Welcome to Day 2 of the BofA 2023 Leverage Finance Conference. I'm Ana Goshko. I cover technology and telecom credits, and we're thrilled to have Cogent Communications with us this morning. Dave Schaeffer, the company's Chief Executive Officer, founder, and chairman. So thank you.
Well, Ana, thank you for hosting me. I'd like to thank all the investors who got up early to pay attention to Cogent, and as always, thank Bank of America for a great venue.
Okay, great. So, Dave, got a lot of moving pieces right now in the business. You know, most of those moving pieces are associated with the Sprint Wireline asset acquisition the company recently completed. But a lot to cover, but since this is a credit and debt-focused audience, I did want to kind of clarify some issues on debt structure first, get that out of the way. So when you first announced the agreement to acquire the Sprint Wireline assets, I recall that you said the intent was, from bond indenture perspective, to keep the Sprint assets in an unrestricted subsidiary. And I think the intent at the time was to insulate the credit from any negative EBITDA of the acquired assets.
But I also think the last time you publicly spoke about this might have been at this conference a year ago. So I just wanted to clarify what structure was ultimately put in place with those assets, the business, and the kind of restricted, unrestricted group with regard to the debt.
Absolutely happy to answer that question, Ana. The publicly traded company is Cogent Holdings. Holdings has no debt. Underneath Holdings, there are two operating entities. In Cogent Group, which is where our debt resides, Cogent's operation success, all employees, all customer relationships, all revenue, sit within the Group entity. When we acquired the GMG business from T-Mobile, we bifurcated that business in order to ensure that bondholders would not be disadvantaged. All of the operations of the acquired business reside in Group. That means all of the customer relationships, all of the revenues, all of the employees, and most importantly, for the creditors, the stream of payments, $700 million from T-Mobile over a 54-month period, all sit within Group.
Cogent Group, a sister company that is unrestricted, only owns the physical assets, the 482, uh, pieces of fee simple real estate, the dark fiber, meaning the 19,000 route miles of intercity fiber, and the approximately 1,200 miles of metro fiber, as well as approximately 4,600 miles of IRU fiber, leased fiber, uh, that a large portion of that is governed by an out-of-market, uneconomical lease that we called out specifically. Uh, that liability and the assets with no revenue sit within Cogent Infrastructure. Cogent Group is purchasing an IRU from the infrastructure entity as it migrates traffic off of the Lumen fiber and migrates it onto the Sprint fiber, as well as an IRU to facilitate the sale of wavelength services. We will also be leasing space in those 482 buildings only on an as-needed basis.
Rather than burden Group with any undue liabilities, we kept those liabilities segregated. Hopefully, that clarifies it. So all the customers, all the revenue stream, all the employees are in the operating entity that is the borrower, and in the sister subsidiary, only physical assets and uneconomic liabilities sit.
Okay. That's... That is helpful, and that's clarifying. A little tough if you really want to be pure about trying to figure out what the, kind of, credit metrics are for the debt to kind of parse between the two. Over time, as you are able to get out of these economic leases because they basically just expire, and you're able to reduce costs, is there a plan to basically merge the entities so you don't have this kind of separation of the assets, which I think you believe are worth at least $1 billion, right? From the debt holder group.
So the physical assets were appraised by KPMG at $1.15 billion. Subtracted from that appraisal amount was $147 million, or roughly $150 million, for the uneconomic leases that we were forced to acquire. So the net value in that infrastructure entity are tangible assets that today have no revenue associated with them, only costs of $1 billion. As we monetize those physical assets, and that will occur in 3 primary ways. 1, we will sell or lease dark fiber and generate revenue against that liability and at least bring that asset into compliance, and at least break even. 1 of those customers will be Cogent Group, but there'll be other third-party customers.
