Cogent Communications Group's CEO Presents at Bank of America Merrill Lynch 2013 Smid Cap Conference (Transcript)

| About: Cogent Communications (CCOI)

Cogent Communications Group, Inc. (NASDAQ:CCOI)

Bank of America Merrill Lynch 2013 Smid Cap Conference Call

May 8, 2013 12:25 PM ET


David Schaeffer - Founder and Chief Executive Officer


David Schaeffer

Good day. I think we're being webcast, so please continue on. But I want to thank you for taking the time to hear about Cogent. What I'd like to do is spend a few minutes giving you an overview of the company, and then we can kind of open it up for questions-and-answers, if that make sense to everybody.

I don't know how familiar you are with the company so at any point feel free then or opt Safe Harbor language. I think from the Safe Harbor, let me describe a little bit about what we do, who we sell to and what we sell.

Cogent is a provider of dedicated internet access service. We operate one of the world's largest IP networks. It's a physical network that serves 36 countries, 26,000 miles of intercity fiber connected to an additional 26,600 miles of metropolitan fiber. On that network we serve 181 markets, about 1,900 end-locations, 36,000 customer connections. So we're clearly approximately 18% of the world's internet traffic. We are solely an internet provider and we win business by being the best value to our two respective customer bases.

51% of Cogent's revenues come from our corporate end-user customers. These are customers that buy the internet to support their businesses in large multi-tenant office buildings, 49% of our revenues come from selling to other service provider. So we're selling bulk internet connectivity on a per megabit basis to other service providers, who could either be content companies or access network operators in carrier neutral data centers.

Our network is solely dedicated to delivering internet, it's a physical network that stretches from Tokyo in Japan to Kharkiv in Ukraine, form Helsinki, Finland to Mexico City. It's a network has served 182 metropolitan markets with about 570 metropolitan rings of fiber. Connected to that are 1,300 multi-tenant office buildings and an additional 650 carrier neutral data centers.

Our network touches approximately one-tenth of 1% of the office buildings in North America that actually serves about 11% of the square footage. We connect to 720 million square feet of multi-tenant office space. The total inventory of office space in North America, U.S. and Canada is about 6.4 billion square feet. And then finally, we serve approximately 4,600 customers in locations that are too far from our network or too small to justifies building fiber and we connect to about 4,000 buildings to serve those 4,600 customers.

Now, why does a customer choose Cogent? For our corporate customers, what we are delivering is a substantially better product. The customers' getting 65 times as much bandwidth, it's on a network that's completely independent from the phone company. Its fiber, not copper; it's ring protected as oppose to point-to-point, and handoff to the customers' internet.

So the most common product we sell to our corporate customer is a 100-megabit per second, non-oversubscribed and non-blocked dedicated internet access service for a flat monthly fee of $700 a month. We typically compete with the incumbent phone companies in our market. So AT&T and Verizon are our two most common competitors; CenturyLink and Bell, Canada. This corporate business is exclusively in North America.

In our service provider business, we are selling bulk internet connectivity to other network operators in a carrier neutral data center is effectively a supermarket for bandwidth. We sell our product at a lower price than anyone else in the market. In fact, we have a standing offer in the market to undercut our competitors in this space by 50%. We win business in the net-centric or service provider market solely because of price. We're selling an undifferentiated, functionable commodity at a lower price.

We see in terms of competition, we pick what we see across our footprint Level 3. On a geographic basis, we'll see Telia, Telecom Italia and Telefónica in Europe; NTT, PCCW and TATA in Asia; and here in North America occasionally AT&T, Verizon, and Sprint would be our competitors.

So in our corporate footprint the most common product we sell is a 100-megabit per second non-oversubscribed and non-blocked dedicated internet access service for $700 a month, where typically we're pricing that copper T1 or a Paris copper T1. The customer will spend about the same with Cogent to get a vastly superior product. Our typical corporate customer will use approximately 11% of the bandwidth that they purchase at peak.

