All right. Thank you everyone for joining us today. I'm Sami Badri with Credit Suisse Equity Research, and we have Dave Schaeffer, the CEO and founder of Cogent Communications. Dave, thank you for joining us.
Hey, thanks for having me.
Yeah.
Sami, thank the investors for their time and Credit Suisse for a great venue.
Absolutely. Dave, I want to kick things off with you and discuss, you know, a little bit about Cogent, but more specifically how Cogent is actually changing, right? 'Cause you guys have announced a fairly large acquisition. If maybe we could kinda discuss that, the merits of the transaction, the direction you're going with the company.
Yeah, sure. Let's start with kind of classic Cogent and the core business. In our corporate business, we have seen improvements, albeit slow and less even geographically than we had hoped. We are seeing increased corporate activity, increased office leasing activity, increased employees returning to office. And as a result, our corporate business, which has averaged 11% year-over-year growth for 17 years organically, had declined to a negative 8% year-over-year growth rate at the trough of the pandemic. We're now back to a slightly positive, about 0.5% growth rate, and we're seeing continued improvement. In our NetCentric business, where we're selling to other service providers, that business has actually accelerated throughout the pandemic and continues to outperform.
Our growth rate has averaged 9% year-over-year. Going into the pandemic, we were only growing at 3%. That business accelerated all the way to 25.5% year-over-year growth. Now, last quarter grew 16.5% on a year-over-year basis. To just remind investors, 53% or excuse me, 57% of our revenues come from corporate users, 43% come from NetCentric users. Our business is transforming with the acquisition of Sprint GMG. We are acquiring that business from T-Mobile. It was the first nationwide fiber optic network. At time of deal signing, there was approximately $560 million of revenue, 28 products were supported and about 1,400 customers.
As part of the transaction, T-Mobile is end-of-lifing 24 of those 28 products as they are subscale and gross margin negative. We anticipate a closing having about $450 million in revenues, 1,200 large enterprise customers. What Cogent is acquiring is a 19,000 route mile fiber optic network in North America, inner city, and another 1,300 miles of metropolitan network. That network is connected to 1,300 pieces of real estate, 47 of which are significant and will be converted to data centers, totaling about 1.3 million sq ft of technical space and about 150 MW of power. In order to affect this transaction, T-Mobile is paying Cogent $700 million in cash to acquire the asset.
The company today is burning cash. The current cash burn rate is about $300 million. Between signing and closing, the restructurings, including unprofitable product elimination, will reduce that burn rate to about a negative 180 of EBITDA. We will quickly migrate traffic onto the Cogent network in Metro. Over a two-year period, we can save $180 million in local access costs. We will eliminate a leased international network, resulting in an additional $25 million in savings. For Cogent, we will terminate an IRU that was prepaid, but we have maintenance obligations on it with Lumen saving an additional $15 million.
The real, I think, advantage of this transaction is twofold. One, allowing Cogent to serve a new customer base, very large enterprises. That's really a market where Cogent has not participated historically. Those customers have a average relationship life with Sprint and T-Mobile of over 30 years. Secondly, and maybe even much more importantly, is repurposing the asset. The network was originally built to carry long distance voice. It has been under-managed and under-invested in, and we will be able to convert that to, one, carry our internet traffic, and two, to sell optical transport networking services or wavelengths.
That is a brand-new market for Cogent. We anticipate taking the $8 million run rate in Sprint, and growing that to about $500 million over a five to seven year period. That is a $2 billion addressable market. We will have some distinct competitive advantages. One, we will have over 800 carrier-neutral data centers in North America where we can offer those services. Two, over 90% of the route pathing is geographically unique. Therefore provides diversity. Maybe most importantly, we have a 225 person sales force that focuses on this market segment and can use existing relationships with hyperscalers and regional access networks to sell these services.