The second relationship will be, as we repurpose those buildings for data centers, we will become a tenant for a small portion of that footprint, where we will operate a data center business, but equally important, there will be substantial power and space available to other data center operators to lease from us on a wholesale basis. We have not built that wholesale revenue stream into our models and projections at this point, because we're just too uncertain about how deep that market is. And then third, there is a small amount of metro fiber, most of which is not relevant to Cogent, and that may also be eventually sold off or monetized. So the idea is to get the infrastructure entity at least to a break-even level on its own. As it becomes cash flow generative, we will have two options.
Do we decide to somehow borrow against that entity with a separate credit, or do we elect to merge it? We have not yet made that decision, but today, those assets are actually cash flow negative, and they're being funded with cash from the holding company. So cash that we would move from the restricted payments bucket in the covenant up to holdings that could be used for buybacks or dividends, now have a third use, which is they are down flowed into the infrastructure company to absorb the carrying cost of those assets.
Okay. Helpful. Thank you. So maybe to move now to the kind of consolidated picture. And if you don't mind, just for the audience, I think it'd be helpful just to sort of do a small recap of you know, where you are with Cogent Classic, as we call it, and then the acquired Sprint Wireline business. So Cogent Classic, which sells internet access to corporate and NetCentric customers, it has been facing some headwinds from the pandemic and post-pandemic corporate office vacancies, but nonetheless, was run rating about $225 million of annualized EBITDA, I think, before the Sprint Wireline acquisition. 2022 cash flow after finance lease, equipment installation payments was about $50 million. That sound about right?
That sounds correct.
Okay, great. Okay, and then on May 1, 2023, you acquired the Sprint Wireline business, and effectively, T-Mobile is paying you $700 million to offset the negative EBITDA, negative free cash flow, and the expense in CapEx that it's gonna take you to get that to break even, right? And then that $700 million dollar payment from T-Mobile, so you get $350 million the first year, so it's $88 million a quarter, and then the final $350 is gonna be over 42 months. So then that's gonna drop to about about $17 million a quarter after the first year. So the beginning—I mean, the middle of 2024.
Actually, a little more, about $24 million a quarter.
Oh, sorry. You're right. Times three. You're right. Got it.
It's okay.
You're right.
It's all right.
8 times 3. Okay, so simple math. In third quarter, you reported 131 of EBITDA, about $57 million, plus or minus, I think was legacy Cogent. You got $88 million from T-Mobile. So then we back into the Sprint business, there was a negative of about $13 million, right? But then you also had a benefit because you reclassified a lease to a finance lease. That was about a $13 million benefit. So you really, and you also had some other uneconomic finance lease payments. So I think what we're trying to back into here is where you stand today, how negative is Sprint business from a cash flow perspective? And I think it seems to be about $100 million or so.
But, so if you could clarify that, and then the bigger question, obviously, is: How do you get that to recover?
Really complicated, but it is not that complicated. So let me clarify at least four different points that you raised.... First, the Cogent Classic business, we sold three basic products. Internet access was accounting for 81% of revenues, 16% of revenues were VPN services, and 3% of revenues were co-location. That business was operating on a 61,000 route mile intercity fiber network and an additional 18,600 miles of route miles of metropolitan fiber, as well as 55 Cogent-owned data centers that comprise approximately 620,000 sq. ft. and about 69 megawatts of power. That business had not done any acquisitions for 18 years and had organically grown every quarter since going public in 2005, at a compounded rate of about 10%, with an average of 220 basis points a year of EBITDA margin expansion.
In the last four quarters of standalone business, we were doing slightly over an annualized rate of $600 million in revenue with 39% EBITDA margins, so about $240 million of EBITDA in that classic business. Because of the pandemic, the growth rate in that business had decelerated from 10% to approximately 5%, and our margin expansion was also decelerated to about 100 basis points a year. That business was slowly improving as companies return to office. That improvement is continuing in that classic business. We have reported all of our metrics for that business in a consistent manner, with no reclassifications or changes in reporting for 18 years. We acquired two things from T-Mobile. They were really two separate transactions, but they were tied together. We cannot do one without the other.
