So why am I buying? The reason is that if you look under the hood, Cogent is fine. The core business is the same as always and the disastrous Sprint Wireline acquisition is on the cusp of turning from a headwind into a tailwind.

Over the next couple of years Cogent will return to top and bottom line growth, net leverage will come down, and investor confidence will be restored. It won’t be a smooth ride, but I’m willing to sit through any near term pain.

The Architect Behind Cogent: Dave Schaeffer

Dave Schaeffer is one smart cookie. At age three he was doing his 14 year old brothers homework, and by age 12 had started College. Upon arriving at the University of Maryland Dave realized that there was no limit to the number of courses you took. Always quick to recognize a good deal, he signed up for 40 credit hours worth of courses, more than double the average student.

“When I went to school, I tried to be efficient. I was too young to drive, so I would get up at about five in the morning, get on a bus around six, and be at school by eight to take my first class. I would typically leave around six in the evening. I literally took classes all day long.

They had a great deal: you could take as many courses as you wanted as an undergraduate for a flat fee of $349 per semester. That was a real bargain…

There was a second loophole, which was that you could take graduate-level courses, which were usually charged by the credit hour, if you met the prerequisites and could do the coursework. So, effectively, I stayed enrolled as an undergraduate but was taking all graduate courses.”

- Dave Schaeffer

Dave took a variety of classes across, physics, economics, and history. His one criteria was that they had to be in adjacent buildings to one another. Less time walking between classes meant more time learning. If that wasn’t enough, Dave also worked part-time in his little remaining free time.

Just two years after starting College Dave was already working on his PhD. However, after passing the last of his PhD examinations and just before defending his dissertation, he dropped out. The reason? His fathers taxi cab business was in big financial trouble. So in 1973 with the business a million dollars underwater, Dave joined the business along with his older brothers to help turn things around. He was just 17 at the time.

In short order the business was profitable and back on sound footing. It would eventually grow to over 2000 cabs, making it at one point the second-largest delivery and ground transportation enterprise in the country.

By the mid-80’s the business controlled enough of the market that they were running into political opposition. Rather than continuing to gain market share, they expanded into two adjacent businesses. The first was buying the land underneath of competitors. This further cemented their competitive position and was a great place to park capital. Cab companies are land-intensive because they need lots of parking and at the same time have to be close to central business districts. This made the real estate land-rich, building-light, and well located. Although they didn’t fully appreciate it at the time, these locations would prove to be lucrative covered land plays.

“We ultimately did 47 deals together. We still own five of them, and they are really covered land plays. Again, it is good to be lucky. We did deals where the convention center was built, where the second convention center was built in D.C., and where the baseball stadium was developed. These were neighborhoods that rapidly gentrified, and on an absolute basis the properties increased 10x or 20x in value.”

- Dave Schaeffer

The second expansion was into insurance. The largest single expense in the cab business, even more than the car itself, is insurance. So in 1979 with $1.5 million in seed capital they started a property & casualty insurer. It was the first of its kind in DC since Geico opened up shop in 1937 and still exists today.

“So we bought the underlying real estate of our competitors. We insured them, and we owned their underlying real estate. While we did not own them, we had effective control over virtually 100% of the industry.”

- Dave Schaeffer

The insurance business would quickly scale to $25 million in annual premiums and ended up with 90% of the market for cabs, limousines, couriers, and ambulances across the tri-state area. The capital generated from their insurance operations was was fed back into their real estate business.

“Our loss experience ratio was generally around 86% of premiums. But the money you make in insurance is really the investment income. We did some real estate deals through it, some mortgage underwriting, and some more traditional business investing. Pretty quickly, we had built a capital pool of about $50 million back in the early 1980s, and that became our base of investment capital.”

- Dave Schaeffer

Concurrent to building the D.C. taxi empire, Dave made a killing trunking specialized Mobile Radio Spectrum (SMR). It started when Motorola, which dominated the radio market, moved from direct distribution to a dealership model. Seeing the obvious opportunity to sell into his taxi cab network, Dave become one of those dealers.

Right around the same time the government opened up a new band of spectrum for private mobile radio. Importantly, you did not have to pay for it if you met certain loading requirements. To prevent spectrum hoarding and speculation, the FCC implemented rules requiring operators to place at least 70 users per channel within five years before you were granted permanent property rights.

While others struggled to find enough users to meet the requirement, Dave turned to his radio dealership. He traveled the country offering cab companies free radios and computerized dispatch systems. In exchange, those cab fleets had to operate their radios on his spectrum. Because cab fleets have hundreds of drivers, Dave instantly met the government’s 70-user mandate, triggering the permanent property rights.

If you had five or more adjacent channels, you could do something known as trunking. This combined channels into a shared pool that could function as a private cellular system.

“As individual channels, they were not valuable. So the typical model was to go into a market, get seven frequencies, find an operator who could meet the 500-unit loading requirement, strip off two as operating channels for the cab companies, take the remaining five, trunk them, and then sell them to a mobile aggregator.”

- Dave Schaeffer

Once he legally owned the spectrum assets, Dave would peel off a small portion for the cab companies to keep using before flipping the remainder to what would eventually become Nextel. In total, he completed 180 SMR channels at a cost basis of about $35 per channel, his average selling price on those same channels came out to $175,000.

In 1994 after 17 years running the cab business, Dave decided it was time to move on and handed things over to one of his brothers. Still only in his mid-30’s, Dave didn’t sit still for long. He would next go into the paging business which meant starting a network carrier from scratch. A couple of years later he sold it for $12 million having invested $1 million growing the business to 65 towers and 60,000 subscribers up and down the U.S. East Coast. The real win however was the $40 million in tax assets he retained. To keep growth capex low Dave insisted subscribers own their own equipment. Rather than outright sales he instead leased equipment to subscribers with a penny buyout at the end. This kept the operation asset light while still allowing him to retain the deprecation for his own account.

Looking to start his third technology business, Dave turned to the burgeoning telecom led fiber buildout and started PathNet. He turned regulatory complexity into an advantage and strung together a series of clever deals to create and aggregate excess spectrum capacity, and to secure continuous right of ways along thousands of miles of pipelines and railroad tracks. With the Spectrum and right of ways secured, Dave began building out a long-distance network serving rural markets, coming to market as a “carrier’s carrier”.

At the time, traditional telecom monopolies heavily marked up long-distance call termination fees in rural areas, charging anywhere from 10 to 15 times urban rates. PathNet’s network would bypass those incumbent networks entirely. On one end, Dave secured supply agreements with various utilities companies, on the other, he locked in take-or-pay contracts with major long-distance giants like AT&T and MCI to carry traffic for them at much cheaper rates.

By 1997, the telecom sector was on absolute fire. PathNet aggressively levered up to accelerate network construction and in August of 1998 were ready to go public. Pathnet was running a fully marketed IPO roadshow to raise equity to rebalance their top-heavy $3.4 billion debt load which was backed by just $36 million in equity (yes, you read that right). But then, disaster struck. Just one day before PathNet was scheduled to price its IPO, Russia defaulted on its debt.

”We were 7x oversubscribed on Tuesday and by Thursday we couldn’t pull a book together.”

- Dave Schaeffer

After the failed IPO, the board decided it was best to bring in a more experienced telecom CEO and Dave sold his position in the company for a profit of $36M in March of 1999. Five months later Dave started Cogent and in the Fall of 1999 raised $500M of equity. The plan: to build a single-product fiber network optimized entirely for ultra-cheap, ultra-fast internet transit.

Before Cogent was able to deploy their newly raised equity, the dotcom crash and resulting telecom meltdown of 2001 stopped them in their tracks. Just as Cogent found themselves flush with cash, fiber assets were becoming available for pennies on the dollar.

”I sat down with my 10 investors at that April [2001] meeting and said, ‘There are three paths forward. Path number one: let’s just go home. We’ll liquidate. There’s $450 million, we’ll split the money up and go home.’

Well, none of the investors liked that answer, because I owned 23% of the company and had only put in $7 million. They had put in $500 million and owned the remainder.”

- Dave Schaeffer

After rejecting Dave’s bold option number two asking investors to double down, they settled on option three: in exchange for subordinating his equity stake, the remaining cash was allocated into a blind pool giving Dave total discretion to go out and buy bankrupted competitors.

Dave would go on to execute one of the more ruthless and capital-efficient distressed-asset roll-ups in corporate history. And when I say ruthless, I mean ruthless.

One of Cogent’s first acquisitions was PSINet. They were the world’s second largest internet provider at the time of their bankruptcy in 2001.

“PSINet had $4.3 billion in debt and $300 million in cash. The secured creditors, mostly telecom equipment vendors, were owed $600 million and wanted to get all the cash quickly before it burned off. The unsecured creditors, who held the other $3.7 billion of the debt, were holding up the bankruptcy in hopes of cutting a better deal.

Schaeffer played the two groups against each other expertly. He persuaded both sides to let him buy PSINet for $10 million and shut down the business. That halted the cash burn, pleasing the secured creditors. His next move did not.”

- Telecom Knockout, Sep 25, 2008

After acquiring PSINet, Dave moved all of the routers and switches into warehouses and told secured creditors they could have their collateral back. They didn’t collect, so Dave had the court issue an abandonment order which voided the equipment from the bankruptcy. With no collateral, the secured and unsecured creditors were now on equal footing. The $300 million of cash was then parceled out pro rata to both groups of creditors.

This arrangement obviously favored the unsecured creditors. Because of this they agreed to pay Cogent $40 million which covered the $10 million purchase price and $30 million in severance costs. This left Cogent with all of PSINet’s assets at no cost.

Between April 2001 and December 2004, Cogent would evaluate 121 targets, bid on 19, and successfully close 13. Of those 13 companies, 6 had been public and at the peak of the bubble commanded an aggregate market cap of $61 billion. In total, Cogent spent only $60M buying all 13 companies. Those same companies had a combined $14 billion of invested capital poured into them by their original backers.

With a massive global backbone cobbled together from the wreckage of the dot-com bust, Dave shifted to a laser focused organic playbook

The Cogent Playbook

Cogent was built on the premise that the internet would ultimately be the only network that mattered.

“It was all about finding a cost-effective way to bring revenue onto the network. You sell only the product that is in structural ascendancy and ignore the products that are in structural decline. The internet fit that description, so we built a network focused solely on internet connectivity.”

- Dave Schaeffer

Dave recognized this 30 years ago, deciding to ignore the long-term losers like voice and video and instead stay laser focused on internet connectivity. That laser focus was built on five foundational pillars.

1. Bandwidth would become a commodity, and to survive, you had to be the lowest-cost producer.

2. Local networks are natural monopolies. As a new entrant attempting to overbuild those networks, you had to be highly selective about where you deployed capital.

3. You needed to achieve scale quickly. A successful network needed everywhere-to-everywhere connectivity.

4. You needed to take advantage of changes in technology more effectively than your competitors. That meant using Ethernet and running IP directly over DWDM.

5. You needed a compelling value proposition to lower the cost of customer acquisition.

In addressing these pillars, Cogent decided to focus on two distinct markets. The first (which Cogent labels Corporate), was large, multi-tenant office buildings in central business districts of major North America cities. The second (which Cogent labels Net-centric) was bulk transit, wholesale bandwidth in data centers around the world. In addition to having ultra low customer acquisition and build-out costs, these two markets complement each other perfectly. Office buildings are inbound-traffic heavy and see peak usage during the day while data centers are outbound traffic heavy with peak usage in the evenings during primetime streaming and gaming hours. This helps keep Cogent’s network balanced and increases utilization throughout the day.

The reason Cogent has been able to capture 25% of global internet traffic can be boiled down quite simply:

“In this business, or in any business, in any industry, the operators of the business have to answer three very simple questions. I am always shocked that public management teams do not think in these terms.

First, do people need what I am selling? Second, why will they buy from me versus the other guy? And third, if they buy from me, how do I make money?

Those are the three questions, whether you are a car manufacturer, an internet provider, a restaurant operator, or a clothing store. Every business has that common set of questions.

The way we create value here is that, honestly, while we are in a bad neighborhood, we are walking. We do not have to create demand. I do not have to take out ads during the Super Bowl to convince people to drink brown sugar water. Everybody needs the internet, so everybody needs what we have.

In terms of the second question, why will they buy from us? Because we deliver more value. It is that simple. It is not because our salespeople are smarter, or look better, or have a fancier business card. It comes down to customers making a rational business decision based on value.

And [to answer the third question], the operating leverage we have established shows that we can take that revenue and convert it into meaningful cash flow. That is how we create value.”

- Dave Schaeffer

In its end markets, Cogent is the lowest cost provider of internet by far. They offer the lowest prices while also providing superior reliability, service, and speed. Cogent is able to trounce it’s competitors on price/value and still make money because of it’s unique, purpose built network that commands the lowest costs per unit by a wide margin.

Unlike the incumbents whose networks support legacy voice, private circuits, cable video, emergency services, and so on. Cogent’s network is purpose built for internet. Their 100% fiber network runs Ethernet-only protocols and they sell 4 core products vs their competitors who sell 100’s of products. Less layers, less equipment, and less complexity translates directly into lower operating costs and faster provisioning times.

Cogent’s network is completely standardized throughout the world, leading to yet further savings and operational simplicity.

“[As of March, 2022] our network is about 60,000 route miles terrestrially. Spanning every continent. In addition to the long haul network, there’s about another 17,000 route miles, about 40,000 fiber miles of metropolitan fiber in 215 markets and about a 1025 rings. Between those locations, we have just under a 1,995 optical transport nodes. And if you walked in to an amplifier site in Ukraine, in Finland, and Sydney or in, you know, Kansas, the rack faces, the diagrams will look identical. It’s all about standard. One vendor. We are end to end a Cisco shop for routing and transport.”

- Dave Schaeffer

“So product standardization, system standardization, we have one provisioning system, one CRM system, one network management system, one billing package globally.”

- Dave Schaeffer

Advances in optical transport and routing equipment have brought the cost per mbps down at about a -30% CAGR for the last 25 years. Cogent’s standardized, purpose built network allows them to capitalize on these cost reductions faster than anyone else.

These Cogent unique attributes create meaningful advantages over competitors. To compete on the same footing as Cogent, someone like AT&T would have to drop all of their voice and cable customers (who account for the majority of revenue) and reconfigure their entire network.

Corporate

Of of the 1.1 million office buildings in North America, Cogent has zeroed in on only the most profitable buildings. A number that stands at 1,875 today. Those buildings average 41 stories in height, 553,000 square feet and 51 distinct businesses/tenants.

While this might seem like a small number, these large A class office buildings that Cogent caters to account for a disproportionate amount of all multi-tenant office space in the U.S.

“By being hyper-focused on the buildings we selected, and by buying dark fiber from other providers to get as close to each building as possible, we minimized the amount of outside plant we had to construct.

We would then pre-wire the building with a vertical riser and install breakout boxes on every third floor. That meant that, when we connected a customer, the horizontal run was never more than one floor up or one floor down. This kept the installation cost to approximately $600 per subscriber.”

- Dave Schaeffer

Cogent has pre-wired the most lucrative office buildings in North America. This means there is very little incremental cost to signing up a new customer and install times average only 9 days.

