Hidden Value – Cogent Communications
Cogent’s (NASDAQ: CCOI) acquisition of Sprint Wireline remains one of the most interesting deals we’ve ever seen. We’ve talked about the Wavelength opportunity in another post, but it’s worth diving into a few other elements that make the deal so uniquely appealing to Cogent. Dave Schaeffer, Cogent’s CEO, has mentioned other pockets of “hidden value” a few times. This post will unpack what he referred to in this quote from a conference presentation:
“And just as we do with other assets, there are things that we discover that we can bring value out of whether it be the 1,300 pieces of fee simple owned real estate, whether it be the 9,750,000 incremental IPv4 addresses that we acquired, whether it be dark fiber, wholesale colocation services, our goal was to find pockets of value, that have really been ignored by a larger company and it’s something that we’ve done in the past.”
Background
Public companies with “hidden value” typically have a difficult time realizing that value in their mark-to-market valuations. Many investors are loathe to invest in sum-of-the-parts stories, especially if parts of the business are in secular decline or are subject to competitive pressures. Investment theses that involve some flavor of “…and you get this segment for free!” rarely generate attractive IRRs. Usually they involve some combination of “good” and “bad” businesses, which leads to a group discount because nobody wants to own the “bad” to get the “good.” Thankfully, Cogent has positioned its assets and business model in the “good” growth areas of telecom – the internet and growth connectivity markets. It has built one of the few enduring growth telecom companies with great incremental margins.
Hidden value can be extracted in two primary ways: (a) generating operating cash flow from accelerating growth, and (b) selling/monetizing assets. We think Cogent can unlock its hidden value through a combination of both methods, but most will come from accelerating growth and generating more operating cash flow. We think this is the healthiest and best way to generate enduring equity value for shareholders.
Now, let’s enumerate the pockets of value Cogent received from its acquisition of Sprint Wireline and build on Dave’s quote from above. Not all of them are hidden.
Cogent will receive $700 million of cash payments from T-Mobile in the form of a take-or-pay IP transit contract. $350 million will be paid in equal monthly installments during Cogent’s first year of ownership, and the other $350 million will be paid in equal monthly installments for the following 42 months. At a 15% discount rate, the NPV of these payments is nearly $530 million.
It received 9.75 million IPv4 addresses. These trade for $50-60 per address, implying a pre-tax value of about $500-600 million.
It received 1,300 technical facilities owned fee simple (i.e. owned outright, no debt). 41 large buildings will be converted into colocation data center facilities that can be sold, leased on a retail basis (by the rack), and/or leased on a wholesale basis (by the building). Those 41 larger facilities will have 130 megawatts of power available. We believe these facilities could be worth $1 to 1.5 billion of enterprise value in a sale.
It received nearly 19,000 route-miles of intercity, long haul fiber and 12,570 route-miles of metropolitan owned fiber. This network was appraised at $1.2 billion of asset value by E&Y.
It received an operating business that was generating about $450 million of revenue but carried a cost base of about $630 million at closing (-$180 million of EBITDA). Cogent expects to extract over $220 million of cost synergies after completing its network integration work, turning Sprint Wireline into a cash flow positive business within a few years. The terminal value of this business is likely around $200-400 million.
If we add these up, we get aggregate value of about $3.5 to 4 billion from these five distinct areas. Post-tax value at a 25% rate would be about $63 per share of value to shareholders. However, selling these assets outright isn’t their highest value and best use, and patience will be rewarded. Let’s walk through why.
To Operate or to Sell?
With each bucket of assets, Cogent’s leadership team must analyze whether it’s more accretive to operate the businesses or to sell the underlying assets. In most cases, we think it makes sense for Cogent to operate the businesses, grow its revenues, and generate more operating cash flow. Below we’ll walk through the strategic options that Cogent has or will be analyzing:
1. We think there are huge pockets of value from extending its business into the Wavelength market (as discussed in our prior Insight), but that market would be inaccessible if it sold the physical fiber it acquired for $1.2 billion. We think the NPV of the Wavelength and Dark Fiber opportunities is far greater than the $1.2 billion of asset value. Therefore, Cogent should not sell the physical network. A $500 million annual wavelength business and a dark fiber business that could generate $300-500 million of upfront payments and $100 million of annual recurring revenue is worth significantly more – likely upwards of $6 billion at conservative valuations (and over $12 billion in more bullish scenarios). With the acquired fiber assets, Cogent has chosen to generate operating cash flows from new, additive business segments that sell network capacity. We wholeheartedly agree with the company’s choice, and it’s also worth noting that the compensation committee recently tied Dave’s cash incentive compensation to Cogent’s Wavelength business.