The Sprint Global Markets business was the nation's first wireline nationwide fiber optic network that was designed to carry voice. That network was built between 1982 and 1989, at a capital cost of $20.5 billion. It was designed solely to carry voice. The second thing that we acquired was a operating business serving predominantly large enterprise customers. That business had been ignored and not focused on for over 15 years. The revenue had declined from $40 billion to $565 million, and the business went from $16 billion of positive EBITDA to -$300 million when we signed the agreement to purchase that business from T-Mobile. In that transaction, we paid $1 for the physical network. 93% of the revenues that were in the operating business never touched the Sprint Network.
The Sprint Network was basically a fallow, dormant asset that had been sitting in place for nearly 15 years with no capital investment. Maybe what best summarizes it is the fact that across those physical buildings that I described, there were 22,500 cabinets or racks, full-size racks, 8 feet tall, of telephone switch equipment that hadn't been in service for nearly 15 years, but had never been removed. We paid $1 for that asset. The operating business had shrunk to just under 1,400 customers buying 28 products. In our due diligence, we determined that 24 of the 28 products were actually gross margin negative. As part of the transaction, T-Mobile, prior to closing, began a cost reduction program that included end-of-life-ing those products. So between signing and closing, the deal was signed in May 2022.
The deal closed, September 2022, closed in May 2023. In that period, T-Mobile spent over $500 million in operational burn and reduced the revenue stream from $565 million to $490 million. The EBITDA burn was reduced from $300 million to $190 million. We have said that we will further reduce that burn over the 12 months subsequent to closing by an additional $110 million, getting us to an $80 million annualized negative EBITDA one year after closing. We do that basically with three major tools. One, the consolidation of networks and the migration of traffic from off-net to on-net. Two, the elimination of uneconomic customer relationships that was begun under T-Mobile and is continuing. And third, the SG&A savings through headcount rationalization and facilities rationalization.
That will continue for the next two years, so a total of three years post-closing, till we are able to get that stabilized business of $450 million in revenue that we acquired, to a positive 20% EBITDA margin. So we will spend about $400 million of the $700 million we are getting from T-Mobile in that integration and subsidization phase as we turn that business around. At the simplest level, a shareholder or a bondholder can look at Cogent and say, "We issued no debt, we issued no stock, and our aggregated EBITDA went from $240 million-$520 million annually. Our margins increased from 39%-47.7%, with no additional debt or equity being issued." That is because of the stream of payments from T-Mobile.
They will step down from $29 million a month to $8.3 million a month in June of 2024, and continue for the next 42 months, for payment subsidies spread over a total of 54 months. Another way to look at it is, Cogent's EBITDA margins went, without the payments from T-Mobile, went from 39% down to approximately 20% because of the acquired businesses burning cash, but off of a revenue scale that was 80% larger. Then adding in the payment stream from T-Mobile, the EBITDA increases to $131 million in the last quarter. Over the long run, we will be operating a business that combines three components. One, the classic Cogent business. It will be all reported as a unified company, but that business will be a significant portion of our business.
The second thing that we will be doing is taking the physical assets that I described earlier, that today are fallow, and monetizing them through dark fiber sales, facilities leases, and then finally, adding a new product, optical transport. It is a market that's adjacent to where Cogent is, but is a market that we have never participated in, either as a buyer or a seller. It is about a $2 billion addressable market. In that market, we have four distinct competitive advantages. We have more endpoints on-net than any other provider. Our routes are completely unique. Over 90% of them are on right-of-way, that is not shared with any other provider. Third, we have very accurate, detailed engineering maps that give us the location of the network within one meter.
This was not a roll-up, this was not an amalgamation of acquisitions, but rather it was an organically built network by one owner for over 40 years. Then finally, we have a $1 cost basis, which will give us the ability to sell at whatever price it takes to gain market share. We are very encouraged by the initial response to selling transport services. We do have a significant amount of reconfiguration work that needs to be completed over the next 13 months or so, but at that point, we will have fully integrated the networks. On a reporting basis, Cogent has always reported consistently. There are two changes to our reporting metrics. One, we've added a third type of customer we never had, enterprise.