Importantly, Cogent’s competitors choose not to pre-wire. If AT&T or Verizon want to provide service and aren’t already connected to your unit, they have to run cables all the way up from the ground floor. This takes closer to 90 days and means getting riser access from the landlord.

Thanks to Cogent’s single purpose network, for the same monthly price you can get 1Gbps of dedicated internet access from Cogent vs a competitor who only offers 100 Mbps on shared infrastructure. Same price, 10x the speed, and faster latency.

Net-centric

Net-centric is Cogent’s wholesale, high-volume bandwidth business. Despite accounting for less than half of the groups revenues, it represents over 95% of total network traffic.

Similar to the Corporate business, Cogent has been highly selective in choosing the 1,744 data centers they serve across the world. These data centers are carrier-neutral, third-party facilities where customers are already co-located. Connecting a new customer requires no incremental spend.

“We sell a metered service at a lower price per megabit and are now the largest carrier of internet traffic in the world, carrying approximately 1.8 exabytes of traffic across our network each day. Just over 25% of global internet traffic travels across the Cogent network every day.

We are the most interconnected network. We have more autonomous systems directly connected to us, operate in more data centers and have more capacity. The better connected you are, the better your service performs, because there are fewer intermediate autonomous systems between the sender and the receiver.”

- Dave Schaeffer

Cogent's network is bi-directional and architected in rings. This makes it fully redundant, there's physical diversity to every customer so they are protected in the event of a fiber failure or catastrophe.

Off-net vs On-net

Cogent’s network reaches the entire global internet, while that seems impressive, anyone can pull this off. That’s because if your network doesn’t connect to somewhere you can simply use your neighbors network to get your data where it needs to go. The catch though, is that your neighbor is going to charge you for it. One of the key characteristics of a network is therefor how much of your traffic remains on-network vs off-network.

If your data can get from A to B without leaving your network, it stays on-net. You don’t have to pay someone else a fee for using their network. Because fiber capacity is essentially unconstrained, there is no incremental cost to sending more data, so on-net revenues come in at 95% margins.

If however, you have to use someone else’s network to get your traffic where it needs to go, they’ll charge a fee. There is an incremental cost to sending more data, so off-net revenues come in at ~45% margins.

This brings us back to the impressive nature of Cogent’s purpose built, optimized network. Through just 3,600 buildings, Cogent is able to keep 3/4’s of all revenue on-net. While a competitor like Lumen has the same degree of network ubiquity, they do so through a much larger footprint, think hundreds of thousands of buildings. That gets expensive to maintain, further widening the cost advantage Cogent enjoys.

Capital Allocation

Dave Shaeffer has a rare combination of technical knowledge, operational know-how, and business acumen. While these traits are required to be a great operator, they are only part of what makes a great CEO. Equally if not more important is the topic of capital allocation. I believe Dave’s record in this regard to be excellent. The below quotes that I’ve pulled from various interviews and earnings calls offer some insight into how he thinks about allocating capital.

“We’ve been pretty consistent in producing free cash from operations since mid ’06. Over the course of the past four years we’ve used that cash for different things. First we bought back a significant amount of stock in ’06, ’07, and early ’08. We then shifted in mid-2008 and started buying back our convertible debt at a deep discount, and we really bought back all we could get our hands on. Then in late ’08 we shifted gears yet again and started accelerating our footprint expansion, buying a lot more dark fiber both in terms of metro routes and long-haul routes. In late ’08, ’09 and ’10 we almost doubled the footprint of the network, giving us a much larger addressable network.”

- Dave Schaeffer, 2011

“We’re very focused on our return on incremental capital invested. We think that based on our product set and our pricing model, smaller buildings or buildings with less tenants in them will not provide us an adequate return on capital.”

- Dave Schaeffer, 2011

“…and most important, is price. Cogent exists today in large part because of our very disciplined price structure in terms of what we would pay for assets. We reviewed a lot of the transactions in the market and the prices that the sellers were looking for, and we concluded that we were not going to participate, as our prices were far lower than what the sellers were expecting to get. That could change in the future, but for right now I think it’s fair to say we’ve been priced out of the market.”

- Dave Schaeffer, 2011

“A lot of fiber companies don’t generate free cash and are focused on growth and growing the footprint. We are focused on growing profitability.”

- Dave Schaeffer, 2013

“We look at numerous opportunities globally, not just North America and Europe but Asia too. We probably review 150 opportunities each year, and we’re willing to look at lots of different business models to bolt onto ours. We have a two-part filter for M&A. The first test is whether it is cheaper to buy the whole business than to buy the individual assets we want. Most of our past acquisitions have fallen into that category. The second is that we’ll look at different business models to buy if we can generate free cash and a return on capital greater than our cost. We’re even willing to buy businesses in decline, they just have to be priced appropriately. Whether it’s managed hosting, VoIP, disaster recovery, video distribution, quite honestly everything in our sector is dramatically overvalued on a free cash flow generation basis. While valuations can be slightly better in Europe than in North America, I believe they are still are too expensive on a discounted free cash flow basis.”

- Dave Schaeffer, 2013

“In 11 years, Cogent has looked at 525 additional acquisition opportunities and has done due diligence on those targets and has not done a single target. There are three reasons why: we don’t need physical assets; we’re comfortable with the scale we have and we’re not comfortable to pay a premium for scale; and most importantly, valuations are unrealistic.

In looking at growth businesses we were basically trying to understand how long that growth would continue and looking at flat or declining businesses what that rate of decline would be. Even when factoring in potential synergies, we could not find a single target that fit that [return on capital] threshold.”

- Dave Schaeffer, 2016

“…It creates a recipe for investors to overbuild and that’s a recipe to destroy value. If Google and Verizon can’t do it, then nobody can do it, because their cost of capital is cheaper than anybody else’s”

- Dave Schaeffer, 2018

“Stock prices may oscillate based on market conditions, but growth in free cash flow per share can only occur through execution, prudent capital allocation and good corporate governance.”

- Dave Schaeffer, 2019

“And as I’ve said repeatedly, I go to the Board and if I saw inorganic opportunities that I was convinced would produce better returns, I would ask for the allocation of capital to that. And we have worked at literally hundreds of potential transactions. And as Sean said, the highest and best use is to give the money to the shareholders.”

- Dave Schaeffer, 2020

“Why did we buyback stock? It was simple arithmetic. The yield on our stock was substantially higher than our incremental cost of debt capital. So if we can rent capital less expensively, and retire permanent capital, it makes economic sense.”

- Dave Schaeffer, 2020

“We’re very cognizant of tax efficiency. The fact that we went from 63% to 79% [of the dividend] being characterized as return of capital kind of skews us more towards dividends because that gives the recipient the ability to defer those taxes, much in the same way a buyback does. We’ll continue to evaluate [buybacks]. We have an authorization. As Sean said, we have excess liquidity on the balance sheet, and we have access to incremental capital. So as markets are volatile, we will be observing that volatility and trying to monetize it.”

- Dave Schaeffer, 2022

“And the important reason for our success is we ignore the millions of other buildings that we could serve, we’ve concluded that we can’t generate a high return on capital when we factor in our sales and marketing costs.”

- Dave Schaeffer, 2022

“There’s no deal worse than a bad deal. And we’re not going to do a bad deal.”

- Dave Schaeffer, 2023

“We will continue to look at assets. I do not think we are doing another deal; there is nothing else in the hopper. But if things come along, we will look at them. If they are not accretive, we walk away.”

- Dave Schaeffer, 2023

“We have refused to deploy capital at low single-digit returns where we would be subject to the monopsony power of that single tenant in that location. So, we will not go out and build to a specific sole tenanted facility unless there is sufficient de-risking in the form of upfront payments, higher returns and also long-term contracts.”

- Dave Schaeffer, 2024

“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.”

- Recurve Capital, 2024

“Dave Schaeffer is famously parsimonious. Despite being a billionaire, he flies Economy class everywhere he goes around the world when traveling for Cogent.”

- Recurve Capital, 2024

“[Dave] runs a very tight ship. If you visit the offices, they are Spartan. They drink water out of Styrofoam cups.”

- Recurve Capital, 2025

“Finally, engineers often forget that you can build the very best network in the world, but without customers, it is a science project rather than a business. It is neither self-sufficient nor self-funding. You therefore need a compelling value proposition that lowers the cost of customer acquisition.”

- Dave Schaeffer, 2025

“You can never fall in love with your business. You can be passionate about it, but you cannot be emotionally attached to it. It is a vehicle to create value for its stakeholders.”

Dave Schaeffer, 2026

“A good business has two attributes, no matter what the industry is: it grows organically and it produces cash at the same time.

Amazingly, very few public companies actually check both of those boxes. And I am not just talking about telecom; I am talking about all industries.

So we start with the fact that we have a good business. We are self-sufficient. We are producing cash, growing that cash flow, and growing the business. So there is no need to do anything.

I would actually argue there are four paths forward.

One is to stay independent and keep doing what we are doing.

The second would be to broaden the product set. That would, to use the word of one of our investors, Peter Lynch, diversify our business and take a good business and make it a bad business. If there is one thing I hope I have learned in life, it is that I do not want to work harder to make less money. My goal in life is always to work less and make more.

The third option would be for us to go out and acquire other businesses. That does not make a lot of sense either, given that we do not have a core competency in those other areas, and many of those businesses have challenges that may be insurmountable.

And then the final option is: do we sell? As a public company, that is something we always have to consider. Listen, I own 10% of the company, but my partners, the public shareholders, own 90% of it. So I am always going to listen to any offer and respond appropriately.”

- Dave Schaeffer

“For 18 years, we did not do a single acquisition. We had looked at 850 targets and passed. We chose to do Sprint because it was additive, and it was priced correctly, where we felt that the challenge that T-Mobile had gave us the opportunity to de-risk the acquisition with the $700 million payment stream.”

- Dave Schaeffer, 2026

In a nod to Dave’s savvy deal making ability, it’s worth visiting how he brought Cogent public. Early in Cogent’s life, optical transport vendors were desperate for new business. For each $1 of Cisco gear Cogent bought, Cisco would lend it $1.40.

To go public Cogent went through a reverse merger with Allied Riser, a failed fiber company with $140 million of cash on the balance sheet. After the merger, Allied Riser’s bonds became junior to Cogent’s $400 million in vendor financing debt and plunged in value. Cogent then bought back the debt, which had a face value of $140 million, for only $9 million plus some preferred stock.

When the dust settled, Dave had traded 13% of Cogent for $132 million in net cash and all of Allied Riser’s fiber network. To top it all off, Dave sued Allied Riser’s board for having agreed to merge with Cogent citing a failed duty of loyalty owed to the bondholders. Allied Riser’s insurers settled, paying Cogent $5 million from its directors’ and officers’ insurance policy.

In Summary

From 2004 to 2022, this playbook saw Cogent’s revenue compound at 10.3% while margins expanded by 220bps annually. Thanks to flat capital intensity, FCF grew even faster, north of 20% annually. Utilizing growth in EBITDA to gradually increase borrowing capacity, Cogent paid out over 100% of FCF through buybacks and dividends.

Over that same period Dave and the management team would evaluate around 850 potential acquisitions. All of which would fail to meet their hurdle rate except for one: Sprint Wireline.

The Sprint Wireline Acquisition

In 2023 Cogent acquired T-Mobile’s Sprint Wireline business for the grand total of one dollar. Sprint traces it’s roots to the Brown Telephone Company, founded in 1899. They were the first to build a nationwide fiber backbone and invested $20.5 billion in the 80’s and 90’s building out their 20,000+ route mile network designed to carry long-distance voice. MCI WorldCom tried to buy this business in 1999 for $129 billion but the deal was ultimately blocked by the DOJ.

After acquiring Sprint in 2020, T-Mobile migrated all traffic off of Sprint’s Wireline assets and onto the core T-Mobile network. This left T-Mobile with a nearly empty physical network burdened by a large fixed cost base. During the merger process they were also under heavy regulatory scrutiny. Rather than having to deal with layoffs, employee relations, and customer service at the same time, T-Mobile decided to offload this money-losing business to Cogent where it could be better utilized and restructured more rapidly.

At closing Cogent received the following key pieces:

• $700m of cash payments from T-Mobile, to be paid out over 4.5 years.

• 9.9 million IPv4 addresses.

• 482 technical facilities owned outright.

• 19,000 route-miles of owned inter-city, long haul fiber and 1,300 route-miles of owned intra-city metropolitan fiber.

• 16,800 route miles of leased dark fiber.

• An operating business doing $450 million of run-rate revenue and losing ~$180 million of EBITDA.

“The [Sprint Wireline] deal was really two deals. Deal number one was taking over their services business for enterprise customers. That business was generating $565 million of revenue, but negative $300 million of EBITDA, with $30 million of capex. It was declining at 12% per year, had a number of very bad contracts, and 93% of its revenue was off-net.

I was paid $700 million to take that business and run it off. They agreed to spend $500 million pre-closing to implement a number of changes. They also gave us a $100 million indemnification and another $25 million for severance. From their perspective, they got speed and certainty, and it cost them $1.3 billion instead of $1.5 billion. So it was a win from [T-Mobile’s] perspective: they de-risked it.

Then they sold us the network for one dollar. The network included 19,000 route miles of intercity dark fiber, all directly buried under railroad track; 1,200 route miles of metro fiber; and 482 fee-simple-owned pieces of real estate, comprising 1.9 million square feet and 230 megawatts of power. So I got that pile of assets for a buck, and we have been busily focused on repurposing them.”

- Dave Shaeffer

By the time the deal closed, T-Mobile had instituted a number of initiatives that brought revenue down to $450 million, and the burn rate from $300 million of negative EBITDA to $180 million of negative EBITDA.

“Yeah, so it is a highly unusual set of transactions when the seller pays the buyer $700 million in cash, indemnifies them, and then also pays employee severances above that, coupled with selling you a $20 billion asset for a dollar. But this was a business that was burning $1 million a day.”

- Dave Shaeffer

Also around the time of closing, Dave and the management team made the following announcements:

• They would enter the North American market for wavelength sales with a ~$2 billion TAM expected to grow at 7% per year over the medium term.

• They would look to sell excess dark fiber nationally and regionally along unique routes and rights-of-way.

• An aggregate revenue target of over $1.5 billion by 2028.

• After an initial decline, EBITDA margins of the combined enterprise in the low to mid 30’s could be expected within five years with the opportunity to subsequently expand EBITDA margins over time to approximately 45%.

• Cogent would not issue any new debt nor issue equity to fund the deal.

• Net leverage ratio of the combined company was expected to decline to approximately 3x EBITDA in five years time.

• Improved aggregate cash flow was anticipated to support their 43 quarter track record of uninterrupted dividend growth.

• Annual run-rate network cost synergies of $215-225m by year 3 after closing.

While some of these projections are on track, the majority are not with most of the original financial benefits now in doubt or pushed to the right. Below I go through all of Dave’s projections and compare them against where we are today.

