Great, everyone. Thanks for coming here to the afternoon session. Again, my name is Anthony Klarman. I cover the telecom cable and satellite space for DB. I wanna welcome everyone to the session for Cogent Communications. Here with us from Cogent is the company CEO, Dave Schaeffer. Dave, thanks very much for being with us in person here again.
Hey, Anthony. Thank you for hosting me. As always, I thank DB for such a great venue, and always thank investors for taking the time to hear a little bit about our story.
Dave, I was just giving you a hard time as we were walking up on stage today. I had prepared my Q&A draft before you announced the most notable transaction the company has engaged in in some time. I thought maybe a good place to start was, a lot of us are familiar with what used to be called the Global Markets Group, the Sprint Global Markets Group. It's been around a long time. It's been up for sale for a long time. What made now the right time to engage in a transaction like that with T-Mobile? What made the structure of the deal so unique and compelling that gives you the confidence to do something like that?
Yeah. Two different parts to the answer. What motivated us as a buyer, what motivated them as a seller. You know, Cogent's history was that we set out to build an all IP over DWDM network protected at layer three, using Ethernet to focus on two types of customers. Large multi-tenant office buildings, where we sell to mid-size businesses, and data centers, where we sell to other service providers. Along the way, the telecom market imploded, and we were lucky enough to have raised $500 million of seed capital. We took that capital and went out on a shopping spree, looking at a number of distressed assets. We ended up bidding on 19, acquiring 13 companies that were then dismantled and reconfigured into Cogent.
Cogent is not a roll-up, but rather the sum of the assets that those companies like PSINet, Allied Riser, Carrier1, Firstmark Communications, had deployed, Verio, and then recompiled into our architecture. For the next 16 and a half years, we looked at over 800 potential M&A deals, and this is the first one that we have done. In doing this acquisition, we acquire two things that Cogent doesn't have today. A large enterprise customer base that are buying services similar to what we sell, and a 19,000-mile inner city owned network and 1,300 route miles of metropolitan fiber and 1.3 million sq ft of technical space, that is fee simple owned. What motivated us at this time was, I think, the need for those additional revenues and customers, coupled with a realistic seller on the part of T-Mobile.
You know, just as background for investors, this is actually the third time that Cogent's worked on this potential transaction. When SoftBank announced its acquisition of Sprint, prior to that acquisition, the CFO of Sprint, Joe Euteneuer, reached out to me. I actually had the opportunity to meet with Joe and Michael Schwartz, who was running strategy. I flew out to San Mateo, met with Masa Son, and then ultimately went to Tokyo, had another meeting with Masa and the rest of his wireline management team, and we had the structure of a deal in place. Masa wanted us to consummate that deal at that point, and we had not done any due diligence or gotten board approval. We did not. We continued to do the diligence, and then quite honestly, he lost focus and was no longer interested in transacting.
Fast-forward to 2019, end of 2018, beginning of 2019, TAP Advisors, a small boutique bank, ran a second process for T-Mobile. We took a look at it in that process and concluded it was unactionable. The only proposal we made was a network management agreement with a profit-sharing split with T-Mobile. Now, we and everyone else who looked in that transaction ultimately did not consummate a deal. Now let's fast-forward to July of this year. I get a teaser from a different bank, Houlihan Lokey. They were looking to sell it. I ignored it. I said, "I've wasted time on this twice before. I don't need to waste more time." I actually knew the senior banker at Houlihan. We had done a deal when we acquired Allied Riser over 20 years ago. He was representing Allied Riser.
He reached out to me and said, "You really do need to do some work on this because economics are not motivating T-Mobile. They are motivated by the maintenance of their brand, treating customers fairly, treating employees fairly, not having any regulatory issues against commitments they had made." They wanted to look good in a transaction. On that list, nowhere was economics.
Mm.
He said bluntly, "We're willing to be a payer for you to acquire this asset. It's deteriorating." I said, "How bad?" He said, "Really bad.
Need interiors.
We went in and did the work, came up with the structure that is proposed, and worked with T-Mobile to achieve their objectives and ours. I think the $700 million payment they're making to us is adequate for us to be able to restructure the business and be assured of profitability. Remember, in addition to that number, between signing and closing, T-Mobile will assume the burn and have other operational improvements in the business, probably spending $500 million in that interim period. Then post-closing, they're providing us two additional backstops, $25 million for any additional employee severance.
Mm-hmm.