All of our growth in this market comes from adding additional customer connections. It is extremely rare for our customer on the corporate portion of the business to outgrow the bandwidth that is purchased.

Now, in our corporate business roughly 60% of this half of our business or 30% of the total business is the corporate on-net business, where we're selling the 100-meg connection in the multi-tenant office building. Often times these customers will have a branch office in a building that we use a third-party to help us serve, and we're selling an off-net circuit. In that off-net sale, the value proposition is not nearly as compelling, we're reselling someone else's loop, we're adding the IP wire to it. We do compete in that markets, usually again with the incumbent and some times the cable companies.

Within the datacenter, the companies that buy from us are purchasing bandwidth to incorporate into the final good or product that they are selling. This market is characterized by customers, who buy the megabit and volumes go up, prices go down. So to understand the market, the market unifying growth has averaged about 30% per year for the past eight to 10 years. We have been able to grow at about twice the rate of the underlying market on average.

In fact, over the last year, our year-over-year traffic growth was up 91%. This market has also characterized by decline in prices. The average price per megabit in the market has fallen at a rate of about 40% per year for a dozen years. The numbers of players has shrunk from over 200 to about a dozen today.

Our price during that period has fallen from $10 a megabit to the most recent quarter at $3.05. On a year-over-year basis our installed base declined on a per megabit basis by about 24%. Our average new sale in the quarter was $1.70. We report these numbers consistently and transparently each quarter.

We have been able to grow our revenues, in what is a flat addressable market. So what is often being surprising to investors is that the total dollar value of the core internet transit market that is the 400 petabytes a day of bandwidth, that is purchased at the core of the internet out of the 650 data centers is a $1.5 billion addressable market and it's been flat at that level for a dozen years.

We today have approximately 12% of the market by dollars. We have about 28% of the market by bits. And maybe helpful at this point, to explain a little bit about what internet connectivity is. The product we sell, which is known as dedicated internet access or internet transit is really the combination of transport, routing and internet connection.

The average bit on the public internet travels 2,700 miles and goes through 2.4 networks from origin to destination. It goes through 8.5 routers in that path and we'll typically have one ingress and one egress point on the network. We bundle that combination of transport, routing and internet connection into a single product, and sell that at the lowest price in the market on a per megabit basis.

We today are the most interconnected network in the world. So the internet is not a single network, but rather it's a network of networks. While there have been about 42,000 autonomous systems, where network numbers issue for people or organizations to run their own network, today about 6,000 of those networks makeup in excess of 98% of the public internet.

Of those 6,000, 4,400 in any of those networks are directly connected to Cogent, I man they go through a no third party. That's actually substantially more than any other network. Probably the next closest to us would be the Level 3 at about 3,900 connected networks. Then it actually drops off pretty significantly to companies like Sprint, AT&T and Verizon that are around 1,200 networks connected to them.

Now, of those interconnections, they come in two forms. Either a settlement-free peer, meaning the two parties chose to exchange traffic without money exchanging hands. We in fact have 40 of those settlement free agreements with other major network operators. And then, the second form is where network operator pulls another network operator for transfer. We pay no third party for connectivity or transit. In fact, 4,440 of the 6,000 network choose to buy their upstream of buy connectivity from Cogent.

Now, this has resulted in the company being able to deliver working on a sequential growth for 31 out of 32 quarters as a public company. Our growth has averaged about 13% year-over-year during that period. In fact, our equipment growth has grown every single quarter 32 out of 32 quarters sequentially. During that period, we've been able to deliver roughly 55% traffic growth on our network.

Our service provider revenue is more volatile for a couple of reasons. 40% of that revenue is outside of the U.S., so we're susceptible to foreign currency fluctuations. The second reason for that volatility is that our business is somewhat seasonal. Typically people use more bandwidth in the winter months, core months, than they do in the summer months when they are outside moving around. And then finally, there are certain business models that are succeeding, where others have failed in a dynamic ecosystem.

