Cogent Communications Group's CEO Presents at Morgan Stanley Technology, Media & Telecom Conference (Transcript)

| About: Cogent Communications (CCOI)

Cogent Communications Group, Inc. (NASDAQ:CCOI)

Morgan Stanley Technology, Media & Telecom Conference Call

February 26, 2013 8:00 pm ET


Dave Schaeffer – Founder and Chief Executive Officer

Unidentified Analyst

Good afternoon everybody and this is [Mike Grimes] pleasure to introduce Dave Schaeffer to finish up today with a very strong, coming off a very strong quarter, very strong year. Before I hand it over to Dave, please note that all important disclosures including personal holding disclosures and Morgan Stanley disclosure appear on the Morgan Stanley public website at or at the registration desk.

So perhaps you can just talk about as you come off a strong 2012, you talked about some really excellent traffic growth in Q4. I think we’ve heard at the conference over the last couple of days that it feels like whether it’s mobile or whether its data centers, now those trends are strong as ever. How do you think about your priorities for 2013?

Dave Schaeffer

Sure. So first of all, I want to thank you all for staying late in the day 5 o'clock West Coast Time, that’s impressive and I would like to thank Morgan Stanley and Simon for inviting us.

In understanding our business, traffic growth is a key driver. However, for our corporate business which represents 51% of our revenue, that business only generates 3% of our traffic. So other than motivating customers to buy a connection from Cogent, as those customers increase their traffic, it does not have a material impact on our revenue. 49% of our revenues are very traffic sensitive. They are the sale of Internet connectivity in about 650 data centers around the world where we are selling bulk internet connectivity to other service providers.

Those service providers are bifurcated into two main groups, we sell to content companies who are looking to push an application or some form of content out to the public internet and we sell to access network operators that do not have their own global backbone. That business is 49% of our revenues, 97% of our traffic.

In the most recent quarter, traffic on our network grew sequentially 27% quarter-over-quarter. We today carry about 20% of the world’s internet traffic. So may be as a bench line just to understand it what internet traffic is could be helpful to investors.

Today, the global internet is about 350 petabytes a day of traffic. That traffic travels on average 2700 miles and will go through 2.4 networks from origin to destination and will traverse 8.5 routers in its path. Cogent provides that combination of transport, routing and interconnection as the most interconnected network in the world. That traffic growth last quarter resulted in our service provider or net centric revenue growing sequentially quarter-over-quarter 6.3%. So as Simon pointed out, that was a great way to end the year.

We clearly have 31 out of 32 quarters as a public company of sequential organic growth, because of the high operating leverage in our business we have been able to deliver better than 200 basis points of EBITDA margin expansion every year since the company has gone public for eight years. We expect going forward to grow 10% to 20% a year and expect our margins to expand next year by another 200 basis points.

So, good 2012 and hopefully we are off to a good start this year.

Unidentified Analyst

When you have that much market share, presumably what you are seeing on lot of those is market expansion. Can you give us some sense of what you think is going on in Q4 which you have some big wins, was there some seasonality, or you think it was literally from Netflix and other kind of drivers of traffic?

Dave Schaeffer

I think it’s a little bit of all the above Simon. So the internet has grown at about 30% year-over-year for the past half a dozen years. A number of independent third parties monitor their traffic growth, whether it be Cisco, TeleGeography, Renesas all measuring aggregate traffic growth. We have historically grown at about twice the pace of the Internet, that’s the good news. The bad news is, that core internet transit market is a $1.5 billion addressable market that’s actually been flat at that scale for a dozen years. Via Cogent have grown our service provider business this net-centric business at a revenue rate of about 15% a year, by capturing market share.

So our growth is coming from three factors, existing customers continuing to grow, secondly, we continue to win number of customers. So the internet is a network of networks, well there has been about 42,000 unique network numbers issued or net autonomous system numbers. The internet is comprised of about 6000 of those AS’s account for 98% of the traffic. 4,400 of them directly connect to Cogent that’s more than any other network in the world. Today we carry about 20% of that 350 terabytes were about 70 terabytes a day.

So we gain share because customers realized that we are delivering a equal quality product at a lower price point.