As you can see from the above Dave has been wayyy off on his targets. There is some hope amongst the disappointment though. In general they have delivered on the things within their control. Cost reductions, data center conversions, and the merger in general all went well. This was no small task either, they took out $220 million of costs on a $490 million revenue business within 2 years. The Sprint facilities hadn’t been used in over a decade. They had to rip out 22,500 3-foot wide by 8-foot tall cabinets full of old telephone switches before upgrading the facilities and connecting them to metro fiber networks. They planned to convert these facilities over three or four-year years but accelerated that down to one year in response to all of the demand from AI. At the same time they migrated all of Sprint’s systems onto their own. Customer care, billing, support, field services, network monitoring, etc. They had that done within 6 months of closing. To sell wavelengths they had to reconfigure their network and deploy new equipment to data centers. They had 800 locations online in the first year and a half. I don’t want to downplay the guidance misses, but operationally wise the acquisition went about as well as one can ask for.

On the demand side of the equation, where things have been more outside of Cogent’s control, it’s been a mess. Dave’s projections for Wavelengths, legacy Sprint, data centers, and net leverage have all been complete flops.

Dave made it clear from the get go that net leverage would come down via EBITDA going up and not by any reduction in gross debt. The net leverage miss has been a combination of 1) increased capex into data centers (this was deliberate so that excess facilities could be monetized), 2) acquired Sprint revenues never stabilized prolonging breakeven and pushing any EBITDA contribution to the right, and 3) the main culprit, Wavelengths ramping far slower than expected.

While the absolute growth of Wavelengths has been impressive, relative to Dave’s projections it’s been very disappointing. This is the most important piece of the Sprint Wireline story and I will get into it further on in the writeup.

On data centers, Dave sounded confident that they would get $10 million per megawatt of power right up until they announced an LOI at ~$8 million per megawatt for two of their largest facilities. That deal fell through and 6 months later they announced a now closed deal for half their portfolio at $4.2 million per megawatt. This one is pretty hard to excuse, especially as demand for data centers has only gone up over this period. If these quirky converted telephone switch facilities had no precedent transactions to go off of and the value in them was really that uncertain, why would you give guidance at all?

Looking Forward

Today’s market valuation implies the Sprint Wireline acquisition destroyed a significant amount of shareholder value. I disagree.

For this investment to work there are two things that have to happen. 1) Cogent has to de-lever, and 2) they need to return to growth. In determining the likelihood of these two things coming true, there are five areas to pay attention to: The old Cogent (Corporate + Net-centric), Sprint’s Wireline’s legacy business, Data center sales, IPv4 addresses, and Wavelengths.

On point one, Dave has laid out a clear path to de-levering through data center sales and Wavelengths growth. He has been so far off in his communications on these two items that the market is not believing anything he says until it shows up in the numbers. I’ll touch more on this in my downside discussion below, but in short, net leverage has peaked and should come down to their 4x target within roughly 4 years time.

On point two, Cogent’s reported numbers show that their business has been in decline since mid-2023. In reality, the core business is fine. It is instead the acquired legacy revenues from Sprint Wireline that have distorted the numbers. Cogent migrated all of the Sprint business onto their own systems. As a result, the declining low-margin legacy Sprint revenues were in part allocated to a new reporting line called enterprise with the remainder being allocated across Cogent’s already existing segments. Uncertainty creates fear, and the consensus is not willing to look dumb on what looks like an over levered business in decline. If you dig into the numbers, it’s clear that Cogent’s group wide revenues are on the cusp of inflecting back to positive growth and that Cogent’s leverage problem is only a temporary one.

The Old Cogent

From 2012 to 2022, Cogent consistently grew their top and bottom line while returning all FCF as dividends. During this time they traded for an average dividend yield of around 4% - 5% (~2% above the 10-year treasury yield) with a peak of ~8%. Immediately prior to the Sprint Acquisition, Cogent was paying a run rate annual dividend of $3.70/share. Underneath all the noise from the Sprint Wireline acquisition, I argue that at the core cogent business, it is still business as usual. In this case the old Cogent is worth a conservative ~$54/share.

I’ve already touched on Cogent’s Corporate and Net-Centric businesses so I will keep this brief and focus on the numbers.

The Numbers

The Sprint Wireline deal closed on May 1, 2023. In their Q2 2023 filings Cogent discloses the amount of on-net and off-net Sprint revenues from May 1 to June 30 as well as the amount of Sprint monthly recurring revenue at close classified as Corporate, Net-centric, and Enterprise.

Using these numbers we can work backwards to get adjusted Q2 2023 financials that includes Sprint revenues for the full quarter. After this they stopped disclosing the Sprint acquired revenues making it very hard to get a read on the core Cogent business. This changed on the Q4 2025 earnings call, and continued in the Q1 2026 call. For the first time since Q2 2023, management disclosed the acquired Sprint revenue included within the Corporate, Net-centric, and Enterprise line items. These were important disclosures and yet the market didn’t seem to really care.

Stripping out the acquired Sprint revenues, Cogent’s core businesses of Corporate and Net-centric are up 9% and 44% respectively. This is complicated by the fact that included in Corporate and Net-centric are IPv4 and Wavelengths revenues. Adjusting for this I estimate that Corporate and Net-centric are up 9% and 28% respectively. Underneath all the noise, the classic Cogent playbook is still in action.

Corporate

Cogent is present in around 16% of all multi-tenant office space, and another 45% is reachable off-net. They win approximately 40% of written proposals. Today they have about a 14% market share within their corporate footprint.

Today they are present in 1,875 office buildings. This number has slowly expanded over time, however the majority of growth will come from further market shares gains. Although the law of large numbers means they’re growth rate will naturally slow over time.

“We probably have another decade of [Corporate on-net] growth ahead of us before we start to bump up against our [addressable] TAM. And our margins will naturally grow as our scale grows.”

- Dave Shaeffer

Corporate revenues grew pretty steadily at around 11% per year prior to COVID. They only started to recover in late 2023 and at a slower rate, however this should improve as office occupancy rates rise.

“Across our national footprint, the vacancy rate in those buildings is at around 15%. That is down from the just over 17 and change that it peaked at, at the worst of the pandemic. There is a modest improvement, but definitely not back down to the 6% vacancy rate in our 1,868 MTOBs that we were at pre-pandemic.”

Dave Shaeffer, Aug 24

Longer term, Cogent can also push harder into the 45% of office space that is reachable by off-net. For example if Cogent is currently serving a large on-net customer, they could sell off-net services to that customers regional offices.

Net-centric

The end market for Cogent’s Net-centric business has historically stayed flat at a TAM of around $1.5 billion. While the amount of data sent over the internet continues to explode, it is offset by a rapid and continuous reduction in the cost to send that data, resulting in a flat TAM over time.

“The 23% long-term decline in price per megabit is probably moderating a little bit, maybe reducing down to a 20% rate of decline. But the price of transit is going to continue to come down for 2 reasons. The market is sufficiently competitive. We are not a monopolist, I wish we were, but we’re not. And then, secondly, that the underlying technologies to produce routed bit miles continue to improve at pretty consistent rates.”

- Dave Shaeffer

In dollar terms Cogent has about 14% of the market, but in volume it’s about 25%. That’s because Cogent regularly undercuts their competitors by ~50% which has and will continue to be the driver of market share gains over time.

In contrast to their Corporate business, Net-centric was a beneficiary of COVID as video streaming exploded.

Next I’ll turn to the new Cogent, which addresses the core Sprint Wireline pieces of: the legacy business, Data centers, IPv4 addresses, and Wavelengths.

Sprint’s Wireline’s Legacy Business

Sprint Wirelines legacy business was and still is garbage. It came to Cogent as a bloated and undisciplined organization burdened by low and no margin revenue. Cogent immediately got to work shredding costs and unprofitable revenue. Dave initially thought these acquired Sprint revenues would stabilize around $400 million off of which he expected to be able to produce $80 million in EBITDA. Instead the business has declined all the way down to a $156 million run rate at breakeven EBITDA.

The legacy enterprise business did not use Sprint’s network, it ran off-net. Not included in the above breakeven EBITDA is the fixed annual cash costs to run the empty Sprint network. These cash costs are about $140 million a year (this was poorly communicated by Dave until recently). On top of this there is about a $30 million of annual capex needed to maintain the network. That brings the annual cash costs of the acquired Sprint assets to $170 million. Included in that figure are costs related to the 24 data centers Cogent is in the midst of divesting. Once this is complete, go forward Sprint cash costs should drop to around $150 million.

The revenue opportunities made possible by the Sprint acquisition include Wavelengths and leasing proceeds from the 9.9 million IPV4 addresses they acquired. Cogent’s ARPU on those addresses is currently $0.44/month and by year end they should have about half the broader portfolio leased. IPv4 revenue comes in at essentially 100% EBITDA margins getting us to a theoretical ~$26 million of Sprint attributable IPv4 EBITDA. That leaves us with just under $125 million of negative EBITDA to make up for. Wavelengths revenue comes in at ~95% incremental EBITDA margins meaning Cogent needs around $140 million of Wavelengths revenue for the acquired Sprint business to be running at break even. As of Q1 2026, Wavelengths was doing $54 million in run rate revenues. Over the last year Cogent has added an average of $1.6M in Wavelengths revenue per quarter. If that continues they would hit a $140 million run rate by year end 2029. Included in Sprint’s network are dark fiber assets, however Cogent has not started to monetize these so I simply ignore them for now.

With a line of sight to legacy Sprint as a whole breaking even, let’s quickly evaluate how much cash the acquired Sprint business has consumed since closing. I’m going to keep this simple. Prior to Sprint, Cogent paid out all FCF through a $0.925/share quarterly dividend that was increasing by $0.01 a quarter. In theory, from the end of Q1 2023 to the end of Q1 2026, this would have produced a total of $564 million in dividends with net debt increasing by about $250 million (excl. lease liabilities and keeping a constant net leverage ratio). However in reality, $477 million of dividends were paid out, $25 million worth of shares repurchased, and net debt (excl. lease liabilities) increased by $836 million. That $648 million delta ($62 million less return of capital and $586 million more net debt than expected) is what I consider to be the total Sprint cash costs to date. Is that number precisely right? No, but it should be close enough. If you follow Dave’s disclosures the total cash costs to date have been closer to $430 million. I prefer to keep things conservative so will stick with my $648 million number.

On June 29, 2026 Cogent closed a $225 million data center deal (these came from Sprint and the cost to convert them is included in my total Sprint cash costs above) and we can expect at least that amount for the remaining half of the portfolio. That will take our total Sprint cash burn down to $243 million. There is also still $195 million of cash left to be collected from T-Mobile.

If I assume that the $648 million of total cash costs was distributed evenly over the three years from closing until now and the $450 million of data center proceeds are received this year. Then for Cogent to make a 8% IRR on the Sprint acquisition Wavelengths only needs to scale to $150 million of run rate revenue by the end of 2029 and then hold steady from there on out.

Admittedly, this includes way too much monkey business. If my Sprint cash costs calculation above was questionable then this IRR math is even more so. Again though, I am simply trying to get this directionally right. I’ve erred on the conservative side in all of the numbers above. The point I’m trying to make is that if Cogent only scales Wavelengths to a $150 million business, a far cry from the $500 million Dave expects, then the Sprint acquisition at worst has created no shareholder value. This is in stark contrast to the large negative number implied by today’s valuation.

IPv4 Addresses

While Cogent has long owned a large portfolio of IPv4 addresses, they only really began to monetize them in 2022, hence why I have included them under the new Cogent.

The protocols that define what the internet is today require a unique numbering scheme. When the internet was initially designed, it was decided that there would be 2 to the 32nd power unique six-decimal addresses. The U.S. government initially owned all of these and wound up keeping approximately 800 million for their own needs while the remaining 3.5 billion addresses were made available to support the public internet.

Addresses were readily available for free until around 2011 at which point they began to be rationed. In 2015, Cogent began leasing out the addresses they had collected over the years, but only to customers that purchased bandwidth from them. In mid-2022, Cogent relaxed that restriction and has been rapidly growing that business since.

Cogent was initially interested in selling it’s unleased IPv4 portfolio. Prices peaked at ~$60 per address in 2022, but have since fallen to around $30 per address as the two largest buyers, Amazon and Microsoft, are no longer active in the market. Because of this they instead decided that leasing out the remaining portfolio would create the most net present value.

Today cogent has 15.2 out of their 38 million IPv4 addresses leased out, 1.5 million of which are with legacy customers for free. The run rate revenue from this business is $72 million and it comes in at essentially a 100% margin. Since Q1 of 2023 this business has grown at a lumpy 29% per year.

Despite increasing IPv4 prices by 30% over the last three years, Cogent still has a lot of room left to go before getting anywhere close to the $2.92 and $3.65 that Microsoft and Amazon charge.

To give you some idea of what these addresses are potentially worth. Cogent was able to issue two tranches of ABS debt totaling $380 million at an average fixed interest rate of 7.3%. This is backed by just 13 of the 38 million portfolio. In 2022 Ernst and Young appraised the 9.9 million IPv4 addresses acquired from Sprint at $458 million. Prior to selling Sprint Wireline to Cogent, T-Mobile sold ~2 million of Sprint’s addresses for $121 million. The above transactions land between $29/address for the ABS debt and $60 per address for the T-Mobile sale.

“As we looked in our aggregate balance sheet, we thought that it would make sense for us to reach out to ABS investors as a new group of investors and realized that our V4 revenues were an optimal candidate, widely diffused customer base, 8,000 customers, 12,000 unique agreements, a very sticky customer base. The churn rate and IPV4 leasing is [ 0.8% of 1% ] annually, almost 15 times better than the churn rate in our bandwidth business. And it is an extremely high-margin business with no real operating cost. So it was an ideal candidate for this market.”

- Dave Shaeffer

It's important to highlight that IPv4 prices are down sharply since their high in 2022, and hence the decision by Cogent to lease rather than sell for the time being. If prices recover then selling addresses could become an attractive option again. The 9.9 million addresses acquired through the Sprint acquisition are carried at a cost of $458 million vs $0 for the remaining portfolio making them a likely candidate for divesture from a a tax perspective.

There is a risk to this business though, which is IPv6. IPv6 is a newer version of the internet protocol that increases the number of unique address from 2^32 to 2^128. With 79 octillion times more addresses than IPv4, prices would drop to zero. However, in the short term IPv4 is protected by switching costs, and in the medium to longer term IPv4 is protected by pretty strong network effects.

“There are three challenges to IPv6 being widely adopted. The standard came into play in 1998. In 27 years, it has only gained 7% market share. The limitations are all equipment needs to be IPv6 compatible and can for the path of the packet travel. That is becoming more common. It is still not the case today, but probably over the next five years, I think that limitation will go away. The second limitation is probably the most important. That is when you use the internet, you want to reach all of the endpoints. If an endpoint is only visible on IPv4, IPv6 cannot see it. As a result, if you elect to use IPv6, you today only get 7% of the public internet, and 93% of it remains dark to you.”

- Dave Shaeffer

“The Federal Government in 2010 put a mandate out for all agencies to be entirely on V6 within 18 months. Today, they’re less than 2% converted. This has a very long tail. It is a very fine resource and the expense of renumbering is not trivial. So even if you had V6 and it would work, the cost of leasing addresses is so diminished versus the benefit it gives. Most companies will take a very long time to renumber. And no one wants a partial view of the Internet.”

Given that choice, almost everyone says, ‘I’ll pay an extra $0.50 a month to be able to see the whole internet.’ The third issue, which is the ultimate process of transition, is very problematic because it has to happen all at once. There is a fair amount of labor involved in going to each device and changing the number. It is kind of like the Y2K problem on steroids, where every device needs to be touched and upgraded to V6, but it all has to happen instantaneously. Because of the complexity, ubiquity, and importance of the internet, I do not think that is practical today. It is probably another couple of decades before we see widespread ubiquitous use of V6.”