So we do have some volatility to one quarter, in which the company had negative topline growth, was actually Q1 of 2012. And the reason for that large negative was that the U.S. government shutdown our largest single customer, who was sole source to Cogent, Megaupload, who accounted for 4% of world's traffic by themselves.

Absent that, we've had a consistent pattern of growing and we've been able to grow this revenue consistently absent Mega. And in fact, in the most recent quarter, our corporate revenues grew sequentially 1.7% and our net centric or service provider revenues grew on a quarter-over-quarter basis by about 3%. Total topline revenue growth for Cogent was 2.3% quarter-over-quarter.

The slide indicate some of our pricing trends. Price per megabit is falling and will continue to fall forever, due to the improvement and the underlying technology necessary to deliver this service. Our on-net ARPU have been relatively flat recently and our off-net ARPUs have actually increased, but the size of those off-net collections tends to get larger over time.

Now, in our corporate business we sell to legal, general corporates and financial services, those will be our mean of vertical. Today we serve 85 of the 100 largest law firms in the world, as their primary internet service provider. We serve almost 1,400 hedge funds. We serve virtually all of the major private equity funds. We serve investment bank, mutual funds and general financial services companies.

And then finally we have a mix of high-end corporate users that typically locate their business, in the most expensive real estate in the city, skyscrapers in the central business districts of major city. So whether they'd be a architectural firm, a management consultant company, a software company, those would be our typical general corporate users.

On the service provider side, we serve today about 1,200 colleges and universities globally. We serve virtually every major access network operator that does not have a global backbone. So we serve ComCast and Charter, Cablevision and Cox, we serve both Canada and British Telecom, but we also serve some international players. We serve almost 80 PTTs or incumbent phone companies around the world, from Telecom Iraq, Telecom Malaysia, Telecom South Africa, Telecom Egypt, Pakistan Telecom, to very small companies such as Telecom Liechtenstein or Telecom the Vatican.

We also sell to a number of the competitive providers, Iliad and SFR in France, Jazztel in Spain, EarthLink here domestically, would be examples of some of our other service provider customers, and the final group of service providers are the content companies. So we sell through CD as well as selling to company that publish their own content or applications on the internet. We sell to Microsoft, to eBay, to Yahoo, to Google, to Apple and Net Plus, and directly with real network, MPC, CDS, Viacom are examples these companies.

So we can have situations where we are providing only the access network operator and not the content company. We can provide just the content, not the access or we can do both, where we are being paid both by the content company and the access company. So example would be, if you're playing an electronic hearts game and you're on Comcast, chances are we're being paid by both parties.

If you buy a Netflix movie and you're using Charter, as an example, being paid on both; whereas if you are a Verizon customer, buying a Netflix movie, we would only be paid on the Netflix side not on the Verizon side. So roughly three quarters of our revenues are on-net, a quarter of our revenues are off-net.

The most common on-net product for the corporate customers is the fast internet; for the service provider, it's the gigabit connection. Our off-net services are a combination of Ethernet as well as TDM. And then finally, about 1% of our revenues come from non-core products that we no longer sell that came to us since through various acquisitions over the years.

This slide gives you an indication of the breakdown of our business by geography, by market segments, and probably on the right side the most important breakdown is on-net versus off-net per margin. And we'll talk about that in a moment. Roughly three-quarters of our revenues, 73%, are in North America. The remaining revenues, 27%, are outside of the U.S. By traffic we're roughly 55% North American and 45% rest of the world, which pretty closely mirrors the internet.

If you look at our market segments, 51% of revenues come from corporate customers. The corporate customers only generate about 3% of the traffic on our network. So 49% of our revenue accounts for 97% of the load on the network, each of the service provider customers that tend to buy large connections, and then fill them up and use them very efficiently.