The combination of the $700 million payment from T-Mobile over a 54-month period and this incremental high margin opportunity really is transformational. Cogent, at closing, will be about $650 million in revenue. We will grow to $1.1 billion with the acquired Sprint revenues, and within five years be at $1.5 billion. EBITDA margins will decline from the 39.4% we delivered last quarter to probably somewhere around low to mid-thirties within five years. The growth rate of the combined company will moderate to 5%-7%, and we should be able to deliver 100 basis points a year of margin expansion. Larger scale, more cash flow, cash flow accretive from day one, product diversification, and real asset ownership are all key attributes.
Got it. Got it. Thank you for running through that. I want to double-click through the corporate customer segment. It returned to growth in 3Q of 2022. One thing is you know, at least all of the investors that are that are focused on Cogent and specifically this segment, are trying to understand what KPIs are you looking to track and how are those KPIs changing that are signaling the direction of that business?
Yeah. Cogent today has about 15% penetration in its corporate footprint. We did see vacancy rates in that footprint during the pandemic increase from 6%- 18%. They have slightly improved to 17.7%, and we're continuing to see leasing activity. We're also monitoring employee entry into the building as measured by security badge swipes. I think what's most important for Cogent's revenue is the level of customer engagement, the number of proposals that are issued, the number of proposals that are converting into orders. You know, what we are seeing is many IT departments finally moving forward with architectural re-engineering that they've been putting off for two years as they didn't really understand what the new normal would look like.
This means increasing internet connectivity, adding diverse points for remote employee aggregation, replacing MPLS networks with VPLS or SD-WAN services. It is this higher level of customer proposals and new orders that's most encouraging. Another KPI that's an important indicator is our ability to grow our sales force and reduce turnover rates in that sales force, coupled with higher productivity. Last quarter, we actually had the highest level of sales force growth in the company's history in a single quarter. You know, we continue to have a robust candidate funnel, but most importantly, with our employees returning to office, we've seen employee productivity increase and therefore lower turnover rates.
Got it. Got it. I wanted to go through from an ARPU perspective, you know, what does the price scheme look like from 100 MB to 1 G, 10 G, et cetera? Like, what are the typical customers that buy the 10 G speed?
For our corporate customers, they buy a fixed connection. It is an all-you-can-eat symmetric, non-oversubscribed, non-blocked service. We offer contract terms ranging from as short as month-to-month to as long as five years. The most common contract is a three-year contract for corporate customers, those customers are buying a 1 Gb connection, typically for about $600 a month. The 100 Mb connection, which had been the previously most common product several years ago, has been end-of-lifed, and we're phasing those out with over 90% of the customers buying 1 Gb or larger connections. We are seeing an increased interest level in 10 Gb corporate connections.
Typically, that's sold at 4x the price of a 1 Gb connection, about $2,500 a month.
The customers that will buy that product tend to be financial services companies that are not very price sensitive, but want to buy the very best that is available in the market. As connection sizes have increased, the average corporate utilization of those connections has actually declined. When the majority of our customer base was on a 100 Mb connection, the average corporate customer was using about 18% of that at peak, or about 18 Mb. As the market shifted to a 1 Gb connection, we've seen the average peak utilization fall to about 9%.
While the average customer's utilization went from 18- 90 Mbps , the percentage of the pipe they're buying is declining. That same trend exists in 10 Gb connections, with most corporate customers having very low total utilization.
Got it. Got it. I want to shift gears back to the Sprint's GMG acquisition. It's projected or aimed to close at some point next year. you know, one thing that we've all been really trying to understand is, can you speak to why you believe that Cogent's ability to reverse the asset, decline or at least, sorry, the assets course of revenue growth, why do you believe that it's going to stop declining once it's actually integrated into Cogent's business?
Again, as background, this is a business that in 2002 had 70,000 employees and $40 billion in revenue and was primarily a long distance company. That company was under contract to be sold to MCI for $129 billion. That transaction was blocked by regulators. Sprint's management shifted its focus after that regulatory setback and almost exclusively focused on its wireless business for 20 years. That GMG business, as voice disappeared, had declined from that roughly $40 billion run rate to $560 million. It had been underinvested and under managed. We look at this and see two key assets. One, the 1,200 large enterprise customers that we're acquiring.