- Dave Shaeffer

Data Centers

Included in the Sprint acquisition were 482 owned technical facilities comprising 1,900,000 square feet and 230 megawatts of power. These facilities were fallow and full of old telephone switch equipment. The initial plan was to take 45 of those facilities (later increased to 125), put a 10,000 square foot data center in a portion of the building and leave the remaining space empty. This would allow Cogent to enter the Wavelengths business using the empty Sprint network and to get rid of some existing leased facilities. As the AI wave accelerated, it became clear that the power available in these facilities was a scarce asset and could be monetized. In 2024, Cogent made the decision to spend $100 million of incremental unplanned capex to convert 24 of the largest facilities into full data centers. These 24 facilities, comprising 109 MW of power could then be divested or leased out.

Because Cogent only needs a small portion of each of their data centers to run their business, the remaining space can be leased out to other operators. The remaining 101 Sprint converted data centers that Cogent will continue to operate offer additional colocation revenue opportunities. This was historically a small part of their business but should become more meaningful over time.

The 24 data centers Cogent is marketing for sale are not equivalent to modern purpose built facilities. They’re typically smaller facilities with lower power density and have no contracted revenues associated with them. On a positive note they’re distributed over a wide footprint which has some value and in some cases they have the transmission capacity and zoning to expand the already existing facility.

In coming up with an expected value of the converted data centers Dave somewhat arbitrarily applied a 40% discount to the average sale price of about $17 million a megawatt for newly built data centers. In hindsight he was way too confident in this go-to-market price of $10 million per megawatt. At the end of the day they only received $4.2 million per megawatt… ouch.

That $4.2 million per megawatt comes from their $225 million deal for 10 facilities that closed on June 29th of this year. The remaining 14 data centers totaling 56 MW look pretty similar to the 10 that have already been sold with one exception, the 14MW facility in Fort Worth, Texas. That data center is worth considerably more on a per MW basis than the rest of the portfolio, however that is offset by a couple of subscale facilities that are worth very little. A relatively safe assumption is that the remaining data centers will fetch ~$225 million as well. Cogent has enough NOL’s and bonus depreciation to offset capital gains from the first $225 million but there will be some tax leakage on the remaining sales.

Wavelengths

Wavelengths is the most important piece of the Cogent story going forward. It was one of the main reasons for acquiring Sprint and it’s the primary culprit for why shares have fallen so far.

To utilize Sprint’s empty network Cogent entered into a new market: Wavelengths. The cheapest way to transmit information is over the public internet. Cogent charges $0.07/Mbps per month to do so. The public internet runs on protocols that find the most efficient available route for a packet of data to get from A to B. While on its journey your data packet competes with everyone else’s, there is no guarantee as to how fast it arrives. It also passes over shared infrastructure, while data is separated and encrypted, it is not 100% secure. Wavelengths fixes these problems by providing a private point-to-point connection. For customers requiring predictable throughput, stable latency, and greater control overall, it is worth paying more for a Wavelengths connection. The business is called Wavelengths because customers are allocated their own optical wavelength on a strand of fiber. Customers include hyperscalers, telco providers, content companies, and government agencies. The market is dominated by Lumen and Zayo with an estimated TAM of ~$2.5 billion.

Just like Cogent’s existing network, their Wavelengths network is purpose built around a single product.

“We have built the sales processes and tools into our CRM system, so when a customer inquires about a wavelength between two North American data centers, there’s a drop-down menu of centers. The salesperson clicks on the two centers. It auto-generates the physical path through the network, a map with latency accurate to within one hundredth of one millisecond, a KMZ map or a detailed geographic map of the network with accuracy of one meter, and the pricing for that customer. No other vendor in the industry can provide this level of detail even with a 120-day installation.”

- Dave Shaeffer

“We could architect it from the ground up in a way that would be much more efficient for this product. Companies like Lumen and Zayo sell hundreds of products across their network. You know, I think they’ve recently announced they’re hoping to reduce their product portfolio to just 300 products. Cogent, prior to selling waves, literally had three products, now it has four.”

- Dave Shaeffer

“Our competitors sell many other products over a heterogeneous network, and therefore, they do not optimize their wavelength services. Wavelengths are not the primary revenue stream at either of the companies you mentioned, or any of the companies we compete with. So like with any multidimensional product, the chain is no stronger than the weakest link. The weakest link is the most difficult product you have to deliver.”

- Dave Shaeffer

Dave has his sights on capturing 25% of the North American Wavelengths market in short order. This is achievable for the following reasons:

• Speed: Cogent provisions Wavelengths within 30 days. This compares to competitors with average provisioning times of 90 to 120 days.

• Accuracy: Cogent’s network is homogeneous and architected from the ground up specifically for Wavelengths vs competitors who operate disparate, unintegrated networks. In a couple of clicks a salesperson can provide latency accurate to within one hundredth of a millisecond, a detailed geographic map of the network with an accuracy of one meter, and the exact pricing for that specific route.

• Breadth: Thanks to the breadth of Cogent’s Wavelengths network, they can offer far more unique point-to-point routes. Cogent offers Wavelengths in 1,107 carrier neutral data centers compared to ~400 at Lumen. Because any location can connect to any other location, Cogent can provide 612,921 unique routes compared to Lumens 79,800. With legacy providers, a customer often has to buy a metro wave from one vendor and an intercity wave from a different vendor. More often than not, Cogent can offer one seamless solution,

• Diversity: Cogent can offer redundancy to customers because 90% of its routes are physically unique to existing offerings. By adding a Cogent wavelength, a customer diversifies their network. A single backhoe can no longer take down an entire point-to-point connection.

• Reliability: Cogent’s Wavelengths run on fiber buried 6 feet under the middle of railroad track, protected within an armored jacket. This compares to conduit networks buried 2 feet deep in plastic pipe, along public highways. As a result, Cogent has 50% fewer splices per mile than competitors.

• Price: Cogent acquired their network for $1, while competitors have spent billions building theirs. They can afford to undercut competitors and use price to gain market share.

After acquiring Sprint Dave guided to capturing 25% of the then $2 billion dollar TAM by mid-2028. Despite reiterating this guidance as recently as Dec, 2025, it looks very unlikely that Cogent will hit this guidance. 80% of the Wavelengths market is bought by the top 200 buyers, 150 of which Cogent already does business with. So with Cogent’s access to customers and their clear advantages over competitors, why have they been so slow to ramp?

Wavelengths has gone from standstill to a run rate of $54 million in 3 years. Last quarter they grew 50% YoY. This is very impressive on an absolute basis but relative to guidance it has been painfully slow.

The slow ramp can be attributed to three things, all of which are temporary: 1) Streamlining their provisioning process took longer than anticipated, 2) Cogent had to prove themselves and their unique network design, and 3) Customers have delayed turning on their provisioned Wavelengths.

1) When entering the Wavelengths business Cogent started with a blank piece of paper and were able to design a network that required only two field visits per installation with no pre or post engineering work. Lumen and Zayo on the other hand typically require six or more field visits and custom engineering to bring a connection online.

However, before they could start provisioning in this more automated manner, Cogent first had to reconfigure the Sprint network. This was a huge undertaking with nearly 10,000 discrete work projects involving 1,000 of Cogent’s 1,900 employees. While Cogent could still provision Wavelengths on a custom basis, the same way competitors do, the focus was on getting the network reconfigured. Dave originally thought that once the top 50 to 100 data centers were reconfigured that would be enough to cover the majority of customers’ required routes. This proved incorrect and Cogent had to finish reconfiguring the entire network before shifting into high gear. This meant their acceleration in Wavelengths provisioning was delayed by about a year.

“We had hoped to generate waves out of a much smaller footprint, and realized that we needed the whole footprint.”

- Dave Shaeffer

2) The Sprint network that Cogent runs their Wavelengths on was built 40+ years ago. It was easy for competitors to paint a picture that this was an old and unreliable network. This is untrue and after 3 years running Wavelengths Cogent now has the performance data to prove it.

“…The third attribute [of Sprint’s network] that is different is the type of fiber, so the fiber that was deployed was SMF-28. It was first generation fiber, and it was thought in the late 1990s that that fiber would be obsolete, as everybody needed to move to higher transmission speeds and that, in fact, would have been correct if all of the transmission technology remained non-coherent.

The movement from half a gig to 40 gigs per wave, and the move from one wavelength to roughly 80 wavelengths, was all done with non-coherent optics. At the time, it was believed that you had to switch to a different type of fiber, non-zero dispersion shifted fiber. The most prevalent of those was Corning E-LEAF. The second most prevalent of those was Lucent TrueWave. Those were the two brands that were most widely deployed in North America, and that fiber turned out to actually be not very good for coherent transmission technology. When the marketplace demanded 100 gigs per wavelength, all of the equipment vendors globally switched their optics from non-coherent to coherent. When that happened, the non-zero dispersion shifted fiber actually performed worse than the original SMF-28…

Now, it is true that the brand-new SMF-28 that’s bought today, you can go still buy that at Corning, is slightly different. It has better optical purity and slightly lower loss, so you’ll gain about one-tenth of a dB per kilometer on the newer fiber versus the older. But the number one reason why the fiber underperforms is actually splice cuts. So each time the fiber is cut and spliced back together, there is loss in that fiber.

We, as a buyer of fiber from Lumen, can absolutely attest that on a per mile basis, there are over twice as many splices on the Lumen fiber today than there are on the Sprint fiber, just because of how it was physically deployed. I do think there is a need for SMF-28. I also believe that AI will generate more demand, but I also think there’s an arithmetic problem, and that is how many fibers you actually need to do what you want to do. We are the largest carrier of internet traffic in the world. We carry a quarter of the world’s traffic. We do that on one pair of fibers. We are utilizing roughly 30% of the lit capacity in that fiber and have the capability of increasing that lit capacity by at least a factor of ten.”

“While our wavelength base is still small, about 1% of the North American market [as of Aug, 2025], it is now at least statistically significant enough that we can measure our performance quality as measured by a number of outages along each route. Today, and this is anecdotal, we are running at about 7x fewer outages per span than at least one of our major competitors as reported to us by customers who have similar city pairs along different fiber with that other vendor and with Cogent.”

- Dave Shaeffer

As mentioned in point one, Cogent can provision Wavelengths anywhere in the country within 30 days. This was unheard of in a market accustomed to multi-month delays and horrible service. Naturally, customers were skeptical of Cogent’s installation times.

“You hear horror stories about Zayo. Someone asks for service in a data center that Zayo is in, and Zayo says, ‘We do not have service there.’ The customer is FaceTiming with the Zayo sales rep, saying, ‘I am looking at your rack. You have service here.

…That is the market customers are used to: horrible service, not turnkey service. So when Cogent goes from 18 months of ‘we cannot install you and cannot promise you anything’ to flipping a switch and saying, ‘We can install you within 30 days,’ that is all well and good. But number one, nobody believes you. And number two, nobody transacts in that market on that timeline, because everyone is accustomed to significant delays.”

- Recurve Capital

Currently Cogent is only being offered a portion of their customers backlogs to bid on. As Cogent continues to prove out their network and provisioning speeds they will see an increased portion of those customer opportunities.

3) As explained in point two, customers have been caught off guard after receiving Cogent’s wavelengths on or ahead of schedule. Unlike the Net-centric business where customers can easily plug in and configure a new connection, Wavelengths is a bit more involved, requiring more planning and equipment setup.

Customers are accustomed to a slow and error prone provisioning process. Cogent has surprised the market with how quickly they can deliver and customers have been left with installed wavelengths before they’re ready to turn them on. Cogent only starts billing a customer once they have accepted the wavelength. With Cogent’s Net-centric business, they implement forced billing after a customer delays delivery more than two times, however they have not yet implemented this with Wavelengths.

“We wanted to be careful not to alienate significant customers by being too aggressive and pressuring them to accept wavelengths. We have several large customers that have been truly shocked by our ability to provision in the windows that we have outlined. And as a result, they were not prepared to take the wavelength services. They typically order their cross connects, order their pluggable optics and accept them in a 3- to 4-month window from placing new order. But in Cogent’s case, we had one very large wave order that we were actually able to provision nearly 100 waves in 7 days. I mean they were truly amazed at that, but they came back and said we can’t take them… I think there will be a point in time when we can implement a forced billing discipline, but we are unwilling to do that at this point.”

- Dave Shaeffer

Customers have also pushed out their acceptance of wavelengths due to supply chain delays. The industry as a whole has run into issues with the availability of power, optical equipment, GPU’s, and technicians. As of now there remains a meaningful backlog of installed connections that haven’t yet shown up as revenue. With time these constraints will abate and customers will adapt to Cogent’s rapid provisioning times.

All of the signs point to Cogent winning meaningful market share from competitors. They have a superior offering and are willing to be aggressive on price. Switching costs are relatively low and contracts average 3 years in length. As Cogent continues to prove themselves they’ll receive a greater share of customers bid opportunities and as customer equipment bottlenecks ease the number of billed connections will catch up with the number of installed connections.

Putting It All Together

As I hope to have proved above, the Sprint acquisition has not destroyed shareholder value. The market is wrong. If wavelengths can scale to ~$150 million in annual revenue then the acquisition will have ended up creating a negligible amount of value. Everything above that is very accretive to shareholders.

In addition, Cogent’s core business has continued growing throughout the integration process and has a number of tailwinds going for it. Net-centric is a major beneficiary of AI and the continued shift towards streaming. While their Corporate segment’s growth has been more muted, the average customer in that business is using 8 times more bandwidth than pre-COVID with 40% less employee days in the office.

The Cogent of today should be worth more than before the Sprint acquisition, not less.

Forced Selling and Incentives

Before getting into my valuation, a quick discussion on the share price and Dave’s equity ownership is warranted.

Shares are down 86% from their peak in January of 2025. The sell off started in February after the first of what would turn into multiple Wavelengths and legacy Sprint misses. This continued through Spring as short sellers sniffed out Dave’s looming margin call which came to fruition in August when he was forced out of his entire equity position. He’s been very transparent on what happened:

“So I as an individual, have received my compensation from Cogent almost exclusively in stock for 25 years. I paid taxes on that stock as that stock vested.”

“As the stock appreciated, it got more expensive. I went to the Board and said, “I need to have the ability to pledge those shares to pay taxes.” I have a tax basis as of [Jan 1, 2025] of $155 million that I had paid taxes on, so almost $80 million in card tax that I paid. That was the sum dollar value of my pledged position. I had to contribute stock to be able to do that [2.7 million shares]. I kept a pool of stock unpledged and a pool pledged. As the stock came down, I went to the Board and said, “I would like to, rather than sell, put more stock into the pledge pool.””

“I was fortunate enough to have income from other sources, primarily my real estate portfolio. As the D.C. real estate market deteriorated, I had additional pressure to reduce leverage on my real estate portfolio, which forced me to begin selling Cogent stock.”