The final chart on the right is the breakdown of on-net versus off-net. With roughly three-quarters of our revenue is being on-net and about a quarter of our revenue is being off-net. Our incremental margins are quite different. And we'll talk about that in a moment.

The principal network that we operate is the world's largest single IP network. It's a network that connects to the highest traffic locations. We have 1,300 plus multi-tenant office building, another 575 data center building representing about 650 unique data centers. We also connect to 43 Cogent-owned data centers. These are our own facilities, where we have backup and power available to our customers in a footprint that represents about 420,000 square feet.

Across that network, if you cut through any of those blue lines, on average you would see about 56 10-gig wavelength across the network. This capacity is used entirely for delivering the internet. So we are not a phone company that is trying to carry internet traffic, we're not a cable network. We have built a network from the ground up solely to deliver internet.

So what makes what Cogent's network very different than others is that it is a network that resembles a corporate land on a global footprint that's running IP-over-DWDM, that's protected at Layer 3 and it uses Ethernet to interface with a customer. On that network the only products we sell are internet. We're not selling wavelengths, we're not selling voice, we're not selling ATM, we're not selling linearly delivered television services.

So our network culminates in two types of structures. We go in to very large multi-tenant office buildings, in fact the average building that we connect to is approximately 41 stories tall, 550,000 square feet, and would have 51 unique tenants. To that building we dedicate on day one 2-gigabits of bandwidth, and we grow that as building needs additional services.

Every customer in the building gets a non-blocked and non-oversubscribed service meaning they get a connection that is 24/7 dedicated to them and they pay a flat rate whether they use it or not. We own the fiber into the building authorized all the way to customer suite. So our definition of the on-net system is literal definition.

We also connect to data centers. Within those data centers, our customers house their server equipment or their aggregation equipment and then they buy a cross-connect to Cogent at which point we then sell them bandwidth by the megabit that they then either publish their content on or use our network to provide connectivity to their end-user customers.

So the Cogent network is very different than our competitors. It's a network that is built out of IRUs, these are indefeasible rights of use. We have purchased fiber from someone else. We did not build the fiber. These agreements typically are 10-to-40 years in term. The average remaining life on these agreements is about 20 years. We've been paid a fixed maintenance fee monthly on this fiber.

On the fiber we own outright all of the equipment that we've deployed, a combination of dense wave division multiplexing at the core of the network and coarse wave division multiplexing at the edge of the network. The network is always configured as the series of interlocking rings. We own the network into the building, our provider. We own our own data centers and we own our own central offices. And then, on that network we connect two other networks most of which in fact 99% of our network interconnections are to our customers.

Now, how do we get to this point? Cogent was very lucky. We raised $500 million of seed capital in 1999 and we were lucky enough not to spend that capital before the market turned down. When that turndown occurred, we went out and bought distressed assets. So between April of 2001 and December of 2004, we did serious due diligence on our 121 acquisition targets, we bid on 19 and we were successful in buying 13 companies, six public, seven private.

Those companies we bought, the six public companies at peak had a market capitalization of $62 billion. The 13 companies, in aggregate public and private had raised $14 billion in invested capital and $4 billion in property plans and equipments. We paid $60 million. We acquired $150 million cash, $400 million of acquired preferences and $500 million of acquired debt. We restructured the balance sheet, never went bankrupt, but equitized the preferences and the debt.

We then generally turned off the operating networks in companies we purchased, fired the employees, fired the customers, turned-off the product and then redeployed the physical assets of those companies into our architecture, creating the world's largest IP network.

So in 2005 when we became a public company we had a big network, we had no debt and we had very few customers. What we've been able to do is leverage that asset in a way that is focused on the fastest growing part of the market. Today, we are using about 19% of the lit capacity in our network. We have a very simple product portfolio selling only eight skews or different versions of dedicated internet access.

We target the most traffic-rich locations and we understand that what we're selling is commodity and we need to be in the low cost provider and provide our customers the very best value in the market. We have been able to do that and ensure that our cost of revenue acquisitions is about one-sixth the industry average, not because our sales people are necessarily six times better than our competitors, but it's because the value we deliver to each of our respected customer base is so great.