The third way we gain share is by customer shifting portions of their wallet spend away from other providers to Cogent. Simon mentioned Netflix, Netflix is in the process of decommissioning or stopping using third-party CDN providers and building their own network. As they build that network, they need by upstream internet connectivity and Cogent is the most cost-effective, the most reliable, and most ubiquitous and they are one of our many customers.

Today, if you look at the global internet, about 75% of the total bit volume of the internet is video, over 90% of the bit volume growth is coming from video. So as we look at 2013 and beyond, the key shift in traffic will be the roughly 300 minutes a day of video consumed in developed world per capital moving away from satellite, broadcast, DVD and cable to over the top. Today, only about 15 minutes of that 300 minutes is over the top. We believe over the next five years to 10 years that will shift and eventually become about 250 minutes of the 300 minutes being over the top video that will drive traffic growth on the internet and therefore drive revenue growth for Cogent.

Unidentified Analyst

We talked a lot here about LTE growth and one thing that LTE does do is it allows video in a way to 3G its pretty constraint, but it's very expensive $10 a gigabit or whatever. How you lever to the growth in mobile data and to the deployment to these networks.

Dave Schaeffer

So, first of all, mobile data growth is much faster in percentage terms than wireline growth growing more at 70%, 80% per year as oppose to 30% for the internet. However, it’s growing from a very small base and you are absolutely correct that the cost per bit is extremely expensive. Anything that brings more traffic on to the internet is a net benefit for Cogent. Many of the LTE operators do not have their own global backbones.

Clearly, three largest providers in the U.S. AT&T, Verizon, and Sprint carry traffic on their own networks and therefore we will only get paid at best on one side of that transaction, the content provider pushing on to the internet. For some of the smaller carriers and regional providers around the world, they do not have their own backbone and represents the opportunity for Cogent to be both the upstream provider for that access network as well as the provider for the content company.

But I think consumers are at the early stages of going through a learning curve understanding there is one thing you can do with a wired network and a very different set of characteristics and features of a completely mobile network. Most of the data that is being consumed in a portable environment is actually not mobile data, but rather data that was transmitted via a fixed line network and then uploaded to a portable device via Wi-Fi. You load it on to your tablet; you take it with you and view it. That’s very different than watching real time.

I don’t want to rein on the LTE guys pray, but if all 300 million consumers in the United States wanted to watch video real-time that would be not enough spectrum independent of what technology and which network you use, it will be far too expensive and the rationing mechanism will be price.

So when you talk about $5 or $10 a gigabyte on the wireless network you're talking about tens of a penny on wireline network. And consumers are rational economic being and they understand that completely mobile communications needs to be for only those time sensitive applications and for kind of synchronous communications not for using the LTE network to download last weeks episode of the Sycamore TV show.

Unidentified Analyst

You touched on your margin expansion and you have this fairly steady 100, 200 basis points a year, can you just talk through your network and your ability to scale back how they particularly on-net margins (inaudible)?

Dave Schaeffer

So, our network stretches from Seattle, Washington to Kharkov in Ukraine from Helsinki, Finland to Mexico City. Roughly 56,000 miles of intercity fiber that has been directly connected to another 26,000 miles of metropolitan fiber in a 180 markets around the world.

Unidentified Analyst

Mostly IR using?

Dave Schaeffer

Completely based on IR use, but those are into feasible rights of use. We have bought the dark fiber not net services, but dark fiber typically prepaying for it for a 20 to 40 year term and then we own the optronics on that fiber. We own the routing equipment and then we build the connections from the metro network into the building up the riser. So today, our network serves two types of buildings, 1300 skyscrapers. The average building is 550,000 square feet it is 41 stories tall and would have 51 unique tenants in it. That business is only in North America, U.S. and Canada and represents just under 12% of all of the rentable office space in North America. There are million commercial buildings we have gone out and cream skin of the market to go after only the buildings that have the greatest customer density per building.