“I did not anticipate the run that we had and effectively got wiped out of my position. I’m a big boy. I get that. That was the bargain I made. The confusion around my real estate. I have never pledged a share of stock for real estate. I have a portfolio outside of Cogent of 42 office buildings. I had office buildings before I had Cogent, and I built that business up, and it has been under tremendous pressure. Over the past 20 months [Dec, 23 - Aug, 25], I injected $152,340,000 of actual cash into that portfolio to negate the decline in value. The portfolio at pre-pandemic levels was 53% LTV. Today, it’s probably above 90% LTV. While I’ve reduced the debt load from $578 million to $422 million, the face value of the assets has fallen from $1.1 billion to below $500 million.”

“I’ve tried to be extremely transparent with investors, probably more than most people in my situation would be. And I am committed to making sure that as an individual, not as Cogent, my lenders are made whole even though many of my brethren in my industry have walked away from their assets.”

“The only stock that I have left are the shares that are held by the company that are restricted. Those restrictions will lift. I will be able to use that stock, but I have no longer any debt on it.

If there’s a silver lining in a very dark cloud, I now have a loss carry forward on my Cogent stock. As that stock vests, I won’t have to pay any taxes.“

- Dave Shaeffer

Earlier this year the board implemented new performance rewards for Dave. In addition to his annual compensation package, he was granted 3 tranches of RSU’s. These expire on Dec 31, 2029 and vest only if Cogent shares trade for a 60 day VWAP of $70, $85, and $100 per share.

I don’t love this compensation structure as it incentives home run/go big or go home thinking. On the other hand Dave is an owner at heart and is a shareholder friendly operator. All in all I am happy to bet on Dave and he obviously see’s $70+/share of value.

The Numbers

Before I get into my valuation. I’d like to spend some time on the downside. Cogent’s high debt load and upcoming 2027 maturity warrant careful consideration.

Debt Load and Downside

Cogent is 7.4x levered. There are some nuances to this number, but overall it is too high and the equity has been punished for that.

I prefer to use the net leverage ratio adjusted for current and future payments from T-Mobile of 10.3x. This includes amounts due from T-Mobile in net debt and excludes payments from T-Mobile in EBITDA. This number will come down fairly quickly thanks to data center sales and EBITDA growth. I estimate that by year end it will be sitting somewhere around 8.3x.

FCF and continued EBITDA expansion should see net leverage drop by another ~2x turns by end of 2028.

With the path to de-levering established we can address the near term liquidity issue. This is the $750 million of senior unsecured notes that come due on June 15, 2027. Cogent’s plan to address this maturity is:

1. Amend the 2032 Senior Secured notes to increase the allowed secured leverage ratio from 4.00x to 4.75x - this is now done.

2. Use the majority of the of $225 million in Data Center proceeds to repurchase the existing 2032 Senior Secured notes until they trade at par - the 2032’s currently trade at 90% of par.

3. Refinance the 2027 Senior Unsecured notes with new 2033 Senior Secured notes.

“Our current $600 million secured notes will mature in June of 2032. While anticipated $750 million of secured notes will mature in 2033, $206 million of our secured ABS IPv4 notes mature in May of 2029. $174,400,000 of our secured IPv4 notes will mature in April of 2030. Whereas $629 million of IRU finance leases or capital leases have various maturities extending through 2046.

But with our current $600 million secured debt trading at a discount to par, we want to try to capture some of that discontinuity and buy back the current $600 million secured until they trade closer to par. And then at that point, the additional capital that we have could be used to result in a smaller new issuance. But today, the current secured debt is trading at a material discount.

We are paying 7% on the current unsecured bonds. Our current secured bonds are trading at just around 8% today. I believe our new issue will most likely price off of the trading of those bonds and will be somewhat similar. We'll have both a new issue concession. It's typically about an 8 point, and it could have a small variance based on duration. If we sell the data centers use a portion of the proceeds to buy back bonds, it is likely that the current secured bonds will trade asymptotically to par, which is 6.5% and then it would allow us to finance probably at a similar rate.”

- Dave Schaeffer, May 26

In case the above refinancing and de-levering were to fail, lets have a look at how a liquidation scenario could play out. As per usual I’ve tried to keep the below analysis very conservative.

A competitor would likely view the above valuation as an absolute steal. Even if I assume all the cash gets eaten up during the liquidation process and then apply a 20% fire sale discount on my already heavily discounted asset values, net liquidation value still lands at $13.50/share.

Valuation

In coming up with a valuation I used two methods. One is the monkey brain method: I take the $0.925/share quarterly dividend that Cogent was paying prior to the Sprint acquisition. This is what I call core Cogent, since then core Cogent has grown revenues 17%, in theory taking our quarterly payable dividend up to $1.08/share. From 2014, which is when Cogent started paying out all of their FCF as dividends, up until the Sprint acquisition in 2023, they traded for an average dividend yield of 4% - 5% with lows in the 3%’s and highs in the 8%’s range. Future core Cogent will grow slower than past core Cogent as they penetrate more of their TAM, so I’ll slap an 8% dividend yield on core Cogent to get a monkey brain fair value of $54/share. I next assume that the Sprint acquisition has not destroyed shareholder value. In a worst case scenario I could see it destroying ~$25/share. That would be if the Sprint network is never monetized and Wavelengths loses money forever. On the flip side, if Wavelengths can get to Dave’s $500 million goal, and even if it takes 3 or 4 years longer than expected, this acquisition creates a ton of value.

While I prefer the monkey brain method, I gut checked those numbers with a simple DCF. I’ve kept my assumptions pretty conservative in the below analysis. Playing around with the numbers under a couple of different scenarios I get anywhere between $40 and $150 per share of fair value. My base case lands at $81/share.

At today’s price of $12.00/share, Cogent is hugely undervalued. The downside is thoroughly protected by a large, diverse, and valuable set of hard assets. I struggle to see how these would net less than $30/share even in a fire-sale liquidation process. My conservative fair value of Cogent sit’s at $54/share, this assumes that the Sprint acquisition was a net zero for shareholders. If wavelengths can scale to $150 million of run rate revenue by year end 2029, then I see fair value landing closer to $80/share. If Wavelengths can grow to 25% of the market within 10 years time, than I see fair value approaching $150/share.

Anyway you slice it, shares of Cogent are available for a bargain price. Net leverage will come down through asset sales and EBITDA growth. When net leverage falls back below 4.0x sometime within the next 4 or 5 years, Cogent will start returning all FCF to shareholders. This is the catalyst and will force a rerating either through a large buyback program (if shares remain undervalued) or as yield seeking investors are attracted back to the name.

The market is not willing to risk owning Cogent through any potential near term pain, waiting 4 or 5 years is an eternity for most market participants. I on the other hand am focused on the long-term. I am happy to wait and will take advantage of any near term volatility to add to my position.

Upside

There are a couple of things I have not included in my valuation that are likely to be of some value.

A) In Q1 2026, 83% of new sales were on-net. That compares to 62% of the current base and 76% of the core cogent base. In my model 75% of core cogent revenues stay are on-net. In reality that number will increase over time, further benefitting margins.

B) Cogent has a significant portfolio of excess dark fiber. The Sprint network has on average, 60 to 70 fibers running along each route mile. Cogent only needs ~6 of those fibers to run their internet and Wavelengths businesses. While there will be some stranded inventory, the majority of that excess fiber should be able to be monetized. Now that the Sprint network has been repurposed for Wavelengths, Cogent can turn some attention to selling this excess dark fiber. There is a lot of price variability and uncertainty depending on each specific dark fiber route. This makes is hard to nail down the potential value, so I have left it out of my analysis.

“…I can tell you what we paid for a pair nationwide, which was $54 million. That’s public record. It’s in our filings. Now, that was some time ago, but dark fiber prices have not really moved. I think we would typically sell dark fiber on a route-by-route basis, rather than on a nationwide footprint. That means you could still assemble a footprint. Different routes have different value, and to be candid, our inventory varies route by route. You know, the Sprint fiber is direct buried. That actually gave it an advantage in not having many cuts and therefore being high quality, but it also means that it is not very easy to add more fiber to that given route.”

“I think if you looked at current pricing and our inventory, it would be realistic to assume that there’s at least $1 billion of value that could be had from that dark fiber sale market.”

- Dave Shaeffer

Cogent pays $15-20m of annual maintenance fees to Lumen for rights to 12,500 route-miles of a single pair of fiber. Cogent acquired 19,000 route-miles of physical network from Sprint with an average of ~40 excess fiber pairs per route.

C) AI is major tailwind for internet traffic. The roughly 800 zettabytes of data that are stored around the world that have been collected over the internet is the raw material for AI training. That data is being collected at an accelerating rate. Historically, about 20% of data transmitted was ever stored while the other 80% was discarded. That ratio has now inverted, with over 80% of data transmitted is now being stored and used for AI training. As the largest carrier of internet traffic in the world, Cogent stands to benefit from the increased collection of that inbound data. In addition, the training of that data in large language models typically occurs in a different location than where the data sits. As such, there needs to be large bandwidth connectivity from those locations to the training location and then back to the original origin of the data. That flow of information is ideally suited to wavelengths and dark fiber.

Timeline and Catalysts

Open market 2032 secured debt repurchases, June 29 onwards: Cogent has committed to use at least 50% of the $225 million data center proceeds for buying back the 2032 notes. Dave’s plan is to buy in the open market until the debt trades up to par. Since June 29th the 2032’s have seen around $100M in total volume. Assuming Cogent has been 50% of that, they will have deployed 50% of the proceeds by around the end of Aug and 100% by around the end of Sept. FINRA caps trade size’s shown on the public tape at $1 million so if Cogent has been buying large negotiated blocks then this timeline would be brought forward. It’s worth keeping a close eye on the 2032 bond price.

Q2 Earnings Release, Aug 6: This is an important earnings release, keep an eye on Wavelengths customer connections. This stood at 2,263 at the end of Q1. Even more important will be Dave’s commentary on the 2032 repurchases and refinancing. Also worth monitoring is the Sprint Wireline revenue vs core Cogent revenue trends within Corporate and Net-centric.

2027 unsecured debt refinancing, H2 26’: Look for a refinancing announcement soon after the 2032 debt repurchases are complete. That should wrap up sometime this Fall, possibly earlier if the 2032’s trade up to par quickly.

Q3 Earnings Release, Early Nov: If group wide revenues aren’t positive quarter-over-quarter in Q2 then they should be by Q3. Again, Wavelengths customer connections is the most important KPI to watch.

Data Center Sales, H2 26’ - H1 27’: I’m expected the remaining 14 data centers to be sold sooner rather than later. I would consider anything over $225 million a win.

Dark Fiber Deals, 26’ - 27’: Pay attention to any commentary from Dave on Dark Fiber. It will be very interesting to see what kind of economics they get out of these assets.

IPv4 Sale, 27’+: If IPv4 pricing recovers, a large sale of unleased addresses would be a likely option. Especially the 9.9 million block acquired from Sprint with the higher tax cost basis.

Net Leverage <4x, 29’ - 30’: This is the big catalyst and is the direct result of asset sales + EBITDA growth (mainly from Wavelengths). The sooner they get under 4x, the sooner the return of capital can kick back into gear.

In Conclusion

Cogent at it’s core is a good company. Everyone needs the internet, and it’s importance is only increasing. Cogent is the lowest cost provider of said internet. They are counter positioned against competitors who rely on legacy revenue lines like voice and video. Cogent uses it’s network for one thing, and one thing only: Internet. That focus creates a structural cost advantage.

Even though the Sprint Wireline acquisition by most accounts has been a disaster, there is light at the end of the tunnel. Net leverage has peaked and unprofitable legacy revenues have just about run off. The transaction is on track to be a net zero for shareholders with ample opportunities for it to still turn into the big win it was initially underwritten to be.

I am buying shares for $12, that’s less than a quarter of my $54 fair value estimate. If the Wavelengths business accelerates, as I believe it will, then shares are worth $100+.

Investors will wake up to the opportunity as 1) the 2027 debt maturity is addressed, 2) the group returns to topline growth, and 3) net leverage falls towards 4x allowing the return of capital program to resume.

For those interested in learning more about Cogent, these are some of the more interesting and valuable sources I came across during my due diligence.

• Aaron Chan of Recurve Capital has written extensively about Cogent, insights. He also has two good interviews out with Dave, 2023 Interview, 2026 Interview.

• How Southern Pacific created the Sprint network: SP and Sprint, Oct 2024.

• Bandwidth Podcast: Dave Schaeffer, founder and CEO of Cogent Communications Part 1 and 2.

• Dave speaks frequently at various conferences. You can find those along with their quarterly call transcripts here.

• A 1994 Washington Post article on the Schaeffer Taxi Empire.

• AIDA’s IPv4 and IPv6 AS Core

• Cogent’s original Sprint Wireline acquisition deck.

• A 2008 Forbes article on Cogent, Telecom Knockout.

• An article from Phil Harvey, Cogent CEO Dave Schaeffer: The World on a Fiber.

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

Valesto Energy owns 6 higher quality jackup drilling rigs, all of which operate in SE Asia. There’s visibility into SE Asia upstream capex continuing to rise through 2028. Valesto has a net cash balance sheet is returning all excess cash flow to shareholders. Shares trade for 5x EBITDA or about 10x earnings with an 11.5% dividend yield.

• FY25 Investor Presentation (Link).

SED Energy Holdings 🇳🇴
NOL: ENH • EV: USD $565M • Last Close: NOK 7.76

SED Energy also owns O&G offshore drilling assets in SE Asia. They own 4 tender barges, 2 tender semi-subs, and 2 seismic vessels. Tender rigs support development drilling on fixed platforms while jackup rigs leans towards exploration and standalone drilling. The brownfield and development work that tender rigs are attached to is more stable than exploration drilling.

Today we are sitting somewhere between scenario 1 and 2 below.

Shares trade for 5x EBITDA or 11x earnings with a 17% dividend yield.

• FY25 Investor Presentation (Link).

Hikari Tsushin 🇯🇵
TYO: 9435 • EV: USD $10.1B • Last Close: ¥37,280

Hikari Tsushin has been described as the Berkshire Hathaway of Japan. They have a 30-year TSR of 9.7%, that’s double the Nikkei 225 over the same period. Book value per share has compounded at 19% over the last 15 years while operating profit has grown at 28%.

Hikari owns a portfolio of asset-light, recurring, cash generative business built around a nationwide agency network. They resell electricity and gas, insurance, telecommunications, water, and more. Hikari targets 10% annual operating earnings growth from this group over the long-term. This group requires very little incremental capital to grow. Hikari invests all excess cash flow into a growing portfolio of public and private securities. This portfolio has produced a pre-tax CAGR of 17% over the previous 9 years.

Hikari’s net investment portfolio is worth ¥25,670/share and the operating group is on track for ¥2,708 of FY27 net income. Included in that net income figure is ¥1,614/share of customer acquisition costs to which they underwrite 15% IRR’s. Backing that out, Hikari’s underlying operating group net income is ¥3,800/share.

• A Writeup from Longriver Partners Q3 2024 Letter (Link, pg. 2).

• FY25 Investor Presentation (Link).

Noodles & Company 🇺🇸
NASDAQ: NDLS • EV: USD $175M • Last Close: $11.92

Noodles & Co is a fast-casual chain with 386 total system restaurants across the US. They are in the midst of a so far successful turnaround. Unprofitable locations have been shut down and same stores sales have turned positive.