We do not believe we can grow the revenue from those customers, but we do believe we can retain them. They have an average relationship with the company of 30.5 years. Most of those customers had gone through multiple technology transitions from X.25 to Frame Relay, Frame Relay to MPLS, and are now being converted to a virtual private line or VPLS service. We will increase their port capacity by 10x. We will migrate those customers to on-net where practical. Today, 93% of the revenue in that business is off-net. In Cogent, 75% of our revenues are on-net, only 25% off-net.
By bringing those customers on-net, we will increase the reliability and the throughput without increasing the cost. We do not believe we can grow that customer base materially, but we can maintain it. The real growth opportunity in buying this business is the value-add repurposing of the network. Liken it to an office building that was built for office purposes, and it's completely vacant, and it's now being converted to residential or hotel use. What we are doing is taking a fiber optic network that was built with SMF-28, which turns out to be optimal fiber for coherent transmission, which is today's standard for 100 Gb and 400 Gb transmission, and repurposing that.
The transition has four discrete phases. Phase 1, we will take all of the traffic off of the Sprint network and put that onto the Cogent network. Today, Cogent carries about an exabyte a day of traffic. Sprint carries about 10 petabytes or about 1%. Step 2, we will take that lit pair of fibers and use that exclusively to sell wavelength services. That is a $2 billion addressable market. We have a sales force that can focus on that market, and we can increase the endpoints where that service is available from 23 endpoints at Sprint today to 800 by tying the metro networks together with the Sprint inner city network.
Third, we will then light a second pair of fibers on the Sprint network to support the Cogent IP business. Then we will sell off the excess fiber that's available. The routes have anywhere from 24–144 strand count, depending on the segment. Then the fourth and final phase will be repurposing this 1.3 million sq ft of technical space, which today is generating no revenue, into useful data center space by integrating it into our network and by selling into it. That will allow us to generate incremental revenue with no incremental cost. Again, it's a asset repurposing. To de-risk this, we're being paid by T-Mobile to effectively transition the business $700 million.
Yep. Actually, about the $700 million, can we just walk through the puts takes and whether or not it's gonna be recognized as revenue, you know, whether it's just gonna be recognized as a cash payment, recognized somewhere else. Can we just kinda go through how we should be thinking about that?
Yeah. The most important thing is the cash coming in the door. It's asymmetrically front-end loaded, so we get $350 million over the first 12 months and $29 million monthly installments, and then over the next 42 months, we get $9 million a month installments, bringing us to the total of $700 million. Above and beyond those payments, we get an additional $25 million from T-Mobile for any employee severance post-closing, and we get a additional $100 million indemnification if there are any unforeseen liabilities. When T-Mobile acquired Sprint, it focused its due diligence on the wireless business, and for that reason, we could not count on their due diligence when we're acquiring the wireline business.
In terms of how we account for that payment of cash, our goal is to account for it as revenue, we cannot be certain of that. The way the deal is structured is we are providing transit ports to T-Mobile. Those ports will be available day one for them to use. They have indicated, at this time, they have no interest in using that capacity, they will evaluate that every quarter. In order to pass the FASB revenue recognition test, there are really two parts to that test. One, was the service really provided? The answer is yes. We have provisioned the ports.
Whether they use them or not is really their decision, much like if you buy cable TV service, you pay your cable bill whether you turn your TV on or not over the course of the month. The second is, was the service sold at a market price? We have given them a most favorite customer pricing clause. We will have a third-party study that evaluates that. We tried to get a firm commitment from our auditors, Ernst & Young, we could count it as revenue. They were unwilling to make that commitment at this time. For that reason, we structured the transaction as a standalone sister subsidiary to our operating company.
It is our belief that we will be able to count it as revenue, but we don't have that guarantee until we get a signed audit opinion, and that probably won't happen till early 2024. What we are certain about is the cash payments. What we are not certain about is does it count as revenue, and therefore EBITDA.
Got it. Following the cash is obviously the key to that $700 million. Then, I wanted to shift gears over to NetCentric, right? We've seen two years of robust growth and robust demand from consumers, and streaming companies have been some of the key kinda drivers of that strength. As we start to see streaming companies provide less strength or less robust forecasts in their business or even reported numbers, how do you think that's gonna manifest itself into Cogent's results and forecast?