The second type of building we connect to are data centers. We have 43 Cogent owned data centers, about 425,000 square feet of raise or we have another 550 data center buildings that have 650 unique data centers in them. Within those data centers we sell to the service providers. That network allows us to touch 98% of the global internet demand or market share. And on that network, we are running a network that is substantially different than any of our competitors. The network was built from the ground up to only deliver internet, we have a narrow product set [HQ’s] we’re running IP directly over DWDM, we are protecting the network of layer 3, we’re using internet as an interconnection technology and we are the most interconnected network connecting to over 4,400 unique networks around the world.

With that every time we add $1 of our net revenue, we have no direct cost of goods sold. We have a 100% gross margin contribution, $0.95 of EBITDA margin contribution. When we add $1 of off-net services to our network, 20% of our sales are to branch offices that are in buildings that don’t justify our capital expenditure to extend the network into those buildings. We have to buy the local loop from the incumbent. We have a 50% cost of goods sold, $0.45 of incremental EBITDA.

With this high operating leverage, last year we did about 34.5% EBITDA margins on about $350 million of revenue. We expect our revenue to grow roughly in the middle of our guidance range 10% to 20% called 15%. And with those incremental margins, you should see our EBITDA margins not only next year but the year after and year after that continue at least 200 basis points a year of margin expansion.

The final piece of the equation that’s relevant to an investor is the capital intensity of this business. Most telecom service businesses are far more capital intense than Cogent. We actually spend less on capital than any of our competitors. We are actually in absolute terms not just as a percentage of revenue have enabled to decline our capital spending. The net result has been, we have generated true free cash flows and not only we’ve been able to grow top line in the mid teens, we’ve been able to deliver a free cash flow growth of close to 30% per year and we’ve been able to use that free cash for a variety of purposes.

We have bought back 15% of our outstanding float, 7.7 million shares out of a maximum total of 53 million shares. We have bought back 55% of our outstanding convertible debt reducing that from $200 million to $92 million. And then finally we’ve issued a dividend that each quarter leadership we’ve increased it sequentially every quarter.

Last quarter, in addition to our interest payments, in intention to our capital lease payments, our capital payments, CapEx expenditures, and our dividend we grew the cash on our balance sheet by about $10 million for the year. So these are great problems to have and we've been committed to returning that cash for shareholders through both buybacks and a growing dividend policy.

Unidentified Analyst

It's nice to see the growth coming with the cash flow as well. Given that low capital intensity you're buying what is the biggest barriers to entry for competitors, that really is your scale and just the pricing that you're able to maintain that just hasn't enable due to (inaudible)?

Dave Schaeffer

So sometimes it’s good to be in the right place at the right time. We raised $500 million in the late 90s and we're walking of not to have spent all of that money fast enough. We were able to then repurpose the money to buy a number of distressed assets at a deep discount. We took those distressed assets, $14 billion of invested capital, $4 billion of PP&E and repurpose them into the Cogent network. But if that was the whole story, once those assets were depreciated, our business would have no competitive advantage.

Our advantage goes beyond our balance sheet, the companies we compete with are anywhere from 20 times to 100 times our size, yet we deliver 1000 basis points better EBITDA margins. So that has to be more to our competitive advantage and it’s really the combination of factors. One, the network architecture that we’ve used; two, the very targeted approach to deployment of capital on high traffic locations; three, the operating efficiency that comes with a very lean set of products that are with the products that are in the rapid growth part of the market. [Meanwhile] it is a sales and revenue acquisition model that is different in the industry.

So most of the competitive datacom and telecom industry failed because their subscriber acquisition costs were greater than the life cycle value of the customers they were acquiring. We have a cost of revenue acquisition that’s once fixed the industry average. I wish I could say this because my sales force was six times better. I hope that’s partially true, but I think the right answer is, we’re just delivering better value to our customers. So our marketing strategy which is very controversial is on the corporate side gets the customer 65 times as much capacity and for the same price is the incumbent.

On the service provider side, go-to-market a guaranteed undercut whoever else is selling by 50%. Remember, we’re selling a product that has no direct cost of goods sold, and we’re only utilizing 20% of our lead capacity. With that model, we have lowered our cost of revenue acquisition to less than $2 to get to $1 of monthly recurring gross marginal on net, with an average customer life of about 57 months. The industry average is to spend $13 to get $1 of revenue, 50% gross margin and average customer life of about two years. That's an equation that doesn't work. All of these factors cumulatively give us an operating advantage. We then have a cost structure and a balance sheet that fits our market, and we’ve used these advantages to gain scale.