Cogent is pretty uniquely positioned to absorb Sprint’s assets and repurpose them into a new growth area in a net additive way. Other fiber companies could have acquired and integrated those assets, but they likely wouldn’t have add tremendous net new revenues since they are incumbent providers of wavelengths and dark fiber already. They would pick up some value from new unique routes, but their net incremental growth would be more limited.
A non-incumbent provider likely wouldn’t have the existing network infrastructure to generate synergies on the money-losing Sprint business. It would be too risky for most management teams to absorb -$200 million of annual cash flow in order to bet on a growth business in Wavelengths and Dark Fiber that was starting from almost zero and very few overlapping customer relationships. Conversely, Cogent has significant overlap (“virtually all”) between its Transit customers and future Wavelength and Dark Fiber customers.
2. What about IP addresses? Core Cogent owns nearly 30 million addresses on its own and is one of the largest owners of IPv4 addresses in the world. Every IPv4 address has been allocated, and they now have scarcity value until IPv6 becomes the standard (if it ever does). As mentioned above, it acquired 9.75 million addresses with Sprint Wireline. These IP addresses carry zero book value, but are worth $50-60 per address. In fact, T-Mobile sold >2 million addresses in 2022 and recognized a $120.8 million gain (see below). These addresses have significant value in the market.
Source: Cogent Communications 8-K, July 17, 2023
Let’s say T-Mobile sold 2.2 million addresses. The $120.8 million gain would imply $55 per address, right in the range that is referenced by transactional marketplaces for IP addresses (there is a difference in value by block sizes, but that is too far in the weeds for this post).
Currently, Cogent leases about 30% of its IP address portfolio at about $0.30 per month, or $3.60 per year. 30% implies about 11 million addresses leased and about $40 million of annual revenue. This revenue comes at 100% margin. It bundles these address leases with its IP Transit services. Would we prefer Cogent to sell them at $55 per address, or lease them at $3.60 per address? $55 / $3.6 = 15.3x FCF – a nice multiple, but actually a discount to Cogent’s valuation. In this case, it seems sensible to keep monetizing them via leasing rather than selling and rolling them into the base of FCF that can be used to distribute more dividends over time. However, if Cogent cannot easily lease all its IP addresses within a reasonable time frame, it could be accretive to sell a portion. We would be in favor of selling a portion of the Sprint addresses to increase flexibility and optionality on the balance sheet, and because it seems unlikely that all could be leased quickly enough to justify not selling. If it were easy to lease 100%, it would have done so already. (Interestingly, on February 1st, 2024, AWS will start charging $0.005 per hour to use its IPv4 addresses.)
Unlike the point above, Cogent’s ability to sell IP addresses in the market is not unique. Anyone with this portfolio could monetize it. In total, Cogent’s IP address portfolio is worth about $2 billion, and it could probably sell up to $500 million without impairing its long-term growth prospects.
3. Now let’s talk about the colocation/data center opportunity. Digital Realty Trust (NYSE: DLR) and Equinix (NASDAQ: EQIX) trade at very healthy multiples (21x EBITDA and 22x EBITDA, respectively). Hyperscalers and AI companies are desperate for more space and power, especially as they ramp up their infrastructure to support generative AI applications. Cogent is in the process of repositioning its large Sprint facilities to be ready for those customers, and its work will be complete within 2024. Once complete, it will have the option to sell or lease those facilities opportunistically. If Cogent were to lease them, we think they could generate $100-130 million of annualized revenue and $50-70 million of annualized EBITDA, implying enterprise value of $1 to 1.5 billion. Interestingly, these projections have not incorporated in Cogent’s long-term guidance, i.e. they present upside to growth and EBITDA targets and/or Cogent may have a preference to sell them.
