continue to moderate. Now with the slowdown in revenue growth, we're growing at about half of that growth rate. We are still achieving close to our margin expansion target. Part of that is the arithmetic of having more on-net sales versus off-net. Part of it is cost controls, some of which we could implement because of the pandemic. The ability to return capital is tightly linked to the ability to grow free cash flow. We have grown free cash flow at high double digits, call it, 18%, 19% over the years. We have 34 consecutive quarters of sequentially growing the dividend. Several quarters ago, we raised the pace of those sequential growth in dividends from $0.02 to $0.025. So there's an arithmetic slowdown in the percentage rate of growth in our dividends just as the dividend gets larger. We have no immediate plans to change that pacing. We have returned about $900 million to our equity holders, roughly $225 million through buybacks, $675 million through dividends. If the stock market went through a catharsis and growth turned negative, we would err more on the side of buybacks. We have substantial excess liquidity on our balance sheet, and we are in the low interest rate environment, where we have substantial unused borrowing capacity. Our goal is to use the balance sheet efficiently for our equity holders, but do so prudently. I think the path that we are on now is the appropriate one. I want to wait and see what the survey says, and then that will give me a chance to complete my answer.

Great. Let's see what the survey says. So yes, but expect slower growth for a flat payout of 75%; and no, expect to cut within 3 years given the risk of slower revenue growth at 25%.

So I think the risk of a cut is very small. I mean listen, a dividend is always at the discretion of the Board. But having 34 consecutive quarters of sequential dividend growth, I think, has only been matched by 8 company -- public companies in history. We view this as an important attribute to our shareholder base. We also, though, remain committed to being very efficient in returning capital. That efficiency in part is driven by the fact that roughly 50% of our dividend has been characterized as a return to capital, reducing the basis for our shareholders but allowing you to receive the dividend in a tax-efficient manner. In fact, we have many shareholders who have held Cogent long enough, they now have a negative basis in their share purchases. If we saw tremendous volatility in the market and our stock's rolled off, yes, we would be much more aggressive with buybacks. Would that jeopardize the dividend? Probably only in the most extreme case. If we saw the type of volatility in the markets that the world saw in 2009, the answer is we may elect to do more buybacks and suspend the dividend. But I think what investors should feel comfortable with is that we are going to be returning increasing amounts of capital to our shareholders. The method will be optimized. And even with the slower growth that we are experiencing because of the pandemic, we are returning larger amounts of capital every quarter.

One of your earlier comments was about the growth of traffic is speeding up some of the capacity deployments that you would normally do for the network. Should investors anticipate significantly higher capital spending, whether it's exiting 2020 or in 2021, to accommodate these investments?

So the short answer is no. We are a beneficiary of substantial improvements in technology. One of Cogent's advantages is that it built a network on a very standardized platform. We constantly upgrade segments of network and move older equipment to more pure flow locations. The network is comprised of those roughly 950 rings that I talked about, roughly 3,100 endpoints, about 604 routers and about 1,800 city pair sets. We are in the process of deploying multiple multi-100-gig lengths. So we went from 2.5 gigs to 10-gig wavelengths to 100-gig, and now we're deploying 400-gig wavelengths. That will continue, but the cost of that optical transport equipment is falling at a compounded rate of 80% a year. The cost of routing equipment is falling at 40% per year. So we feel quite comfortable that we will not see an uptick in our capital intensity, even though we are seeing a material uptick in traffic.

Well, David, our final couple of minutes, are there any other thoughts that you want to leave our audience with on the business and the opportunities in front of Cogent?

Yes. I think the most important thing to remember about Cogent is we are different than other telecom companies. We're solely focused on the internet. If as an investor you believe the internet is the future, Cogent will be there to monetize that. If you believe in legacy platforms, Cogent is probably not the ideal investment for you. Secondly, we have built a network that has a network effect. More networks connect to Cogent. Therefore, creating a better quality of service than any other network that is part of the internet. We have an extremely efficient distribution model with our direct sales force, and we have very high operating leverage. Those factors combined make Cogent a good investment. Fast growth, high operating leverage, low capital intensity and a disciplined allocation of capital. We remain focused on building value for our equity holders. And rather than a lot of companies projecting some kind of hockey stick in growth that is yet to materialize, all we ask of our investors is to believe that the linear growth that we've exhibited will continue going forward. And if that in fact happens, our ability to grow cash flow per share will continue for the next decade, allowing us to materially increase that cash return to investors from the roughly $3 a share today to anywhere of $10 per share. An investor bank can value what that $10 of cash per share is worth today.

Dave and Sean, I want to thank you for joining us today. I appreciate it. And it's great to see you, albeit in this virtual environment, and look forward to the opportunities to doing this in person in the future.

Next year, Mike, we'll be back in Vegas. Thanks. Thanks for all of this.

Thank you, and thank our audience. Thank you.