At the end of last year the company announced a strategic review and gave management retention bonuses that pay out if a change of control occurs by Dec, 2026. Shares currently trade for a forward EV/EBITDA of 5x while comparable transactions suggest a floor of ~6.5x. Every EBITDA turn above 5x translates to almost $6/share of additional equity value.

• A Pitch from Special Situations Investments (Link).

Check Out: The Decline of Deviance

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

Proficient Auto Logistics is a LTL freight transportation company focused on auto-hauling. They transport new cars from auto OEM’s to dealerships. The auto-hauling LTL market recovery has lagged the broader LTL recovery. This looks like it will soon change due to the bankruptcy of competitor Jack Cooper and other smaller players.

Today they’re trading for an EV/EBITDA of ~7.5x. On more normalized/mid-cycle numbers this would be 5.5x. LTL peers trade for around 12x to 18x.

• A Great Writeup from Everyone Hates Poetry (Link).

• A Writeup from Everyone Henrik Alex (Link).

Tab Gida 🇺🇸
IST: TABGD • EV: USD $1.2B • Last Close: ₺233.40

Tab Gida is the master franchisee for Burger King and Popeye’s in Turkey. They are twice the size of the nearest QSR competitor and control their own processing facilities, bakeries, and logistics.

“[The Kurdoğlu family] bring an engineering-minded mindset to fast food. I have never seen any fast-food team anywhere in the world that is as good as theirs. These guys are better than any franchisees I have ever interacted with. They backward integrated heavily. They are detail-oriented on all the finer points. They are looking at all the nuances carefully. That is the reason why they made mincemeat of McDonald's.”

- Mohnish Pabrai

Shares trade for ~12x earnings with a long runway for continued expansion of both current and additional brands.

• A Writeup from Lunch Investing (Link).

Ferrari Group 🇮🇹
AMS: FERGR • EV: USD $634M • Last Close: €7.95

Ferrari Group, different from Ferrari NV, is a family run luxury logistics company. They ensure that very high-end products like jewelry, diamonds and watches make it safely from companies warehouses to stores, between stores, and directly to customers. Ferrari’s services are integrated into customer operations through shared facilities and IT systems. Their security employees, armored vehicles, vaults, and procedures are all designed specifically for luxury items. Transport costs are very small compared to the value of customers products. It is not worth the risk of disappointing a client to save a tiny amount of money.

Ferrari has grown throughout the recent luxury downturn highlighting the durability of the business. Shares trade for an EV/FCF of 8x. They earn a 30% ROIC and management has targeted medium term 6% - 8% topline growth.

• A Recent VIC Writeup (Link).

• A Writeup from The Dutch Investors (Link).

Check Out: Same Process. Different Outcomes.

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

SEA Limited started life as a mobile gaming company. Founder Forest Li used the huge success of his mobile gaming platform to fund a new e-commerce division in 2015 called Shopee and since then have expanded into digital finance products like payment processing, mobile wallets, and credit.

Shopee is the largest e-commerce player on Southeast Asia with a ~50% market share. They are also rapidly gaining share in Brazil. Investors are concerned over increased competition and a decision to accelerate investment into their logistics network. SEA’s dominate market share and mobile gaming cash flows gives it the fire power to reinvest in the business and continue to expand their logistics moat.

SEA’s EBIDTA margin has fallen to 0.7% of GMV during their recent reinvestment cycle. This compares to their previous peak of 1.2% in 2023 and to Amazon and Mercado Libre’s current margins closer to the 3% - 4% range. Management guides to a longer term margin of 2% - 3%. At a 2% margin, SEA trades for an EV/EBITDA of 10x.

• An Article and Podcast from The Investors Podcast Network (Link, Link).

• A Pair of Writeups from Alexander (Link, Link).

Mattel 🇺🇸
NASDAQ: MAT • EV: $5.4B • Last Close: $13.42

Mattel is a global toy and media company. They own brands like Barbie and Hot Wheels. The market is pricing this as a stagnant business and are worried that kids will shift towards digital content. Underappreciated is Mattel’s IP library. It’s one of the last independent portfolios of high quality brands and would be worth a lot to a variety of buyers.

Shares trade for 8.5x Mattel’s last 5 year average FCF and 10x their guided FY26 adjusted EPS. They’ve repurchased 21% of shares outstanding since 2023, a pace they expect to continue.

• A VIC Writeup from 2022 (Link).

Check Out: Invisible Companies

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

“Age as a catalyst works especially well in Japan. Often, the reason they do this is because of tax on the family side. If they die while the family stock is 2.5x higher, their heirs are taxed 2.5x as much. But as soon as they die, the overhang is gone, and the situation can be fixed.”

- David Orr

Icom Inc.
TYO: 6820 • MCap: USD $281M • Last Close: ¥3.070

Net Cash / Market Cap: 65.5%

Price / 3yr avg. Op Income: 12.6x

Dividend Yield: 2.45%

Controlling Shareholder Age (Chairman): 95

Icom is a long‑running maker of professional radio‑communication gear used in marine, aviation, emergency services, and amateur radio. It’s a durable niche hardware business with a global customer base and steady replacement demand.

SK Kaken
TYO: 4628 • MCap: USD $910M • Last Close: ¥9.410

Net Cash / Market Cap: 69.3%

Price / 3yr avg. Op Income: 12.0x

Dividend Yield: 1.28%

Controlling Shareholder Age (Founder): 93

SK Kaken is a specialty coatings manufacturer focused on architectural paints with weather, mold, and fire‑resistant properties. Its products are embedded in long‑cycle construction and building maintenance markets.

Kyoritsu Computer & Communication
TYO: 3670 • MCap: USD $13M • Last Close: ¥1.698

Net Cash / Market Cap: 86.1%

Price / 3yr avg. Op Income: 5.8x

Dividend Yield: 3.83%

Controlling Shareholder Age (Founder, CEO): 91

Kyoritsu Computer & Communication is a small IT integrator providing custom software, network setup, and ongoing system support for regional businesses. It operates in the sticky, recurring world of legacy IT maintenance.

M-mart Inc.
TYO: 4380• MCap: USD $33M • Last Close: ¥1.110

Net Cash / Market Cap: 46.3%

Price / 3yr avg. Op Income: 10.3x

Dividend Yield: 2.25%

Controlling Shareholder Age (Founder): 90

M-mart operates B2B online marketplaces in Japan, connecting restaurants, hotels, food-service operators, and other commercial buyers with wholesalers of food ingredients and related products.

Ohmura Shigyo
TYO: 3953 • MCap: USD $16M • Last Close: ¥0.744

Net Cash / Market Cap: 79.1%

Price / 3yr avg. Op Income: 8.0x

Dividend Yield: 4.03%

Controlling Shareholder Age (Chairman): 87, died 1 year ago.

Ohmura Shigyo is a regional producer of corrugated cardboard and packaging materials supplying logistics and consumer‑goods companies. It’s a steady, volume‑driven paper‑products business tied to local distribution networks.

Artnature Inc.
TYO: 7823 • MCap: USD $183M • Last Close: ¥0.863

Net Cash / Market Cap: 72.2%

Price / 3yr avg. Op Income: 10.3x

Dividend Yield: 3.71%

Controlling Shareholder Age (Chairman, CEO): 84

Artnature is a wig and hair‑replacement company, selling custom hairpieces and operating nationwide salons. It serves both cosmetic and medical customers with recurring service revenue.

Aoba‑BBT Inc.
TYO: 2464 • MCap: USD $29M • Last Close: ¥0.316

Net Cash / Market Cap: 66.3%

Price / 3yr avg. Op Income: 10.3x

Dividend Yield: 3.48%

Controlling Shareholder Age (Founder): 83

Aoba‑BBT is an education company running business schools, online MBA programs, and corporate training. It focuses on adult learners and digital professional development.

Hokuyaku Takeyama Holdings
SAP: 3055 • MCap: USD $103M • Last Close: ¥0.814

Net Cash / Market Cap: 85.5%

Price / 3yr avg. Op Income: 8.0x

Dividend Yield: 2.46%

Controlling Shareholder Age (Chairman): 83

Hokuyaku Takeyama Holdings is a medical distributor supplying pharmaceuticals, devices, and hospital logistics services across Hokkaido. It acts as a key regional link between manufacturers and healthcare institutions.

PRAP Japan
TYO: 2449 • MCap: USD $30M • Last Close: ¥1.050

Net Cash / Market Cap: 107.7%

Price / 3yr avg. Op Income: 6.2x

Dividend Yield: 3.90%

Controlling Shareholder Age (Founder’s Widow): 81

PRAP Japan is a PR and communications firm offering media strategy, crisis management, and corporate messaging. It helps companies manage public relations and public reputation.

Musashi
TYO: 7521 • MCap: USD $127M • Last Close: ¥2.600

Net Cash / Market Cap: 109.1%

Price / 3yr avg. Op Income: 6.1x

Dividend Yield: 2.92%

Controlling Shareholder Age (Chairman): 82

Musashi supplies information/printing systems, industrial equipment, and election related equipment like vote-counting and ballot-processing. It also has paper/product distribution operations.

Nippon Primex
TYO: 2795 • MCap: USD $32M • Last Close: ¥0.949

Net Cash / Market Cap: 111.0%

Price / 3yr avg. Op Income: 8.8x

Dividend Yield: 2.63%

Controlling Shareholder Age (Chairman, CEO): 81

Nippon Primex is a maker of specialized printers and barcode systems used in retail, logistics, and industrial automation. Its devices are embedded in POS terminals, kiosks, and factory equipment.

Check Out: Planet MicroCap Las Vegas 2026 Pitch Session

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

Pershing Square USA is a closed end fund managed by Bill Ackmans Pershing Square. They went public in a combined IPO at the end of April raising $5B. This was initially supposed to be $10B, however the demand wasn’t there. Retail investors received 100% of requested shares which contributed to the 18% drop in share price on IPO day (receiving a 100% IPO allocation is rare so investors typically request more than what they actually want).

PSUS now trades at a NAV discount of 16%, most closed end funds trade at a discount in the 6% - 8% range. Bill Ackman and Pershing Square have a 2+ decade track record with the predecessor fund to PSUS returning 16.2% annually over that timeframe. Pershing Square’s London listed closed end fund also has a history of repurchasing NAV discounted shares to improve returns.

Potentially more interesting than PSUS though is the investment manager Pershing Square Inc. PS receives fees from their 2 closed end funds as well as the Howard Hughes Corporation. These are all permanent capital entities

• A Podcast on PS/PSUS from the Yet Another Value Blog (Link).

• Bill Ackman’s Pershing Square Inc. Letter (Link).

• A Recent VIC Writeup (Link).

MDB Capital Holdings 🇺🇸
NASDAQ: MDBH • EV: $10M • Last Close: $2.79

MDB Capital is a registered broker-dealer focusing on early public offerings and follow‑on financings for emerging deep-tech companies.

Over the last 5 years they’ve been building a clearing platform. This has weighed on profitability as they worked to clear technical and regularity hurdles. With this buildout now behind them they are actively discussing a strategic partnership and are open to an outright sale. MDB has also been investing in PatentVest, their AI powered patent prosecution platform which is planned to be spun out later this year. Post clearing platform and PatentVest transactions, management has guided to annual OpEx of ~$6M vs ~$10M today.

In addition to the clearing platform and PatentVest. MDB own shares of eXoZymes and Paulex Bio. eXoZymes is public while Paulex Bio has an IPO slated for September of this year. Their eXoZymes stake is worth $30M while the Paulex Bio stake is marked at ~$8M. Net cash sits at $20M.

They’ve been around for 30 years and are working towards their stated goal of 3 - 5 listing per year vs a historic pace of 1 every 18 months. At that 3 -5 IPO pace they’d be running FCF breakeven while maintaining upside through stock and warrant participation. At todays market cap your buying this core franchise for a negative value after backing out net cash + investments.

• MDB Capitals Presentation at Planet MicroCap Las Vegas 2026 (Link).

• MDB Co-Founder & CEO Chris Marlett on the Planet MicroCap Podcast (Link).

Mercer International 🇺🇸
NASDAQ: MERC • EV: $1.6B • Last Close: $0.73

Mercer International is the largest producer of mass timber in the US and is one of the world’s largest producers of Northern Bleached Softwood Kraft (NBSK) pulp. The entire industry has gotten crushed during an extended downturn in pulp and mass timber.

Mercer is currently doing ~$120M of run-rate operating EBITDA. This puts them at an annual burn rate of ~$70M after interest and capex. Remaining liquidity stands at $229M and they’ve received covenant waivers from lenders. Most of their debt matures in 2028 and 2029.

Mid-cycle EBITDA should fall somewhere around $400M putting today’s valuation at 4x EV/EBITDA. Previous mill transactions point to valuations in the $750/ton of capacity range, Mercer’s 2.1M tons of pulp capacity are considered mid-tier lower cost assets.

Management is engaged with the majority lender group suggesting a potential LME. The equity is left for dead and an investment here looks more like an option. There’s a serious risk of a zero but if the industry turns before liquidity runs out then you’ve got line of site to a 10x return.

• A VIC Writeup on the Debt Stack (Link).

Check Out: The iPhone’s Last Stand

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

Toast is a vertical software platform built specifically for restaurants. They’re the leading point-of-sale provider for SMB restaurants in the US and offer everything from digital ordering to payroll.

“And most people underestimate just how complex the customer workflow is at a restaurant. But most importantly, customers really love Toast. We do a ton of custom survey work on all our companies. Some notable takeaways as to why customers love Toast so much. Toast has a net promoter score of roughly 50, with 95% of respondents indicating they would recommend Toast. Just for reference, a net promoter score of 25 is considered really good. 30 to 50 is considered exceptional. They've got elite customer satisfaction metrics. On average, customers are using Toast for seven modules, so this isn't just a point-of-sale system like you might see with some of the competitors. It's a multi-product, mission-critical operating platform for its customers.”

- Sean Barrett, Founder and CIO at Counter Global

Toast’s hardware is a loss leader and is more complex than meets the eye. This helps keeps new entrants at bay, while their large data advantage is proving to be a major tailwind when rolling out new AI offerings. They have a ~20% market share in the US but are winning half of all new restaurant openings suggesting they’ve got a lot of growth remaining ahead.

Shares trade for ~20x forward earnings today, that’s too cheap for a sticky business that should compound EPS at 25%+ for the foreseeable future.

• A Great Podcast from Business Breakdowns on Toast (Link).

• A Writeup and Recent Writeup from Bob Hammel (Link, Link).

• A Short VIC Writeup (Link).

Princes Group PLC 🇬🇧
LON: PRN • EV: £327M • Last Close: £3.12

Princes Group is a broken IPO with limited liquidity, a complex governance structure, and an unexpected CEO replacement all contributing to the 35% share price decline since their debut last year.

In 2024 Newlat (BIT: NWL) acquired Princes from Mitsubishi for £700M and renamed themselves NewPrinces. NewPrinces (formerly Newlat) then merged their foods businesses into the newly acquired Princes Group to create a pan‑European food and drinks group before listing it on LSE. The purpose of the IPO was to raise cash for further M&A. NewPrinces continues to own 87% of the shares outstanding.

Princes Group owns a portfolio of #1 and #2 food brands across pasta, canned fish, sauces, and more. They also do a significant amount of customer own brand business for major supermarkets and retailers.