Above and beyond the $700 million. Then finally, a $100 million blanket indemnification. This is an asset that traces its roots back over 100 years.
Mm-hmm.
Through a number of roll-ups. I think both SoftBank and T-Mobile never fully understood all of the potentials, liabilities, and issues, and we needed to know that we had that backstop. From Cogent's perspective, the price was right, the assets were seminal to our business model, and we felt we could create value very quickly.
In a half kidding way, I have to ask, are you actually going to wire a dollar at the closing table when the transaction closes?
I actually think that's going to be required for it to be a legally binding transaction.
Yeah.
Under Delaware law. I think that $1 nominal payment will be made, and I'll have to wait till the end of the month till I get my $29 million first installment from T-Mobile.
Right.
I'll front them a buck.
Perfect. Let's talk a little bit about the structure and how that helps you finance what the market believes could be a tricky integration and kind of getting the network sort of Cogent ready. Half of the $700 is kinda front-loaded into the early first 12 months, and then the other $350, I think, is spread out fairly ratably over the network agreement that they signed with you. How should we think about the pacing of your integration spend? Do those payment streams that T-Mobile has committed to get some of the way, all of the way towards the funding of the integration expense that you'll need to do?
They actually get more than 100% of the way there. You know, my ask in the negotiations was that we had to be confident that we were better off at the end of each month than we would've been if we stayed independent and had not done the acquisition. The step has a lot of complex pieces to it. The first piece was we had to actually demerge under Delaware law, the wireline GMG assets from the wireless business. We left behind two business units that we had no interest in owning, that T-Mobile had made regulatory commitments on. That being secured federal operations, a very small, somewhat secretive operation with an independent board of flag officers from the military.
Mm-hmm.
The transcription relay service business that it wanted to continue to support for regulatory reasons. We took what was the GMG business. After that demerger step one is we take the entire lit T-Mo customer base and port that over to Cogent's network.
Mm-hmm.
Now, we can't do that till the deal closes. We can do some of the pre-planning, some of the integration, but much of that transition has to wait on the 50 states and 24 countries to approve the transaction. Once we have migrated that traffic, and to give investors a sense of scale, Cogent today carries about an exabyte of traffic. The T-Mobile wireline network or Sprint network was carrying about 10 PB or about 1%. We move that over, and it looks just like another customer to Cogent. At that point, we free up a pair of fibers that are lit on the Sprint network. We augment those fibers to sell wavelength services to customers.
Step three is we light a second pair of fibers on that network and then swing the Cogent network back onto the T-Mobile network or the now wireline network that Cogent owns on a segment-by-segment basis. That allows us to exit our Lumen agreement and save about $15 million in operating expenses. The fourth and final stage of this would be the monetization of the dark fiber that is not being used. You know, a simple way for an investor to kind of understand this transaction would be Cogent is acquiring two things, a customer base of services that it wants to support and an underutilized network, like a value add office building that's being repurposed.
It was a network that was originally built to carry voice. As it turns out, the fiber that was used, SMF-28, is actually superior to the non-dispersion-shifted fiber that was installed a decade later. It is armored and deeply buried, so it actually has very few cuts or splices. It actually has better characteristics in terms of loss and transmission than the rented Lumen network that we're on today.
I wanna understand those pieces to make sure that it comes out clear as we talk about it. It sounds like you're actually only gonna temporarily use the Sprint network, that ultimately it migrates all back to the Cogent network, and you'll sell off both lit and dark pieces of the asset that you're acquiring.
Not exactly, Anthony. We actually will utilize the Sprint network from day one to sell wavelength services.
Right. Okay.
On the pair of fibers that were
That's on the lit 2 fiber pair.
That is already lit and just needs to be augmented with additional transponders. Step three, which is the lighting of another pair of fibers, and then the repointing of the enlarged Cogent network onto those fibers. At the end of this process, Cogent Communications Group, its current customer base, and its newly acquired customer will be riding on a newly lit pair of fibers on the Sprint network. The legacy fibers that were lit will be used for the wavelength business, and then all of the incremental fiber is available for dark fiber sale.
Got it. That also is what allows you to exit your third-party network deals that you have in place.
That is correct. We can exit approximately 14,000 miles of IRU that we have and port that onto the 19,000 miles with some unique routing on the Sprint network. The other thing we're doing in this transition is for the customers. The non-core products are being jettisoned. Today, T-Mobile supports 28 products. 24 of those 28 products are being end of life because they are gross margin negative. We think about 200 of the 1,400 customers will go away because they're predominantly purchasing these negative margin products. The 1,200 remaining customers are very large enterprises that buy two major product categories, dedicated internet access, that's a product that Cogent sells, about $220 million of the $450 that's remaining.