Today, in our corporate buildings, we have 10.5 customers out of 51. The only party that’s even close to us is the incumbent. On the service provider side of our business, we have 20% market share. We are the second largest provider in the world. We continue to gain market share and that becomes a sustainable barrier to entry. So whether it would be the fiber we’ve assembled, the technology we’ve used, the targeted buildings, each of those some level could be replicated. What could not be replicated are the interconnection agreements and traffic scale that we have that became a material barrier to entry.

And finally, it's important for people to understand this is a $1.5 billion core addressable market. Investors made the mistake last time in investing $1 trillion to chase a $1.5 billion market. I don't think they make that same mistake the second time, and with that I think it's highly unlikely that we're going to have a direct competitor come behind us. For those reasons, I think Cogent is truly uniquely positioned.

Unidentified Analyst

Excellent. Okay, we have got time to some questions from the audience.

Question-and-Answer Session

Unidentified Analyst

Can you just talk about that 1.5 market, I mean you have growth obviously been with growth and obviously some price decline, is that the reason why it’s not growing, it’s been with offsetting price declines from most of your competitors.

Dave Schaeffer

So first of all, it was a market when we entered it that was much more crowded than it is today. In 2001, when we started selling service, there were 200 networks that claim to be global backbones. Today, there is a dozen of those networks. The unit volume growth during that period has been about 30% per year. The price decline for the market has actually been better than 40% per year. The process has been very Devonian meaning a lot of less sufficient operators from the market either through bankruptcy, through merger or just plain old attrition.

We have been able to be the price leader that whole time. Our go-to-market price in 2001 was $10 a megabit. Today, our average new sale as of the end of the fourth quarter was $1.80. So our price has fallen, but its fallen less dramatically than the market average which has gone from $300 a megabit to about $4 a megabit. We continue to be at the forefront of leading price.

Most of our competitors are vertically integrated telephone companies. The dirty little secret in our industry is telecom service companies hate the internet. The internet is very deflationary to their business models. It separates the application layer from the network layer. That dichotomy has created a tension in which those companies have intentionally either under investment in their internet who are in worst case were taught at the quality of their product before they’ve exited the market. We have been a pure play and because we offer the same product in a same location with equal or better quality and half the price, we’ve been able to grow twice as fast as the market, and we’ve been able to experience price declines that are about half of the market rate of price decline. Those two factors have allowed Cogent to grow its revenues at the midpoint of its guidance while the market has stayed flat.

Unidentified Analyst

If you look just at the Ethernet market versus those TDM plus Ethernet, what type of growth rate you’re seeing in Ethernet market?

Dave Schaeffer

So Ethernet is a technology, it’s a type of interface. Starting on a machine-to-machine communication, Ethernet is by far and way the dominant way in which information goes from one computer to another. Over 99% of traffic on machine-to-machine is Ethernet, there’s a little bit of Fiber Channel and ESCON still in the mainframe market, but its progress of 1% of the market. Today, there are about 6 billion ports of Ethernet in the world. So Ethernet is a huge business and the cost of those ports thus far on below $5 a port, it is the ubiquity standard.

Ethernet is often times used to describe a interface that is then used for point-to-point connectivity, as oppose to point to internet. So for Cogent the product we sell is from a single point to the rest of the world internet, without any specific end locations predetermine. What most other providers sell in their Ethernet portfolio is using Ethernet as the interface technology from point A to point B. That market today is a several billion dollar market, it is growing, but it is growing because of the replacement of TDM or wavelength services on a point-to-point basis.

Ultimately most of that traffic will migrate to the public internet. The key trend in telecommunications is twofold, unit volume growth driven by applications and ubiquity of the internet coupled with a much lower cost per bit mile on the internet. So historically, people who had private closed networks are increasingly comfortable to build VPNs on top of the public internet. The bet that Cogent made was that it could build a viable business selling only internet connectivity. That bet has turned out correct, the companies are selling point to point services are chasing what is ultimately a market that will shrink.