We saw a rapid transition in the movement from linear television to streaming. That had been occurring pre-pandemic from basically 0% of video being streamed in 2013 to, by 2019, 18% of all video in the developed world was streamed. Today, that's over 44%. We saw a material acceleration in that transition during the pandemic. That 44% is continuing to increase but at a more normalized rate. We are also seeing streaming adoption much more robust internationally than domestically. We are also seeing a much larger number of entrants in the market. That actually has bode well for Cogent because that balkanization of the demand allows us to charge a higher price per bit.
We also get higher pricing internationally than we do domestically. Finally, we have seen the percentage of traffic that we're getting paid, both by the sender and receiver, increase from 50%- 73%. All of those factors accelerated the growth in NetCentric from 3% in 2019 to a peak of 25.5% year-over-year at the height of the pandemic. Even as the pandemic has waned, we are continuing to see outsized performance in that business, growing at 16.5% last quarter, effectively similar growth the quarter before that. We believe that we will eventually revert to kind of a 9% long-term average growth rate, but that reversion appears to be slower than we had originally expected.
All in all, we feel even though the streaming providers are struggling to be profitable because of their outsized expenses on content production, there's still plenty of demand, plenty of eyeballs globally, and plenty of minutes that can move from linear to streaming.
Got it. Thanks for going through that. One thing I wanted to hit is the slowing growth rate of your dividend that you introduced and reported in 3Q of 2022. Could we go through the puts, takes, and tell us what— I guess the reason why I'm asking this question is because your dividend growth has been consistent for so long, you've kind of gotten us calibrated on that expectation, but it kinda slowed down and I want to go through that with you.
Over the long run, the growth rate in the dividend has to mirror the growth rate in free cash flow. Cogent initially started with a $0.01 sequential growth rate. It then accelerated to $0.02 and then $0.025. Our EBITDA growth over the past 41 quarters since we've been issuing a dividend and growing it has been about 14%, and our dividend growth rate has been about 13%, in alignment. With the pandemic, Cogent's top line growth rate decelerated from 10% to 5%. Our EBITDA growth decelerated from 15% to about 6%.
As a result, and the fact that the impact of the pandemic has been more prolonged than we expected, we felt it was prudent to lower the growth rate in the dividend to $0.01 a quarter or basically 5% year-over-year, bringing it back in alignment with the growth in free cash flow. As that growth rate begins to re-accelerate, as top line growth accelerates and margins expand, we will reevaluate the growth rate and the dividend. Our leverage is slightly ahead of our targeted range at 3.8x versus the 2.5–3.5 x range that we've outlined. By slowing the rate in dividend growth, we will bring that leverage ratio back down within our targets. It all seemed prudent.
Cogent is truly unique among public companies in having a over 10-year history, 41 quarters of sequential dividend growth. I think there's only been six companies in history, public companies, have had that rate of dividend growth pacing.
Got it. Got it. I wanted to kind of wrap up the fireside chat, anything on your mind that you think we should express that we didn't cover?
You know, I think the one topic I would touch on is a leading indicator you touched on around KPIs. That is our ability to grow our sales force, improve the retention rate of those salespeople, and improve their productivity. You know, that is a key area of growth for the company. You know, I often get asked by investors, what do I worry about? It's sales. You know, Cogent is about 1,050 employees. 750 are in sales, 300 in operations. Sprint was 1,320 employees, probably 1,000 at closing, of which only 60 are in sales. Our effort is going to be to reduce the operational headcount and increase the sales headcount and increase the productivity of those salespeople.
You know, we feel very comfortable that there are plenty of candidates. We have the right addressable market, the right value proposition for customers, and now that we're back in the office, the ability to make those people productive quickly. We're feeling pretty optimistic about where we're at.
Got it. Well, Dave, thank you very much for spending time with us and attending our conference. Thanks to everyone else in the room for attending.
Hey, thanks for having me, Sami. Thank you all very much.