All right. We are gonna go ahead and get started with our next session. We're very pleased, as always to have Cogent Communications CEO, Dave Schaeffer. Dave, welcome to the conference.
Thank you, Matt, for hosting me. Thanks Deutsche Bank for a great venue here in Las Vegas, and thank all the investors for their time and interest in the company.
Well, maybe just to start, we got a lot of topics I wanna dig into. Maybe just to start, can you talk about your top strategic goals, priorities as we close out the year and start thinking about 2023?
Yeah. Cogent has been able to grow throughout the pandemic, where most wireline telecom companies have not been able to achieve organic growth. What we're looking for is the ability to re-accelerate our growth. Our number one priority is to continue the improvement in our sales force. That is, to both add additional sales reps to be able to sell our services and continue to see their productivity increase, as it has over the last several quarters. You know, our NetCentric segment, which is 43% of our revenues, has actually fared very well with the acceleration of the transition from linear video to streaming and that phenomena continues around the world.
It's been our corporate segment, representing you know a little over 55 -57% of our revenues, that has struggled in that many of the customers that we have are located in the central business districts of major North American cities and many of those businesses immediately went to remote work during the pandemic and are only now in the process of reoccupying their offices. We've seen a stabilization in that business. We've seen some improvement, but what we really expect for the remainder of the year and into next year is a significant acceleration in our corporate growth rate.
Maybe that's a good segue into the next question. I was gonna ask you about the macro backdrop. Obviously, it's been really top of mind. Many investors, it's a big topic that's come up at the conference. What are you seeing on the macro front? More specifically around whether it's having any potential impact. Maybe are you seeing any sort of deterioration worsening? What's sort of the macro impact on your business? Maybe we can talk through it across both the corporate and the NetCentric businesses.
Yeah, sure, Matt. We are fortunate in that we sell a necessary utility. Typically we're the last bill that someone would consider not paying because the internet is a business' lifeline to connect to the rest of the world. Our corporate customers are much more creditworthy than the typical corporate business, in that they have elected to put their offices in what is usually the most expensive real estate in a given market. While we have seen a limited number of closures at the beginning of the pandemic, we have seen no deterioration in the credit quality or payment history of our customer base. Our DSOs remain at historically low levels, roughly about 22 days. Compare that to AT&T or Verizon, which are closer to 70 days. Our bad debt as a percentage of revenues is also at historic lows.
Our corporate customers are faring pretty well in the economic turmoil. What we have seen is an increase in vacancy rates in the buildings we serve. The roughly 1 billion square ft of office space that Cogent offers its on net services has seen its occupancy rate go from 94% to 82%. Now what we are seeing is a stabilization in that increase in vacancy rate. We're seeing new leases sign and most of those new leases tend to be slightly smaller and for slightly shorter duration than the leases that they replace. I think this bodes well for Cogent's total addressable market. When those businesses do make a decision to relocate, that becomes a unique window for us to win that business. On our NetCentric side, you know, the majority of our growth has been international.
We have been impacted by the extremely strong dollar on a reported basis, but on a unit volume growth, we've seen growth substantially above industry averages. We've seen our nominal rate of price decline remain consistent, but our effective rate of price decline has actually improved. We also know that our NetCentric customers are also gonna pay their bills. We sell to about 5,000 content generating companies, from names as large as Netflix or Amazon or Disney to small web hosters or regional content delivery networks. In addition to those 5,000 content generating businesses, we sell internet access to 7,700 regional and national internet service providers, mobile phone companies, cable companies and telcoms. They are in turn distributing our bandwidth to their end users. You know, there are about 5.3 billion people in the world who regularly access the Internet.
The Cogent customer base allows us to serve about 98% of that addressable market, where just over 5 billion end users can reach the Internet through Cogent.
Let's maybe drill in a little bit on the corporate business for a second. We've seen, I think, revenues decline sequentially, you know, I think for about eight quarters now since the start of the pandemic. But there were a lot of underlying factors, it sounds like that, you know, when these vacancy rates begin to stabilize and actually compress, could actually work in your favor. I'm wondering, I think in the past, maybe you had talked about an expectation that we see revenue trends inflect in the second half of the year, although it seems like that may have been pushed out a little bit. I'm just wondering, how do we sort of think about when we see that inflection right now?