We intend to preserve that customer base by offering them 10 times the bandwidth for the same price point. The second major product group is MPLS. In the MPLS transition, it's actually a two-part value proposition for the customer. Phase one is the conversion of their MPLS protocol to Virtual Private LAN Service or VPLS. We have developed a process and technology to do that completely transparent to the customer. The customer gets a platform that alleviates your need for CPE equipment and will support their needs for the next 20 years. Almost all of these customers had been through a previous migration with Sprint moving off a Frame Relay.
Mm-hmm.
To MPLS in the late nineties. The average customer relationship with Sprint's over 30 years. This is a very durable customer base. The second thing we're gonna do for those VPN customers that will be riding VPLS is give them the same 10x bandwidth improvement and maintain pricing parity. For those customers, they see a more modern network and 10x the throughput, but no disruption to service. It is truly treating the customers well and minimizing any potential harm that this multi-step transition will cause.
To the untrained finance ear, this all sounds a little complicated, or that it at least has multiple steps. What can you tell the market that gives you the confidence of how you thought about de-risking these various steps that you have to take to make this deal a success?
Let us start with the Cogent team. Most of my team has been with me since the beginning. The average tenure of our team, my senior reports, is 18 years, and they have been through other acquisitions and integrations. Two, we are selecting key people inside of the Sprint organization who have been through other product and network transition programs. At the physical layer, we've done this, been there, and we feel quite confident this is a linear path. While a lot of steps, something that we've done routinely. To just give an investor a sense of scale, in the current Cogent network of 79,000 miles, in any given point in time, we have 85 open network projects, and we complete about six of those a day.
Now, most of those projects are just adding more capacity to the existing network. This project is physically extending the network to allow the traffic to flow between the two networks. I think on the physical layer, we have a great deal of confidence. On the customer side, we're retaining the 60 reps that are in the Sprint organization that have an average of over 20 years tenure in the organization and have continued to work those accounts to help manage the customers during this process. The third thing we've done to de-risk this is we placed these acquired assets into a sister entity that is outside of our credit group. That allows us to isolate any negative effects from our current bondholders. The $700 million payment from T-Mobile comes in at the holding company level.
Sure.
It is then discretionary for us to down flow it either into the group, which is our current borrower, or into infrastructure, which is the entity that will be owning the physical network and selling wavelengths. I think both financially and operationally, we've taken a lot of steps to mitigate any potential downside. I don't wanna trivialize it. It is gonna be complicated, it is gonna be hard, but it is not beyond what we have done routinely over the years.
Maybe just to close out the topic, 'cause I do wanna use the time to get back to the core business. Just your thoughts on time to close, and then what the big major approvals are that are required as conditions to closing.
We don't anticipate any regulatory hurdles, but it does require all 50 state public service commissions to weigh in. California is notoriously-
Yeah.
The longest pole in the tent, and we've begun outreach to some political groups in California to help us. We were very instrumental in the California Net Neutrality rules.
Mm-hmm.
I had the opportunity to testify in front of both California State Senate and California State Assembly during that process, and we hope that some of those resources will be helpful. There's also a requirement for about two dozen federal agencies to weigh in. Cogent today is unregulated, and there is virtually no overlap. We think.
Sure.
This is straightforward. It will probably take at minimum six months, as much as two years, with the most likely timeframe somewhere between nine and 15 months.
Great.
Let's talk about Cogent.
Let's talk about Cogent, Dave. So on your 2Q call, you reiterated your view of trying to return to historical type levels of sequential and longer term growth rates, but that it was gonna take a little longer than expected. Obviously, there are things that happened in the pandemic, and then there are the slower return to work that has happened sort of post-pandemic. Can you just sort of refresh those views a bit now that you've also seen some additional activity from the quarter, and give us a sense as to where you think those, the return to growth rates is turning out?
Just to remind investors, Cogent's corporate business is focused on central business districts of major cities selling into skyscrapers. That business had grown for roughly 14 and a half years, sequentially all but two quarters. The only two negative quarters were Q4 of 2008 and Q1 of 2009 in the Great Recession. The average growth rate was 2.5% sequentially, 11% year-over-year. Since the pandemic, we have 10 either flat or negative quarters of growth. During the pandemic, we saw our growth rate go from +2% sequentially to -4%. In Q4, as companies returned after the second lockdown, we saw very encouraging growth, and we returned to basically flat sequential revenue. We thought that momentum was going to continue. It stalled Q1 to Q2. We didn't decline, but we stayed flat again.