So the key to Cogent's success since going public has been our sales force. And it will continue to be the major focus at the company. Today, approximately 355 of our 620 employees are in the sales organization, 262 of them have the direct quota responsibility, the remainder being sales engineering, sales ops and sales management.

We continue to grow that number in the quote-bearing part of our sales organization. We actually report and have tracked our sales force productivity, both in terms of dollar spend to acquire a dollar to monthly recurring revenue and also report the number of unit sold per rep. In fact, in the most recent quarter, our sales force productivity increased another 10% and we have actually tied for the highest productivity in the company's history completely organic.

This is where we spend the bulk of our time, is focusing on the sales force, growing this organization and adding to it. And as we go forward we expect our sales force to continue to grow, yet we have some of the lowest SG&A in the industry.

Our topline growth has grown at 12%, but that included the loss of Mega, without Mega it's about 13.5% CAGR over a year period. Because of the high operating leverage on our business, our EBITDA margins have expanded over an eight year period at an average annual rate of expansion of 220 basis points a year.

Investor should expect on a going-forward basis that the company will grow its topline 10% to 20% and deliver at least 200 basis points of margin expansion. The results has been our EBITDA has grown nearly 40% over that period. Now, when we add a dollar of on-net revenue, remember three quarters of our revenue is on-net, we get a 100% gross margin contribution, $0.95 of EBITDA. When we add a dollar of off-net revenue we generally have to buy the local loop at a 50% margin contribution growth, up $0.45 of EBITDA.

We have used that capital for a variety of things. We've bought back 15% of our flow. We've brought back 54% of our convert. We've actually expanded the network going from 70 markets to 182 markets. We've increased the longhaul footprint over doubling it since going public, and increased the metro footprint nearly tripling it over the same period. We are almost complete or are at the end of that expansion phase in the business.

The growth has come also from adding buildings, but within those buildings increasing the penetration within those buildings i.e. the number of customers that we serve. And we expect the majority of our growth going forward to continue from our increased penetration in the footprint we're in.

Now, we're able to do this with declining capital intensity. So our CapEx that is the amount spent on prepaid capital lease principal payments. And straight capital has declined for the past three years, $72 million, $62 million, $61 million, and it will decline again this year.

The result of this capital efficiency has been a substantial growth in free cash flow. Our topline revenues grew 2%. We typically get some seasonal SG&A costs in the first quarter, so our margins were essentially flat and our EBITDA was essentially flat, but going forward you should expect to see margins continue to expand. In fact, on a year-over-year basis our EBITDA margins on the first quarter were up 430 basis points.

We have a consistent track record of growing EBITDA, the negative year-over-year EBITDA growth was a result of the loss of Mega, but absent that, very consistent growth. Even through the depth of the recession, we were able to grow topline in a quarter where GDP was declining at 7% annually at an annual growth rate of about 6%. But if you look at that average growth rate though, it's been about 13% over that period.

Now, that has allowed us to build a strong balance, return capital both through buybacks, converted product of sales, expand the market and issue a dividend. So the company has a recurring dividend. It's actually grown that dividend three consecutive quarters on a row. We expect to continue to evaluate the dividend, grow the dividend, produce increasing amount of free cash and give investors hopefully by the end of year a firm commitment on a payout ratio of what percentage of that free cash will then return. And then in addition to that we are working to add our balance sheet to see whether we should put additional leverage on the balance sheet to give us some additional flexibility to return cash to shareholders through either dividends or buybacks.

So in summary we have built the worlds leading internet service provider. We have the most efficient network in the market producing over 20 times as many bit miles for employee as our next closest competitor. We focus on the highest traffic locations. We have a management team that's very stable. It's been intact and in place for about a dozen years.