They currently have £433M of net cash and in FY25 did £130M of underlying free cash flow. Today’s market cap sits at £760M. At an EV/EBITDA of 2x, Princes Group trades well below it’s peer group.

• A Writeup from The Oak Bloke (Link).

• A NewPrinces (Princes Group’s parent co) Witeup from The Value Desk (Link).

• A Writeup from Moram Capital (Link).

Clarus Corp 🇺🇸
NASDAQ: CLAR • EV: $88M • Last Close: $3.08

Clarus owns the Black Diamond climbing gear and Rhino Rack roof racks brands.

Chairman Warren Kanders owns 20% of the company and has a strong history of value creation at previous public companies. Clarus however is the exception, with multiple value destructive acquisitions and a business that has run at breakeven for the last couple of years.

Clarus’ share price struggles started in 2022 after a hedge fund blew up resulting in huge forced selling that sent shares down 70% in a matter of months. Shares never recovered and have only trended downwards along with deteriorating business fundamentals.

They’ll do $250M of revenue this year and precedent transactions suggest Clarus' outdoor adventure brands are worth 1-2x sales. The company is debt free and owns meaningful real estate in the Salt Lake City area.

Greenhouse Funds has increased their ownership to 15% and should encourage improved capital allocation going forward.

• A Writeup from Ragnarok Research (Link).

• A Podcast with Black Diamond Equipment founder Peter Metcalf (Link).

Check Out: Why I Avoid Tech…

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

MIAX owns a variety of securities exchanges with a focus on derivatives. They were co-founded in 2007 by CEO Thomas Gallagher, and launched their first options exchange in 2012. Since then MIAX has grown its total share of US multi-listed options volume to ~17% thanks to it’s modern tech-stack and well incentivized capital structure with the large trading companies having receiving warrants based on liquidity volume targets.

MIAX’s profitability has been skewed by the launch of their equities and futures segments, the former of which just hit breakeven and the latter is ramping spend ahead of a major Bloomberg products launch.

Looking under the hood, MIAX’s core options segment is doing ~$300M of options EBITDA at 75% EBITDA margins while growing revenues 40% YoY. 13x EV/EBITDA is way to low for this high quality exchange business. The Equities and Futures segments both provide large upside optionality as well.

• A Recent Writeup from Moody (Link).

• A Writeup and Recent Writeup from Collin Cook (Link, Link).

• A Writeup from Raging Bull Investments (Link).

WM Technology 🇺🇸
OTCMKTS: MAPS • EV: $5M • Last Close: $0.36

WM Technology operates the leading online marketplace for cannabis, Weedmaps. Consumers use Weedmaps because it brings together info on strains, clinical effects, pricing, THC/CBD content, and in-stock availability. Retailers pay Weedmaps for premium deal listings and advertising exposure.

In addition to being in the hated sector that is cannabis. Weedmaps came public via SPAC in 2021, has a complicated corporate structure with TRA liabilities, has been flagged for material accounting weaknesses, just delisted from the NASDAQ, and is run by an overall dog shit management team.

Because of this shares trade for a market cap of $62M despite having $57M of net cash and doing $40M of EBITDA last year.

A widely expected Cannabis rescheduling will allow cannabis firms to write off expenses other than COGS. This should be a tailwind for Weedmaps as tax savings are in part redirected towards advertising.

• A Writeup from TripleS Special Situations (Link).

• A Report from Rod Alzmann (Link).

• A Pair of VIC Writeup (Link, Link).

Check Out: Investing Lessons From a Midnight Revisiting the SpaceX Valuation: A Post-Prospectus Update!

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

From selling playing cards to the Yakuza to a newly hired art school graduate single-handedly saving the company with the overnight success that was Donkey Kong, Nintendo’s origins are interesting to say the least. For the full story, the two-part Nintendo series from the Acquired Podcast is a must listen.

I first got to know Nintendo via Crossroads Capital’s work. The below writeup draws heavily on their thinking and analysis. For further reading I highly recommend their 2018 and 2021 annual letters.

*Most of the charts and figures in this report are AI generated. The writing remains entirely my own. All figures are in USD unless otherwise noted.

The Ecosystem Blueprint

The success of Apple and their ubiquitous iOS ecosystem are well known. Investors have been rewarded handsomely with the stock returning 26% annually since the iPhones debut in 2007 (that’s an 80-bagger). While there are some important differences, the Nintendo of today looks very similar to the early days of the Apple iOS journey.

Over the past two decades Apple has created an incredibly sticky hardware-software ecosystem, one that ~90% of iPhone users choose to stick with when buying their next phone. The success of this ecosystem is largely attributable to the following key attributes:

1) Apple releases a new iPhone every year, each with its own models across different price points. New iPhones are launched with smaller incremental upgrades. They don’t have to worry that each new iPhone might be a total flop, because they’re all still basically the same device that millions of users love.

2) Apple maintains a virtually identical user interface across different device generations. Your 83-year-old mother or grandmother upgrading from an iPhone 12 is able to immediately start using today’s latest and greatest iPhone 17 with essentially no learning curve.

3) Apple ties users to its platform via user accounts and cloud-based data storage. Transferring apps, photos, message history, and contacts from old iPhones to new ones is seamless.

4) Last and most important, Apple maintains forward and backward compatibility. New iPhones run all existing iOS apps and older iPhones continue to run almost all new iOS apps. Developers know that all Apple iOS devices will continue to run their apps across generations. They also know Apple will continue to sell more and more iOS devices, growing the market for their apps over time. This dynamic means that the App Store has a far better selection and quality of apps than what is available on Android devices. This in turn makes Apple’s App Store more valuable to iPhone users. This self-reinforcing software-hardware iOS ecosystem is the key to Apple’s success and is what drives their ~$120B of high-margin App Store and services revenues.

“Apple’s iPhone is not simply a single purchase device, but a ticket into a one-of-a-kind indefinitely-lived software-based ecosystem.”

- Crossroads Capital’s 2021 Annual Letter

Nintendo: Past vs Present

The Hardware-Software Ecosystem

Prior to the Switch, Nintendo’s business model was one of feast or famine. After every console generation they needed to rebuild their user base from scratch. This meant huge swings in profitability as they went from hit success to flop (Wii to Wii U).

A successful console needs lots of users. To get lots of users, you need lots of good games. The catch though, is that game developers only make games for consoles with lots of users. Without a certain critical mass of users to kickstart this network effect, a new console inevitably turns into a flop. This dynamic drove once-mighty companies like Atari and Sega out of the console business and is the reason why Nintendo has historically held such a large cash balance in reserve.

But what if rather than reinventing the wheel every console generation, you instead continuously improved an already successful one indefinitely? Well, Nintendo has hinted at doing this as early as 2014.

“Consoles and handheld devices will no longer be completely different, and they will become like brothers in a family of systems… To cite a specific case, Apple is able to release smart devices with various form factors one after another because there is one way of programming adopted by all platforms. Another example is Android. Though there are various models, Android does not face software shortages because there is one common way of programming on the Android platform that works with various models. The point is, Nintendo platforms should be like those two examples.”

- Satoru Iwata, Former Nintendo President/CEO, 2014 Investor Call

And again more explicitly during a 2018 investor presentation:

“Up until now, the hardware lifecycle has trended at around five or six years, but it would be very interesting if we could prolong that life cycle, and I think you should be looking forward to that.”

- Shigeru Miyamoto, Nintendo Creative Lead

By transitioning to an iterative rather than one-off hardware model, Nintendo is creating a user base that continuously grows and never resets. As the iPhone has proven, one of the keys to success here is forward and backward compatibility. With the Switch 2, Nintendo has now implemented this all important feature. New devices run already existing games while old devices maintain forward compatibility to run most new games.

Nintendo has also launched Nintendo Switch Online (NSO), their subscription-based online gaming service which gives users free content, exclusive offers, and access to vintage games. More importantly, for the first time in company history, users can now easily transfer games, services, and data across devices.

“Nintendo’s relationship with customers is no longer based on a hardware device, but on their Nintendo Account that persists across multiple devices and which can now be continuously monetized.”

- Crossroads Capital’s 2021 Annual Letter

Just like Apple’s iPhone family of devices, the Nintendo Switch family of devices is now “forever”. The market does not fully appreciate this.

As mentioned already, lots of good games sells lots of consoles. The Switch’s indefinitely-lived hardware-software ecosystem means Nintendo and third-party developers can sell software titles, both new and old, into a massive and continuously growing installed base. The number of third-party software titles available on the eShop continues to grow and now represents over 50% of gross sales compared to ~30% in 2019.

Adding further momentum to the flywheel, the Switch, like the iPhone and unlike past Nintendo consoles, relies on off-the-shelf mobile components instead of custom ones. This lets Apple and now Nintendo ride the continuous mobile cost curve downward. Similar to how a flat screen TV that cost $2,000 in 2005 costs $200 today. The price of older Switch models will drop, boosting unit sales and expanding the Switch family’s user base.

Furthermore, today’s third-party game development engines like Unity make it much easier for game developers to make games for multiple platforms. The Switch is more friendly and more attractive to third-party developers than ever before.

Thanks to Nintendo’s transition to an indefinitely-lived Switch ecosystem, the eShop is quickly becoming an App Store-like third-party software distribution platform. One where software sales, not hardware sales drives earnings power.

The IP Flywheel

While Nintendo’s IP library has historically been under-monetized, this is quickly changing via a renewed push into theme parks, movies, and retail locations. While profitability is not the primary focus, Nintendo’s IP segment is nonetheless creating multiple high-margin, annuity-like revenue streams. More importantly, the real benefit of these new IP initiatives is reaching present and future Switch users.

Let’s take a look at the developing Nintendo Cinematic Universe. In 2023 Nintendo returned to the movie business for the first time in 30 years with The Super Mario Bros. Movie. A blockbuster that reached over 170M people in theatres and generated $1.4B in gross revenue at the global box office. Its sequel, The Super Mario Galaxy Movie, released this April has already grossed $1B. A live-action Legend of Zelda film is slated for 2027, after which management has confirmed it will “work up to a regular release schedule of one movie per year”.

These unique exposures to Nintendo’s IP reinforces the core Switch Platform business. The Super Mario Bros. Movie for example reached well over 400M viewers across theatres and streaming services. 6 months after the films release, Nintendo released a hugely successful new Mario game. Nintendo is now regularly gaining more exposure to potential customers than ever before in its history. Not only that, but they are doing so at a negative customer acquisition cost. Super Mario Bros earned an estimated $560M on a production budget of $100M! Reaching the same number of viewers via 30-second YouTube ads would cost up to $20M.

Nintendo is running the same playbook with theme parks. Super Nintendo World has proved a huge success for Universal who pays royalties to Nintendo, likely in the high single digit range of all park revenue.

Comcast has repeatedly cited Super Nintendo World as a contributor to growth at Universal’s theme parks.

“Domestic theme parks revenue increased, reflecting higher revenue at our theme park in Hollywood due to the continued success of Super Nintendo World, partially offset by lower revenue at our theme park in Orlando…”

- Comcast Q4 2023 Earnings Release

Nintendo’s IP is drawing more users into the Switch ecosystem, which encourages more games to be developed for the eShop, which in turn draws more users into the ecosystem, and so on.

The Pokémon Company

Nintendo owns stakes in a handful of various businesses, the most material of which by far is The Pokémon Company (TPC). While Nintendo’s reported stake in TPC is 33%, its real ownership is likely at least 50%.

TPC is owned one third each by Nintendo, GameFreak, and Creature Inc. What isn’t disclosed is Nintendo’s equity stakes in either Creature or GameFreak. Under Japanese securities law, Nintendo is not legally required to disclose its stakes in them. While not reported, Nintendo almost certainly owns material stakes in each. For example Game Freak already develops almost exclusively for the Switch and literally works out of offices on Nintendo’s corporate campus.

These undisclosed stakes are part of a classic Japanese cross-shareholding tactic aimed at obscuring Nintendo’s de facto control of TPC and makes uncovering the extent of Nintendo’s empire difficult by design. In past eras this made sense if you wanted to guarantee survival for the next 100 years. However with Japanese corporate culture slowly changing, this is no longer the case. Fully consolidating Nintendo’s full economic interest in TPC, as was done in 2021 with another of Nintendo’s partner studios, Next Level Games, remains a major future catalyst.

Also unmentioned in Nintendo’s filings is its ownership of the actual Pokémon trademark and the associated trademarks of individual Pokémon characters, which it licenses to TPC. TPC merely acts as its agent, overseeing the global licensing operations of the Pokémon business empire. With this in mind, ask yourself, who really owns The Pokémon Company? While impossible to narrow down, everything points to Nintendo’s true ownership of TPC landing at at least 50%. For further reading on the topic I’d recommend this article.

TPC is perhaps the most valuable media entertainment IP on Earth.

TPC’s latest reported annual revenue was $3.5B, which given the low-cost nature of collecting royalties should throw off annual FCF of approximately $2.6 billion. A high-margin, highly recurring media franchise such as Pokémon is worth 20x FCF any day of the week. Conservatively assuming Nintendo’s stake in TPC at 50% gets us to a value of $26B.

Other stakes include DeNA, Bandai Namco, Niantic Spatial, and 10% of the Seattle Mariners (yes, you read that correctly, Nintendo saved the team from moving to Tampa in the 90’s). Together, these are worth around $1B.

What is The Market Missing?

The IP fueled transition to an indefinitely-lived hardware-software ecosystem discussed above is a multi-year, if not multi-decade process. While Nintendo and long-term investors such as myself are focused on the health of the business over the long-term, the market at large is not.

“As the CEO and Manager of Nintendo, I should not be too concerned with very short-term share price fluctuations, otherwise we will lose sight of what’s really important.”

- Satoru Iwata, Former Nintendo President/CEO, 2004 Interview

“A year is actually less than the average product development cycle… development for most of them started two or three years ago… some software even have a timeline of 3 to 4 years because what we start making today needs 3 to 4 years before they can exhibit their real worth.”

- Satoru Iwata, Former Nintendo President/CEO, 2009 Investor Q&A.

“It is difficult to predict the changes in the external environment… but I think it is not an appropriate approach to be excessively influenced by short-term trends…”

- Shuntaro Furukawa, Nintendo President/CEO, 2026 Investor Q&A.

Investors are fixated on short-term headwinds that are not relevant to Nintendo’s long-term earnings power. These include 1) margin compression from memory card prices, and especially 2) weak FY27 guidance. These two factors when combined with the markets overemphasis on quarter-to-quarter results has shares down 50% from their 52-week high. This is a blessing in disguise as the market is missing the forest for the trees here. While investors panic over misplaced fears, we get a chance to purchase shares at their most attractive valuation in over 5 years.

Memory Card Margin Compression

As part of the AI compute demand bonanza, memory card prices have gone through the roof. Investors are concerned over input costs and the resulting pressure on Switch hardware margins.

Management expects a ~$600M FY27 earnings drag from memory/components and tariffs. The timing of this headwind with the launch of the Switch 2 is unfortunate and while the numbers are certainly not insignificant, the hardware margin pressure obscures the longer-term picture. Ultimately, it’s Nintendo’s software sales and third-party monetization that drives the earnings power of the business.