We think the recovery is going to take longer and be less even than we had originally expected. We do believe we will ultimately get back to that corporate growth rate of about 10% or 11%. Now, the silver lining in this is Cogent's net-centric business has outperformed its going in growth rate and its historic growth rate. That is 43% of our revenues. It is highly internationalized. It had grown at an average of 9% a year. At the beginning of the pandemic, it was growing at 4%. Post pandemic, it's averaged about a 16% constant currency growth rate. Last quarter was fairly typical at 16.5%. We think that the entire business, which today is growing at about 5% year-over-year, will return to a 10% growth rate.
The margin expansion that we are delivering, which today is about 100 basis points, will revert back to the 200 basis point norm that we were delivering for the previous 14 years. You know, with that, we have kind of guided investors to look at our dividend growth rate. We have 40 sequential consecutive quarters of growing our dividend. That dividend growth rate today is about 12.5%. Our cash flow growth is more like 6%, between 5% and 6%. Eventually, those two numbers need to come into alignment, and either we'll see a re-acceleration in the corporate business or we will slow down the growth rate. It's not a question of cutting the dividend or not growing it.
It's just a question of is it prudent to keep growing it at 12.5% when our corporate business is underperforming. Finally, as I get to travel around the country post-pandemic, there's not a unitary answer. I was in San Francisco a week ago. It was like a ghost town.
Yeah.
Yet, you know, I was in Minneapolis yesterday, and it felt pretty much like business as normal. You know, I've been to Atlanta, I've been to Chicago, I've been to New York, I've been to Boston, where our offices are, and it's very lumpy. In the South, business is booming.
Mm.
In, you know, Seattle and in San Francisco are probably the worst markets. For places like L.A. or Chicago are somewhere in between.
I wanna talk just a bit more about the corporate side and then get your views 'cause you have some really interesting forward-looking views. How much of that returning to the prior growth rates on the corporate side is contingent upon that vacancy rate that you've talked about, which got to 18%? Does it have to get all the way back to six, which I think was more sort of in line with historical trends? Can you get there on that growth rate by finding someplace in the middle?
The answer is yes. We do not need to get back to historical growth rates, although it will make our job easier and will probably push us above the long-term average growth rate. We today have about 20% market share. That means there is 60% of the building that is still ripe for sale, 18% vacant, and then the 20% that we have, there's 60% that are not Cogent customers.
Mm-hmm.
What we have seen is a stabilization in office occupancy. It had been declining through mid-year last year. It's plateaued, and it's actually slightly improving. It appears new leases are for less term and about 20% less square footage. That will ultimately be a positive by increasing the number of unique businesses in the building. Our biggest impediment to returning to 2.5% corporate growth is not a lack of addressable market in our footprint, but rather the ability of customers to get comfortable that this time is different and they're ready to make an architectural change.
Mm.
The IT departments have been whipsawed at least twice in both Delta and Omicron, hurry up and open and then shut down. I think a lot of IT managers are using this post-Labor Day as a litmus test. We do see spoke to's improving, we see proposals improving, and we are seeing improved corporate growth. It's just isn't improving as fast as we hope the answer is now.
Let's talk about NetCentric for a second because that's been your tailwind. Corporate's been your headwind. You always talk about you guys. You're very open about you win on price. You're the price disruptor in the market, and your marginal cost to carry an incremental bit is zero.
Yeah.
Basically, 'cause the network's already there. If I look at who the other big four international transport carriers are, they could probably make a similar argument. These are big multinational telcos that have some form of global transport. What is unique about your ability to continue to win share? How much of the tailwind in NetCentric is just a pure macro driver for everyone, that the desire for more bandwidth and more capacity, given the fact that you can price it almost at zero for the incremental bid, kinda grows the pie, and it kinda grows it for everyone.
I agree with part of your question, but not all of it. Cogent's short-term advantage was the way it acquired its assets, but those assets have been depreciated. Our sustainable advantage comes from our network architecture. There are two variable technologies that we deploy, wavelength-division multiplexing and optically interfaced routing. Our architecture allows us to capture those improvements in technology more effectively than any of our legacy competitors due to the fact that they support a heterogeneous product set. Sprint was in the exact same situation. There was a period for eight years where Sprint was the largest internet backbone in the world.