We got one more.

Unidentified Analyst

Yeah, we will make this last question.

Unidentified Analyst

I actually had two questions for you. One is on that $1.5 billion market, it sounds like your pricing is less than 50% of what others pricing. And it sounds to me like it's not a enterprise or SMB market. It’s a market – its commodity market people can see price transparency why aren’t you winning share faster than you are today. And then the other side, I wanted to ask about capital intensity and what you think of it maintenance CapEx for the network and if you would ever go if you have the network you need or the backbone you need or you would never need to go out and invite buy new backbone fiber?

Dave Schaeffer

Sure. So let me take the two question separately. First of all, in terms of the market, our biggest competitor is [Inertia]. The market used to have 200 players but still got a dozen every one of the companies as larger than us and our household brand names. We as a relatively thinly capitalized start up. Remember, we raised $500 million compared to AT&T or Verizon or Sprint or even level 3, that’s the de minimis amount of invested capital and yet we’ve been able to achieve 20% market share. We’ve done that with no branding campaign, no advertising, but through a direct sales effort. We believe we can grow that market share.

Secondly, we use price as a competitive tool. Investors struggle with pricing models for a product that has no direct cost of goods sold. Today, we are utilizing about 20% of lid capacity in our network. 80% of that capacity that’s lid on our network goes follow everyday, it’s not used. With that type of dynamic, we look to price to whatever it takes to win the business. So the interesting dance that goes on between Cogent and its net-centric customers, our goal is to get as much money from them as possible and we are kind of indifferent to how many bits we have to give them.

Now they think of the world in a price per bit that’s the way our competitors have trained them, but our goal is to give you as many bits as you consume just give us as much money as you are going to spend for that service. That dynamic is not a transparent just we would like it to be. We use a variety of price discovery mechanisms including shows an end voice more undercut of that 50%, a less price program and a set of targeted promotions. All of those things have allowed us to gain market share, probably do better possibly, if we done well, I think pretty well considering what we've had to work with. And our business is balanced, that's half our business, the other half is just kind of specific access business that is screamed skimming a very specific corporate demographic. Those two businesses work together very symbiotically, allow us to use our assets efficiently, and generate superior growth and superior cash flow.

Now to your second question, as I’m already two minutes over my limit here, and it is late in the day, we have a network that spans 36 countries; there is 198 countries in the U.S. The reality is most of those countries don't generate enough traffic, most of those countries don't want an open internet, most of those countries don't have infrastructure that has cost effective for us to buy.

So for the most part, we have completed our network. Could there be mild or small instances where we expand yes, now on the network we have we are running an average of 60 to 10 gig wave lengths. The technology we can support today can run 160, 100 gig wave lengths, so we can connect the throughput for way three X number of waves, we will consistently ad capacity to our footprint, but that's does not require us to buy new fiber. The cost to transport a bit since 1985 has fallen at 80% per year, I saw [JDSU] was presenting down the hall you could hear what their cost per laser has fallen. And then the cost allowed a bit as fallen at about 40% per year.

Our job is to incorporate those improvements in technology and our footprint and to capture part of that benefit as profit in past part of on to the consumer and that's kind of the art in praising our service. But I feel very comfortable based on the structural changes in the industry, the secular growth in internet traffic, we will deliver the top line growth that we’ve laid out and the margin expansion. And I ask investors not to take what I’m saying is a leap of faith, but rather go back and look at what we’ve done at these eight years as a public company with a demonstrated track record and just project from that linearly into the future.

Again I want to thank you all for staying late today. I want to thank Simon for hosting.

Unidentified Analyst

Thank you for coming. I appreciate it.

Dave Schaeffer


Copyright policy: All transcripts on this site are the copyright of Seeking Alpha. However, we view them as an important resource for bloggers and journalists, and are excited to contribute to the democratization of financial information on the Internet. (Until now investors have had to pay thousands of dollars in subscription fees for transcripts.) So our reproduction policy is as follows: You may quote up to 400 words of any transcript on the condition that you attribute the transcript to Seeking Alpha and either link to the original transcript or to All other use is prohibited.


If you have any additional questions about our online transcripts, please contact us at: Thank you!