In this case, the industry’s multiples offer a premium to where Cogent trades and therefore, on paper, an outright sale of the portfolio appears accretive to shareholders. However, the company could use that available space and power to improve its value proposition and win rate (via bundling) in other parts of its business (e.g. IP transit and/or Wavelengths), and therefore it could be net accretive to keep these assets within Cogent to maximize its probability-adjusted enterprise value. Additionally, it certainly wouldn’t be a bad outcome to lease up those facilities and generate another >$1 of FCF/share annually.
Anyone with unutilized data center space and power could sell these assets, and there is high demand for them in the current environment. This monetization opportunity is not unique to Cogent.
4. Many of Cogent’s other 1,300 Sprint technical facilities and other assets (e.g. a small portfolio of cell towers) can be sold off without any impact to the business. However, some must be retained as network regeneration sites to operate the physical fiber network. We believe there may be another $50-100 million of value that could be realized without affecting future growth.
To summarize, even though Cogent acquired around $4 billion of pre-tax asset value, we believe the assets that the company should consider selling are (a) the Sprint colocation facilities, (b) the non-core Sprint technical facilities, and (c) a small portion of IP addresses. The value from these, combined with the cash payments from T-Mobile, could reach $1.5 to $2 billion without sacrificing any future growth embedded in the company’s long-term guidance.
The other roughly $2b of asset value is put to better use by growing revenue and operating cash flow, which we think could generate more than $10 billion of enterprise value to Cogent. Building a significant Wavelength business and opportunistically selling Dark Fiber will drive many years of healthy growth in high-margin revenue, all of which can be used to fuel future capital returns.
Putting it all together, we think the Sprint Wireline acquisition should catalyze value creation to the tune of over $11-12 billion, a minority of which could be realized relatively quickly via asset sales, and the vast majority of which should be recognized as Cogent accelerates revenue growth from its expanded asset base. We think this ratio is extremely attractive for a “hidden value” situation because so much will be unlocked through an acceleration in growth, which is highly coveted in the telecom market and under Cogent’s control through solid execution. Additionally, it won’t hurt that the secular tailwinds driving Cogent’s revenue acceleration will come from serving core modern cloud and AI infrastructure companies. Being an arms dealer to those industries is not a bad place to be.
Conclusion
In the next 12-18 months, we believe Cogent’s enterprise value (currently $5 billion) could be rapidly adjusted down by up to $1.5 to $2 billion as it collects cash payments from T-Mobile and if it sells some of its monetizable assets, as discussed above. Should it choose to, selling colocation facilities, IP addresses, and/or non-core Sprint facilities could rapidly recapitalize Cogent’s balance sheet which would likely increase near-term capital returns to shareholders.
We wouldn’t trust the likes of AT&T, Verizon, Lumen, or any other traditional telecom company to execute on a large portfolio of value creation with so many pockets of asset and business value. However, Dave Schaeffer, Cogent’s Founder, Chairman and CEO, is uniquely qualified to do this. For anyone doubting his capabilities, we recommend watching our recent interview with him. Dave is wired for capital allocation, value creation, and positioning his companies into secular growth markets (while shedding the “bad” businesses). At Cogent, he has done so by buying his competitors for less than free (cumulatively, he has been paid net cash to acquire them) and then scrapping them for parts and/or repurposing their assets and businesses to focus on growth areas – historically within the internet connectivity markets where there is very high customer density (multi-tenant office buildings and carrier-neutral data centers).
Once the Sprint assets are fully integrated by the end of 2024, Cogent, will be able to access and address large, new end markets with a disruptive operating model, differentiated internal tools and processes to run and provision services more efficiently, and a negative cost basis to price aggressively should the need arise. Cogent’s competitors are saddled with mountains of debt; high-cost, low-efficiency legacy businesses; heterogeneous and less efficient networks; and significantly less cost discipline…but they have all the market share. It’s a ripe opportunity for Cogent to steal existing market share over time and to be significantly disruptive when competing for incremental growth in the market.