Mm.
Today, they're a fraction of 1% of Cogent's traffic, and they just could not compete. We continue to gain share based on our focus on this market segment and our cost structure. The other players have elected to sell higher value services, focus on complex enterprise solutions, and that makes them less competitive in transit. The fundamental building block of Cogent, and why we're still here after 23 years, is the Internet eventually replaces every other network. It's cheaper, it's more ubiquitous, and it's actually higher quality. For a long time, the Internet suffered this best efforts, you know, rap sheet that has been proven to be invalid. I think the pandemic proved that. Then on the NetCentric side, it's been our ability to go after the largest content producers, whether it be Netflix, Amazon, Google.
Mm-hmm.
Microsoft, Facebook, all Cogent customers were their primary upstream provider. On the other side, we have 7,700 access networks from cable companies like a Spectrum or a Vidéotron in Canada to a telco like Bell Canada or BT, to Telecom Vietnam, to Telekom Malaysia. They rely on us for upstream. Because we're more data centers connected to more networks, we can offer lower pricing, and we actually get paid on a much greater percentage of traffic on a two-sided payment. The Internet works by sending a bit and either peering that bit away, meaning you get paid by the sender.
Mm-hmm.
The peer is settlement free, or you send it to a customer. Our percentage of traffic that we're paid on two sides has gone up from less than 20% to 73%. That gives us a sustainable competitive advantage over those four other providers who don't have the breadth of data center connectivity and the breadth of access networks that we have.
That's a great clarification. Thank you. I do wanna use the last couple minutes to talk about the balance sheet and capital allocation policies. You touched on it a little bit with-
Mm-hmm.
The dividend, but this is a leveraged finance conference, so I feel like I need to start with net leverage. It's at or near recent highs, and I guess as you think about a balance sheet with your current leverage, yet the growth rates have come down off of normal historical norms. I think you've recently talked about being comfortable with leverage where it is, but I guess we'll get to the dividend in a second. What would it take for you to say, maybe reallocate some of the free cash flow towards deleveraging as opposed to reinvesting it in the growth or distributing it through dividends?
The easiest way to delever is through growth. Again, we are organically growing in an industry that is shrinking, and we see no end to that growth in sight. Secondly, our guidance range on leverage is between 2.5x-3.5x on a net basis. We took on leverage completely voluntarily to accelerate returns of capital to our shareholders. We have breached that 3.5-fold metric, and we're at 3.7. We think that will come down. We think the Sprint transaction is one more tool in the toolbox to help us bring that down. We think our corporate business will return to its normal growth rate. If that happens, we can continue to grow our dividend as is.
If that doesn't happen, it would be prudent for us to slow the rate of dividend goes down, still maintaining growth and dividend, but then naturally delevering by cash flow growing faster than the dividend, because internally we have all the capital we need to run the business.
Sure. With the last minute or two that we have, the recent refi transaction that you completed got rid of your most restrictive covenant that you had in the structure, and now you have much better flexibility on moving cash between the opco and the holdco. Presumably, that would support, you know, a buyback program or something like that. I guess, how should this room think about how you might utilize some of that new flexibility that you have, and how do we think about buybacks as a portion of that?
The first thing is bondholders should be cognizant of the fact that we've had leverage for 11 years now, and we've never done anything rash. We've always been very measured in our return of capital, and we have used both buybacks and dividends. We will probably continue to use both as appropriate. Buybacks make sense when markets are very volatile. The 1% surtax makes it a little less desirable. The fact that our dividend can be treated as primarily return of capital has tax advantage to dividend. 79% of that last year was return of capital, and only 21% was taxable. We'll continue to monitor that situation and figure out which is the appropriate mechanism. I think there's three parts to the answer. The first one is we are gonna return more and more capital. That's a certainty.
Two, we will use the appropriate tool at that time. Three, we will monitor our cost of capital. As interest rates go up, leverage is more expensive, and we're looking to bring that down. I think bondholders should feel comfortable at the amount of equity underneath of them, the recurring nature of the business, the increase in cash flow, and maybe most importantly, the moderate pace of return of capital. I'm not gonna go out and do a huge dividend recap or a huge special or a huge buyback. It'll be more gradual.
Great. Well, we are out of time, and that is a great place for us to leave it. Dave, I wanna thank you very much for being with us.
Hey, thank you, Anthony.