So now Cogent has corporate customers, NetCentric, which are basically wholesale, hyperscalers and access networks, and now we have very large enterprises, Fortune 1000 companies, as customers. The other thing is, every dollar of revenue at Cogent has always been reported four ways: customer type, I've just explained. Geography, U.S., rest of the world. On-net, meaning it's entirely on our network, no third parties, or off-net, where we buy a loop from another provider, and then by product. So Cogent historically had three products, now has four. So it still sells internet access, still sells VPN services, still sells co-location, and now has optical transport as a fourth product. Any on-net service, which our NetCentric business is 90% on-net, and our wavelength business or transport business is 100% on-net, carries 100% gross margin and 95% EBITDA contribution.
The real upside is our ability to take a portion of that dark fiber and repurpose it into a very efficient optical transport network.
Okay, great. That's super helpful, a lot. We've got, actually, just a few minutes left. I did wanna take all of this and shift to the issue of capital allocation, and then again, since we're at a debt conference. So we did start out, I think some people walked in late, talking about how there is a separation of some of the Sprint Wireline Assets into an unrestricted subsidiary versus a restricted subsidiary for all of the payments. But, nonetheless, from the public information that we have to work with, we really do look at leverage on a consolidated basis. So, two questions.
So one, right now, on an LQA run rate of about $525 million of EBITDA, company's leverage is 4.8 gross, 4.4 net. But when these T-Mobile payments start stepping down, once you pass the first year mark, do you expect leverage to initially increase, or will you be on a glide path where you'll be able to maintain or lower leverage over time? And then, you know, what the second part of this question is really on the dividend. So you currently have a dividend that annualizes to about $180 million a year, and that is in excess of your free cash flow.
So if you could just comment on the dividend, your outlook for dividend sustainability, and then basically, the overall capital allocation policy for the credit versus the shareholder return?
So first, let me start with the leverage. Your numbers were correct, pre-acquisition, but with an LTM test and using. We actually delevered last quarter by roughly one-half turn, both net and gross. Over the next three quarters, where we will have the full effect of the acquisition calculated, that delevering will continue. So Cogent has gross debt of approximately $1.4 billion, approximately a little over $1 billion of high yield in two tranches, secured and unsecured, and then finance lease obligations that count as debt of about $400 million, including that uneconomic finance lease. We will have EBITDA of approximately $500 million during that period.
So we're going to be, on an LTM basis, delevered down to a gross leverage of probably somewhere around 2.5 times, which is the low end of our guide range, and we will be on a net basis of even below that because of cash on the balance sheet. It is true that our EBITDA will step down as the T-Mobile payments step down, but that will be offset by the improvement in operating deficits from the restructuring work in the enterprise acquired business and the sale of wavelength services. So for the next five years, we anticipate our EBITDA to remain relatively constant at between $400 million and $500 million, against a total debt load that will actually come down. With that, we have raised our dividend for 45 sequential quarters consecutively.
Today, that rate of increase is at about 4.5%. We will continue to return capital to shareholders through either a combination of raising the dividend or supplementing that with share buybacks. Our goal is to operate comfortably in a net leverage range of between 2.5x and 3.5x EBITDA. After the T-Mobile payments end, we anticipate the business doing at least $1.5 billion in revenue, from the $1.1 billion run rate today, and EBITDA of about $500 million, with top line growth of 5%-7% on a combined basis, and about 100 basis points a year of margin expansion. Which tells us, in even this interest rate environment, we have a very prudently capitalized balance sheet.
Well, with that, we're out of time. Dave, it's a pleasure, as always. Thank you for making the effort and spending time with the investors at our conference.
Well, thank you for hosting me. Thank you to all.