We have tremendous operating leverage. So we're delivering EBITDA margins that are industry-leading. In fact, they are over 1,000 basis points better than our next closest competitor. They are expanding and we are doing that lower capital intensity. That has allowed us to continue to grow the amount of capital that we're retuning to shareholders. And we expect those trends to continue going forward.

With that, I'd like to maybe open it up, and say we've got a couple of minutes for Q&A.

Question-and-Answer Session

Unidentified Analyst


David Schaeffer

So on the corporate side of the business, we don't discount that way. Prices are the same, better product. In the service provider space there were 200 players a dozen years ago, the market price was $300 and our company market price was $10. We actually priced it a 97% discount to market.

We were much smaller than our competitors and we grew from 0% market share to today carrying 20% of the world's traffic and generating 12% of the revenue. The competitive landscape has thinned. We believe that our largest single cost is sales, the only way to make your sales costs, which is a cash cost lower is to make your sales force more efficient and price just the way you do that.

So our cost of revenue acquisition is about $2 to acquire $1 of monthly recurring revenue with about a six-year customer expected life. The industry typically spends $13 to get $1 of revenue with a two year customer life, so a great payback versus a bad payback. Finally, we're only utilizing 19% of the lit capacity in our network. So everyday we don't sell that underused capacity. It's gone, it disappears. So this is a way to help us fill up the empty seats.

Unidentified Analyst

Maybe if you could just repeat about, why you're low cost producer? Is that essentially, because you've bought other peoples assets that was on the dollar or is there something more than that? I think some of your competitors also bought that.

David Schaeffer

Right, our only advantage was either going through bankruptcy or buying bankrupt assets. When those assets depreciate, we'd have no advantage. Our advantage is much more fundamental than that. So we made two bets in 1999, when we founded the company that we're right and one bet that was very wrong.

The first bet was the internet was going to be the only network that mattered. Second was the architecture redeployed of a corporate land on a global footprint over single pairs of fibers purchased through IRUs would be far more efficient. That we got wrong was the size of that addressable market.

We like everyone else thought it was bigger and growing faster in dollar terms. So were lucky and that we didn't spend all of that capital, and we're able to go and buy those assets, but in addition to the asset base that we purchased with a clean balance sheet. We have a lower incremental cost. So if you look at our network, we have a pair of fibers as opposed to maintaining a large bundle of underutilized fiber.

Secondly, we have a network that every time we add more capacity, it's cheaper for us to do it than our competitors based on IP-over-DWDM protected at Layer 3 using Ethernet. Third, we have a simple product set, we have HQs. The typical telecom company will have between six at 10,000, which has a huge number of implications for back office cost.

In addition to that we have this very efficient sales model, and we focus the sales people at the highest traffic locations. So one way to measure long-term efficiency is to look at the number of bit-miles produced per employee. Cogent is over 20 times more efficient than its next closest competitor. A second way to look at efficiency is EBITDA. EBITDA ignores balance sheet, so it kind of forgets depreciation and capital structure.

We have EBITDA margins that our 1,000 basis points better than our competitors, yet our competitors are at minimum 20 times our size. So there is clearly something beyond cheap assets. And it is the focus on product, network architecture, geography, sales models that are sustainable advantages and what ultimately becomes the biggest barrier to entry scale.

Telecom is an industry where marginal cost is always below average costs. This is a business that doesn't have the concept of a backward-bending supply curve. It is our ability to become the largest that becomes a sustainable costs advantage, independent of what our competitors do. I see that you have right point you got on there. There is somebody else standing here, so if you have any other questions?

Unidentified Analyst

Do you have any thoughts about whether at some point where you should add more values-added products or say that this comes through?

David Schaeffer

We have looked at that. We sell to a number of value-added providers. I believe the new trend is continued separation of the application layer from the network layer. We have no real core confidence in those applications. So for the future we're going to remain just a bandwidth provider and allow others to sell the applications on our network.

I want to thank you everyone. Take care.

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!