Robert Reynolds @RobertMERC: "Below I'm making the argument..."

Spain

Below I'm making the argument for $CCOI (Cogent Communications Holdings Inc) which is arguable one of the most misunderstood company in the portfolio but one which I have very high conviction in. Happy to hear you're feedback. There are at least four distinct pieces here of the business, you have the legacy on-net/IP transit business, the inherited Sprint off-net and enterprise runoff, the long-haul wavelength opportunity, and the non-core data-center/facility monetization. Treating all of those as one declining completely misses the forest for the trees. The T-Mobile payments were part of the economics of taking on a distressed Sprint wireline asset. The point of those payments was to bridge Cogent while it cut uneconomic inherited revenue, reduced network costs, and converted the useful parts of the acquired asset base into higher-quality revenue. A large part of the inherited Sprint revenue was off-net, enterprise, low-margin, non-core, or dependent on third-party last-mile/tail-circuit costs. So pruning that revenue is not necessarily failure its necessary despite on the surface it is giving the impressions of a dying business. If you look at the margin profile, you will see why the declining revenue is necessary The most important asset for Cogent is the long-haul wavelength platform. That is a different business from basic IP transit and a different business from owning data-center facilities. Long-haul wavelength is closer to an oligopoly-type market because credible scale requires scarce fiber routes, rights-of-way, lit locations, optical equipment, operational capability, and the ability to provision high-capacity transport across large geographies. So there are only 3 players in the North American market that can provide the long-haul solutions, Lumen, Zayo and Cogent. Of the 3, Congent is the only one that offers redundancy running along rail lines rather than their peers via highways. In an AI/cloud/data-center interconnect world, the redundancy is going to be paramount because there are multi-billion dollar data centers that can afford zero down time. Cogent’s wavelength business in my eyes is the crown jewel and will act as a utility with stable, consistent and long term growth in revenue. Wavelength revenue, connections and ARPU are all growing strongly but it’s small so it can’t carry the company yet. But dismissing Cogent as only a declining backbone misses the part of the business that is improving and strategically relevant. Cogent own data centers, data-center, these were not the core reason to own. Many of these facilities were excess switch-center assets inherited through the Sprint transaction which cost $1. Cogent invested $100 million to convert dead or underutilized facilities into commercial data centers and edge data centers so they could be monetized. The company is selling 10 former Sprint facilities for $225m, equal to roughly $4.3m per gross MW or just over $6m per net IT MW. Those 10 facilities had only about $30m of direct capex invested into them, implying a 7.5x return on the capital deployed. The remaining 14 marketed facilities represent 56 MW gross and roughly 40 MW net. Applying the same transaction valuation implies another ~$240m of potential proceeds. This is all money in the bank. The balance-sheet impact is now much cleaner because this cash inflow is ear marked for debt reduction as per the amendment of its 2032 bonds. If roughly $225m of proceeds are used to buy back debt carrying a 6.5% 7.0% cost, that alone implies roughly $14m to $16m of annual interest savings, before any extra benefit from repurchasing debt below par. A lot of the “leverage” is actually operating leases that are accounted for on the operating section of the income statement with no end balloon payment which this data center sale removes most of the bearish argument related to debt. There is also another non-core digital asset sitting inside Cogent that the market tends to underappreciate, IPv4. Cogent owns roughly 38 million IPv4 addresses and was leasing 15.2 million of them as of Q1. This is a very different asset from the data centers, but but quite similar, Cogent inherited scarce digital infrastructure (for the same $1 price tag) that can be monetized with very little incremental input. Leasing blocks of IPv4 addresses is a high-margin cash stream, and the company has already used the asset base to support IPv4 notes. As lease performance improves, restricted cash tied to that financing is being released. So alongside the data-center sale, IPv4 is another example of Cogent turning non-core digital assets into balance-sheet flexibility and cash flow. Cogent only needs time 😊

Not investment advice. The author may have financial interests in the mentioned instruments.

CCOI

Cogent Communications Holdings Inc

9.92

0 (0.00%)

1 Mentioned

CCOI

Cogent Communications Holdings Inc

Strictly Necessary Cookies

Performance Cookies

Functional Cookies

Targeting Cookies