Yeah. We were extremely encouraged by the rate of improvement in our corporate business from Q4 of 2021 to Q1 of 2022. We were anticipating that rate of improvement to continue. In fact, what happened was it stabilized from Q1 to Q2. It did not deteriorate, but did not continue to get better. You know, prior to the pandemic, our corporate business had 56 out of 58 quarters of average sequential growth of about 2.5% quarter. The only two quarters prior to the pandemic where we saw negative growth were during the great financial crisis at the end of 2008, beginning of 2009. During the pandemic, we've actually had ten quarters of negative growth, although that rate of negative growth has improved and is effectively flat at this point. We think we will continue to see gradual improvement.
I think the corporate business will return to positive growth and eventually return to that 2+% sequential growth rate, but I don't think it happens in the next quarter or two. I think the rate of improvement is more mild and I think the main reason for that, in talking to our sales force, is that many corporate customers have now come to realize that a hybrid work environment is part of their network architecture, but they are not all ready to rearchitect, afraid that there will be yet another variant or another shutdown. I think what we're seeing is a portion of the market move on beyond COVID and get on with their network designs and begin that cycle of MPLS replacement and larger internet connectivity.
I think there's another segment of the market, though, that continues to be cautious and just push out decisions. I think that's the reason why the growth rate has not accelerated more rapidly.
Would it be fair to say if we see vacancies, let's call it 18% now, compress and customers more actively spending to rearchitect the network, in theory, those two would be, I would think maybe the bigger drivers of a return to growth?
Absolutely. I would actually say it's more the latter. It's the customer's decision to modernize their network architecture that will be even a bigger driver than the improvement in office occupancy. You know, Cogent's greatest competitor is actually not another service provider, but rather the indifference and the status quo. The final point I'd like to make on that is, this does not require an expenditure on the customer's part. It just means they need to have a vision of what their new architecture needs to look like. Our services are typically sold at price parity to our competitors, to a corporate customer. They buy a fixed connection on a monthly subscription. We deliver that service 9x faster than our competitors on average and once it is installed, it is 3x more reliable.
The final and maybe most significant attribute is the fact that the customer gets 30-60x the effective symmetric throughput when compared to our competitors. For our corporate customer, the definition of value is not purely price, but it's really the superiority of the product. That is very different than our value prop to the NetCentric customer.
It's a good segue, so I'm gonna pivot to NetCentric. You've seen outsized growth. I think constant currency growth was close to 16-15% last quarter. That's I think despite some tougher comps. Can you talk a little bit about what's driving the outsized strength here? Then maybe also discuss what your outlook is for growth over the next several quarters.
All right. Let me start by outlining the average historical performance of that business. It's grown on average about 9% year- over -year, over the past 16.5 years as a public company. Different than the corporate business, it's never been as consistently, sequentially growth business. It's had a fair amount of volatility. The pandemic is a great example of that. We actually entered the pandemic with our NetCentric business only growing at about 3.5% versus the trend line of nine. We then shot up to a peak growth rate of over 25% year -over- year, which is the highest growth rate in the company's NetCentric segment in the company's history. We've now kind of stabilized in the mid teens, roughly 16.5% last quarter.
The primary driver of that is unit volume growth and that unit volume growth is being driven by the conversion of linear television to streaming, both domestically and internationally. Our NetCentric business operates in a market that is typically characterized by mid 20% volume growth in conjunction with 20% price reductions per year. That's really been the pattern of the Internet for the past 25 years. Our ability to grow at 9% is a direct result of our gain in market share. We gain that share for three reasons. One, we are able to offer our services in more locations than any other NetCentric provider. 51 countries, 1,450 carrier-neutral data centers. That's over 500 more than our next closest competitor. Secondly, we have more networks directly connected to the Cogent network than any other network in the world.
The Internet is not a single thing, but rather an amalgamation of about 60,000 networks. We have 7,700 of those networks directly connected to Cogent. That's probably 3,000 more than our next closest competitor. That direct connectivity allows us to provide a higher quality of service. The final point, which is maybe most important to a NetCentric customer, is price. It is a metered service that is a fungible, undifferentiated commodity and we win because our price points are 50% of our competitors. Now, our growth rate has come by taking market share. It's also been accelerated by the fact that more of our growth has been outside of the U.S. than in, because it's been those international markets, particularly in developing countries, that are growing more rapidly.
The final point is the fact that a higher percentage of our revenue is now derived from customers who are both senders and receivers. Let me explain that. When some individual or company connects to the Internet, they send out request bits, gather information and get return bits. Typically, the way that traffic exchange works is the company that is aggregating those end users, say a Bell Canada, a British Telecom, a Charter, all examples of Cogent customers, would then buy upstream connectivity from one of the global backbones. We get paid by those aggregating networks based on the traffic that they consume. On the other side of that request, there are those content generators, the Hulu and Disneys of the world, the Paramount and Netflixes and Microsoft or you know, any of the gaming or content companies.
There, we're getting paid by those companies to connect to the Internet. We have seen the percentage of traffic that receives a two sided payment go from 15 years ago, less than 20% of our traffic, to today's 73%. That increase in two sided payments help reduce the effective rate of price decline much more substantially than the nominal rate of price decline. As a result, our NetCentric business, on fairly typical volume growth, has actually outperformed on revenue growth.
I want to ask about sales force and productivity. It's one thing you referenced, I think, in the opening remarks. Productivity for sales reps actually improved pretty meaningfully, it looks like, the last several quarters. I think you hit 4.9 in the second quarter. What's driving this?
The number one issue that I deal with is sales force size and efficacy. It was what my objectives were for the remainder of the year. It's where I and most of the management team focus the majority of their time. Cogent's about 1,050 employees, of which 750 are in sales. We have a highly automated, highly consistent network. Even though we carry nearly a quarter of the world's internet traffic, we're the second largest carrier in the world. We do that operationally with only about 300 employees due to the level of automation. Our challenge is filling up that network that today is only utilized at about 28% of lit capacity. That's sales organization. We have an outbound telesales model. We have 80 sales teams located around the world. They physically sit in 45 different offices.
I spent the day yesterday in Phoenix with the team in that office. I'll be spending the day tomorrow in Boston with the team in Boston. I try to get around and hear what the salespeople are actually seeing in the market. It's a really tough job, cold calling. Anybody who's ever done it knows how hard it is. The improvement in productivity has come from getting our reps back in the office, giving them the support infrastructure of training managers and supervisory managers to help mentor them and grow them. Not everyone is gonna succeed at selling over the phone, but we need to make sure that we give those individuals every opportunity to succeed in terms of tools, product and training and support. During the pandemic, we had to pivot to a remote training model.
We continued to hire at record paces throughout the pandemic, but we also saw a record number of turn overs due to the fact that those reps were not hitting their key performance objectives and quota objectives. Now that our sales teams are back in the office, we've seen an increase in productivity. We expect that to continue. We're actually still at 4.9. Clearly, substantially better than a 4.2, which was the trough in the pandemic, but still slightly below the long term average of five orders installed per rep per month. We expect that number to continue to trend up now that we're back in the office, and the fact that our underlying markets are also strengthened.
If we think about margin, so obviously we talked about some of the challenges in corporate. We've seen some slower consolidated top line growth as a result. EBITDA margins north of 39% this past quarter, they expanded about 70 basis points year-over-year. I assume some of that has to do with some of the strength in NetCentric, which tends to be primarily on net. Maybe you can help us think about what's driving the resilience and strength in margins, and then maybe talk a little bit about your near-term outlook as well.
Yeah. Our EBITDA margins last quarter were 39.4%. We have averaged better than 200 basis points a year of organic EBITDA margin expansion for the past 16 years. To do that, however, we need to have 10% top line growth. Roughly 75% of our sales are on net, 25% are off net. When we sell a $1 of on net, it carries a 100% gross margin contribution and $0.95 of EBITDA. When we sell a $1 of off net, because we have to buy a local loop from a third party, it carries only a 50% gross margin, $0.45 of EBITDA contribution. The blend of those two different incremental revenue streams has resulted in a combined contribution margin in the mid 60s.
We anticipate that our EBITDA margins will continue to expand till they plateau at around 50%. We have about another 1000 basis points of margin expansion left in the business. It could expand beyond that if the addressable market was larger, but ultimately, we can't deploy capital to locations that are not giving us a high enough return on capital, so our total addressable market becomes somewhat fixed. With our growth rate today at about half of our historic average, our margin expansion is at about half of our historic average. We expect that growth rate to improve and we expect the rate of margin expansion to improve. We're doing all this while at the same time maintaining modest CapEx, basically flat capital expenditures, which has allowed us to grow free cash flow per share.
We think we have the ability to continue that for at least another decade.
Maybe one sort of derivative of EBITDA, we've seen leverage, net leverage was about 3.37 turns last quarter. It's ahead of your sort of target 2.5-3.5 range. I'm just wondering, you know, is that still sort of a relevant range to consider, particularly with everything going on in terms of rates and maybe some of the more moderate EBITDA growth?
We have been able to grow our dividend for 40 sequential quarters. That's not year-over-year. We're growing it sequentially every quarter for ten years. Our rate of dividend growth is about 12.5% today. We've been sequentially growing that dividend $0.025 a share. Our dividend last quarter was actually paid out yesterday, was $0.905 per share for the quarter. We have excess cash on our balance sheet. We were fortunate that we built our business off of the backs of distressed companies rather than incurred debt on our own. Cogent took advantage of the telecom meltdown in the early 2000s. It bought $14 billion of invested capital and PP&E and $4 billion of PP&E for a total of roughly a hundred.
for $60 million, but we actually acquired $115 million of cash along with those acquisitions. We were effectively paid to take over these distressed assets. That allowed us to build the world's largest IP network with no debt. We went public with no debt on our balance sheet. We began buying stock back a year after being public. We also started a dividend policy and we've added leverage. We today have excess cash on the balance sheet and we're slowly disgorging that. The final point is that if our growth and free cash flow does not re-accelerate, we will have to slow the rate at which we're growing the dividend.
We are quite comfortable, however, that we're still maintaining a prudent net leverage ratio in what is a rising interest rate environment, but one that, by historical standards, is still a low interest rate environment. We are committed to returning greater amounts of capital. We have returned over $1.1 billion to our equity investors. I actually met with an investor earlier today who reminded me how large their negative basis in Cogent stock is. This particular investor bought, not in the IPO, but right after we went public. We went public at $6. I think he bought at about $7.50 a share. He's held those shares, and we've been able to characterize the majority of our dividend as a return of capital. Today, he's received nearly $15 of cash back on a $7.50 investment and has a negative basis.
That's a great outcome if you ask me, and being able to continue to do that is what we are committed to. Now, we have also supplemented those dividends periodically with buybacks. When there's periods of extreme market volatility, that's when we've been a buyer.
One thing I noticed, I think that you'd mentioned this on your last call, you recently eliminated. There was a restrictive debt covenant. Now you've got, I think, it's $287 million in unrestricted cash you can return to shareholders. Anything to read into that in terms of maybe it being a leading indicator for any change in the pacing of shareholder returns?
In order to lower our cost of capital, we have a two tier corporate structure. We borrow money at the operating company level and provide a security. The public equity has the holding company for the dividends and there is no debt at the holding company level. We had previously had covenants that limited the amount of cash we could move from the operating company to holding company. While we still have restrictions, the new debt instruments give us much more flexibility, and we've moved almost $290 million to the holding company level. That means it's fully available for either buybacks or dividends. We will continue to methodically grow our return of cash and whether it's through buybacks or dividends or a combination of both, we are committed to using our balance sheet efficiently.
Maybe last question to sort of tie this all together. Obviously, seeing the stock get hit with a little bit of volatility. I'm wondering, what do you think is most misunderstood by the Street about the Cogent story, and where do you see the potential for upside surprise over the next 12 months?
With a company that has a 16.5 year history of 10% organic growth to be trading with a 6% dividend yield seems somewhat irrational to me when compared to other companies around us. We sell a necessary utility that every business needs. The unit volume of what we are selling is growing and will probably continue to grow for the foreseeable future. We're uniquely competitively advantaged based on our network architecture. I think what investors misunderstand is the exposure in our corporate segment to the office market. If you believe no one will ever go back to an office and 100% of our population will work from home permanently, Cogent's the wrong company to own. You should not own Cogent because we are heavily exposed to that segment.
If you believe that some portion of most white collar workers weeks will be in an office and they need the VPN flexibility to work from home those other days, Cogent is a great business to own. I think investors have over reacted to the downturn in Class A office space and projected that into Cogent. It is true that rents are coming down in major cities. Occupancy rates have declined, but the office is still here. I think the second thing that's misunderstood is on our NetCentric business, which is a commodity business, how sustainable Cogent's market advantage is based on our network architecture. We have seen many other companies exit that market segment.
We've seen some of our customers struggle in that segment, but they do not have the architecture that we built 20 years ago and have continued to build on, which is uniquely advantaged for the Internet. We started with the simple premise that the Internet would be the only network that mattered. The architecture that we deployed allows us to capture the advances in wave division multiplexing and optically interfaced routing much more effectively than any other network. Matt and I, before we stepped up here, were talking about some of his equipment companies that he covers and those companies have been able to achieve miraculous rates of productivity improvement in their products. Optically interfaced routers have a 30 year history of a 40% compounded rate of improvement. Wave division multiplexing has a 35 year history of an 80% price performance improvement.
The problem is for our competitors, they can't monetize that. For Cogent, we are able to monetize that, sell at a lower price and achieve better margins.
Okay, looks like we're just about out of time, so we'll end it right there. Dave, thank you again as always.
Thanks, Matt. Thank you all. Take care.