Nintendo announced Switch 2 price increases in May as a direct result of rising component costs. This takes us to what investors are primarily concerned about, go-forward Switch unit sales.

FY27 Guidance

The Switch 2 sold 19.8M units in FY26, its first year on the market. Expecting continued momentum, investors were disappointed with management’s FY27 guidance of only 16.5M units. The recent price hike from $449.99 to $499.99 has further raised fears over the declining Switch sales narrative. This narrative is wrong though, looking past the noise there is ample evidence suggesting the future ahead is bright for the Switch 2.

Nintendo is infamous for sandbagging guidance. They initially guided for only 15M Switch 2 units in FY26 despite selling almost 20M. Since FY2018 they’ve under-guided hardware units by an average of 10% and software units by almost 30%. The light FY27 guide should not be of concern and is simply a reflection of management’s overly conservative nature. According to a recent Bloomberg article, Nintendo has already asked partners and suppliers to assemble 20M Switch 2 units for FY27.

Unlike the original Switch which had a stacked first year of “system seller” game releases. Nintendo has strategically back-loaded its biggest Switch 2 games. The biggest driver of hardware sales is games, and Nintendo is positioning itself for sustained momentum across the Switch 2’s life cycle rather than risking a mid or late generation drought.

“…it is not an appropriate approach to be excessively influenced by short-term trends. The second and third years for Nintendo Switch 2 are very important, and if we can expand the hardware installed base, we can use that as a basis to greatly expand software sales.

- Shuntaro Furukawa, Nintendo President/CEO, 2026 Investor Q&A.

The Switch 1 benefited from an unusually strong launch year, including Super Mario Odyssey, Mario Kart 8 Deluxe, and Game of the Year winner: The Legend of Zelda: Breath of the Wild. The Switch 2 on the other hand launched with just one new system seller, Mario Kart World, which was later supplemented by Donkey Kong Bananza. Upcoming major Switch 2 releases include Splatoon Raiders, Pokémon Pokopia, and importantly, a rumored new 3D Mario and Legend of Zelda remake.

Upcoming first-party Nintendo system sellers aren’t the only thing to be excited about, the Switch 2 has also seen a step change in the availability of third-party AAA games. The Switch 2's more powerful hardware allows today's major multi-platform titles to run natively, without the technical compromises that accompanied the original Switch. While AAA games on the Switch were limited and with release dates well behind PS4/Xbox deputs, the Switch 2 already has an impressive AAA lineup including titles like Call of Duty and Cyberpunk 2077 with release dates matching PS5 and Xbox.

“For the first time in history, gamers are able to play high-powered AAA titles not just at home, but on the go with dedicated controls.”

- Crossroads Capital’s 2021 Annual Letter

Again, it’s critical to highlight that when it comes to video game consoles, software sells hardware, not the other way around. With this in mind, the Switch 2’s most important software releases are ahead of it, not behind it.

Valuation

The most important Nintendo KPI’s are 1) the number of Annual Playing Users (APUs), and 2) net software revenue per APU. My view on valuation and the overall health of the business can be backed into these two numbers.

The numbers of APUs, which Nintendo defines as the number of users that have logged into their Switch account at least once in the last year, has flatlined since 2024. At face value this is concerning and describes the drought mentioned above that Nintendo is hoping to avoid with the Switch 2 via a more back loaded game release schedule. If the number of APUs were to start dropping, it would break the entire Switch ecosystem thesis. Obviously I don’t expect this to happen. In fact, holding APUs flat at the end of a 9-year hardware cycle amid weak new software releases is actually pretty impressive.

Normalized Switch unit sales of 18-20M per year seems about right as a conservative guess. This checks out when compared to the Switch’s average unit sales of 17M over its first 9 years, and also when considering the 5-7 year average device cycle. Normalized Switch unit sales should grow in line with APUs, hence its importance. I’m also assuming a gross margin of 20% on hardware, below the Switch 1’s ~30% margins from 2022 but above today’s temporarily pressured Switch 2 margins.

Net software revenue per APU has trended downwards since 2018, this number includes both video games sales and NSO subscriptions. This decline was to be expected because 1) The most hardcore Nintendo fans will buy Switch first and they have the highest revenue per user. As the less fanatic gamers join the Switch platform, they pull down the average software revenue per user. And 2) COVID saw a huge demand pull-forward which has now run off.

Going forward, net software revenue per APU should trend upwards. Despite the Switch being the more popular platform, it is well behind the Xbox and PlayStation when it comes to monetization. AAA games also come with higher average selling prices which given the influx of their availability on the Switch 2 will be a tailwind for net software sales. Nintendo Switch Online also has lots of untapped pricing power. The top NSO tier costs $49.99/year, unchanged since its original launch. Meanwhile the top tiers of PS4 and Xbox versions of NSO cost $159.99/year and $276/year respectively.

For longer-term software margins, I’m assuming 85% compared to ~70% today. This bakes in: 1) Digital sales expanding to 80% of total software sales vs 55% today and a mere 15% in 2017. There’s no cost to produce a physical copy and Nintendo gets the full retail instead of wholesale price so digital sales carry a ~90% gross margin instead of ~55% for physical. 2) Third-party gross sales rise to 60% of eShop sales vs ~50% today. Nintendo charges an estimated 30% take rate that carries near 100% gross margin. 3) Further expansion of high-margin NSO subscription and digital content sales.

R&D, marketing, and G&A should run around 22% of revenues. Again, these are conservative numbers. The corporate tax rate in Japan currently sits at 31%.

Putting this all together I get longer-term, “steady-state” hardware margins of 20% and software margins of 85%. At today’s hardware/software mix, which will only improve going forward, we arrive at an operating margin of 22%. We can then apply these steady-state margin assumptions to Nintendo’s current hardware and software revenues to get an idea of their underlying earnings power before applying an appropriate multiple.

The above steady state earnings estimates definitely err on the conservative side, and give zero credit to Nintendo’s growing IP business. I think we can also all agree that a 12x multiple is way too low for a business of this quality. In light of this, I don’t see how Nintendo’s core Switch Platform business could be worth anything below $25B. Realistically, I would assign a fair value upwards of $60B, and this still assumes essentially no APU growth and a below market multiple of 18x.

At today’s market cap of $54B, the downside is so protected it’s almost laughable. TPC and Nintendo’s other stakes are worth a conservative $27B. Add in their $14B of net cash and we are paying an implied $13B for the world’s best and most beloved game franchise along with the single greatest content creation track record of all time. Not to mention Nintendo’s swath of various other undisclosed assets, for example a portfolio of Kyoto real estate amassed over the previous century.

Don’t forget, that this business, for which we are buying for $13B, is in the early innings of its business transformation. My $60B valuation estimate assumes the Switch Platform largely plateaus at its current state. If instead Nintendo continues to execute, which all of the signs currently indicate, and APU growth resumes its trend upwards, there is a credible path to $110B+ in value.

Conclusion

Nintendo has released a staggering 21 of the 25 best-selling console and handheld games of all time. They are the undisputed video game console king and are backed by one of the most valuable media franchises on earth. Thanks to unwarranted short-term market fears, we have been offered a 2007 Apple-esque opportunity to buy into the IP fueled Switch ecosystem at a significant discount to its fair value. As this ecosystem continues to build momentum, investors will eventually wake up to the reality that Nintendo is no longer a low-margin hits-driven business, but instead a high-margin software fueled family of indefinitely-lived Switch devices.

I’m buying Nintendo today at ¥7,500/share, a 46% discount to my conservative fair value estimate of ¥14,000/share. If I’m wrong, and the Switch Platform initiative fails, my downside is more than protected by a large net cash position, TPC, and Nintendo’s various other hidden assets.

*Shares of Nintendo are also available via their ADR listing on the OTC Market.

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

Luckin Coffee is the largest fresh-made coffee store operator in China. Per capita annual coffee consumption has grown from 4 to 22 cups over the last decade and is only accelerating. Per capita annual consumption in Japan and South Korea is around 300-400 cups.

Luckin won the coffee race in China with it’s mobile first, low-cost offering. Online ordering eliminates wait times and compresses store footprints. A new Luckin store costs 350-450K RMB and needs 5-6 staff while a new Starbucks costs 1.5-2.5M RMB fand needs 10-15 staff. Luckin’s average selling price is 13 RMB vs 20+ at Starbucks.

Shares trade for 14x earnings while EPS should CAGR in the 20% range for the foreseeable future thanks to market growth of 15% and expanding profit margins.

• A Recent VIC Writeup (Link).

• A Writeup from Keeping Identity Small & Light (Link).

MercadoLibre 🇺🇾
NASDAQ: MELI • EV: $89.9B • Last Close: $1,672.83

MercadoLibre is the dominant e-commerce and fintech platform in Latin America. They combine features of Amazon, eBay, and PayPal. Shares have returned 24% annually since they’re 2007 IPO.

Meli is a network effect and scale advantages power house. They have 30% market share of all e-comm in Latam and 90%+ of items sold on its marketplace are shipped through there proprietary logistics network.

26 years of purchase history, browsing behavior, and payment data have created a swath of valuable data. Meli has been increasingly monetizing this data through their fast growing lending and advertising businesses.

Investors are concerned over competition from Amazon, Shopee, and Temu which risk pricing pressures, and higher customer acquisition costs. The rapidly growing consumer credit book has also raised some eyebrows. Meli’s real advantage though is their physical infrastructure. The reason someone uses Meli isn't the website, it's that they can get a package delivered to their door in hours, at a similar price and lower effort than leaving their house. Meli’s dominate positions means they can spread the cost of this infrastructure over more users than competitors.

Shares trade for an EV/EBITDA of 18x, an all time low for the company.

• A Short VIC Writeup (Link).

• An Article and Podcast from The Investors Podcast Network (Link, Link).

• A Writeup from Rebound Capital (Link).

• A Writeup from Heavy Moat Investments (Link).

Check Out: Investing Lessons From a Midnight Car Wash

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

In recent years Vantage Drilling has sold rigs and pivoted to an asset light management business. They have a JV with TotalEnergies for the Tungsten Explorer, a 6G ultra deep water drillship. Vantage owns 25% of the JV and receives $17m per year under a 10-year management agreement.

Vantage’s last fully owned rig, the Platinum Explorer, is on a 3+1 year contract with ONGC at $240K/day which should translate to ~$40M of annual EBITDA. They also manage jackup rigs for ADES for which they receive $7.5M a year. Corporate cost are running at $30M a year.

The real opportunity is if Vantage can build out their rig management platform. They have a long track record working with the major rig owners and recently signed a marketing agreement with Eldorado for a 7G drillship.

Very little liquidity and no analyst coverage has caused a disconnect in the share price compared to fair value.

• A Recent VIC Writeup (Link).

Petra Diamonds 🇿🇦
LON: PDL • EV: $366M • Last Close: £0.15

Petra Diamonds owns 2 of the highest quality underground diamond mines in the world. They are unprofitable, leveraged, and diluting shareholders with PIK coupons. The risk of bankruptcy was high until recently when Petra completed a restructuring extending their maturity wall from 2026 until 2030.

The combination of lab grown diamonds and a weakening luxury market has not been kind to diamond prices which have been on a steady march downwards since 2022. Improvements in technology have collapsed the cost of lab grown diamonds turning them into a commodity. Going forward they should be less of a threat to natural diamonds as the exclusivity of rare natural stones widens vs the now abundant low cost lab grown diamonds.

While the debt load means there is real risk of a zero, the right time to buy into a cyclical industry is when things are priced for death. Meanwhile, Petra has been making real operational and cost reduction progress. When the diamond cycle eventually turns, high fixed costs means revenue drops disproportionately to the bottom line.

• A VIC Writeup (Link).

• H1 2026 Results Presentation (Link).

Nichols plc 🇬🇧
LON: NICL • EV: £295M • Last Close: £9.46

Nichols sells Vimto, a fruit-flavored soft drink that first hit the shelves in 1908. While majority of sales come from the UK, the brand is becoming increasingly popular in Africa representing the main growth opportunity ahead.

This is a simple, steady business with a 30%+ ROIC that should grow profits organically at a mid-single digit rate. FY25 saw £34M of PBT at a 19% margin, managements 2029 target is for £225M revenue at a 20% margin.

A business of this quality should trade higher than 14x earnings. While waiting for shares to rerate you get paid a 5% forward dividend yield .

• A Great Writeup from Roland Head (Link).

• A Pair of VIC Writeups (Link, Link).

Check Out: Three Things – Is This a Bubble?

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

KITS is a vertically integrated manufacturer and e-commerce retailer of glasses and contacts. By owning their own optical labs and selling direct to consumer, KITS can sell eyeglasses for less than a 1/5th the price of competitors while remaining considerably more profitable. Customers also have the benefit of receiving their KITS glasses in days rather than weeks.

Founder and CEO Roger Hardy scaled his previous company, Coastal Contacts, from $30M in sales in 2004 to $200M+ before selling to Essilor for $430M in 2014.

KITS CAC/LTV is exceptional. Revenue has grown at 30%+ over the last 5 years which has been funded entirely by cash flow. Management is aiming for $500M in sales by 2030 which would see EBITDA margins in the 10-15% range.

• A Writeup from Pound the Rock Investing (Greystone Capital) (Link).

• A VIC Writeup (Link).

Playboy Inc 🇺🇸
NASDAQ: PLBY • EV: $261M • Last Close: $1.20

After a rough couple of years, Playboy has just one non-core asset left to sell (Honey Birdette) to complete their transition to an asset light royalty company. They’re collecting $33M of guaranteed licensing revenue from deals with Byborg and UTG plus ~$15M from other non-guaranteed licensing deals which are growing at a healthy clip.

Honey Birdette should fetch $50M at the very minimum. Proforma, Playboy will be left with a ~90% gross margin royalty business abd $24M of overhead which has a lot of room to come down.

Check Out: 18% IRR for 57 Years

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

Wireless plan pricing in Canada is notoriously high compared to the US. Quebecor is a low cost wireless upstart out of Quebec who is now expanding nationally after the acquisition of Freedom Mobile. They’re counter positioned against the incumbents who have high debt levels and big dividend payout ratios. Quebecor is operating at close to 50% EBITDA margins despite selling pans at a ~40% discount to competitors.

They currently have an 11.5% share of total industry subscribers and are rapidly outgrowing the incumbents. CEO Pierre Karl Peladeau owns $4Bworth of stock. Shares trade at ~10x FCF.

• A Writeup from Third Point Capital (Link, pg. 6).

Costar 🇺🇸
NASDAQ: CSGP • EV: $13.0B • Last Close: $33.05

At it’s core, Costar is an info services business for commercial real estate. Through four decades of on the ground property inspections, they are they sole comprehensive data provider on properties, leases, sales, tenants, market trends, and more. Investors are concerned that CEO Andy Florence will continue allocating the majority of free cash flow to his battle against Zillow through Homes.com.

Third Point and DE Shaw are pushing for changes and the board has formed a Capital Allocation Committee. Ex-Homes.com, shares trade for ~10x EV/EBITDA. Historically they’ve traded for 30x and info services peers are currently trading at 20x+.

• A Writeup from Show Me The Incentives (Link).

• A Writeup from MBI Deep Dives (Link).

• A Recent VIC Writeup (Link).

Check Out: The Wu Tapes

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

In 2021, GreenFirst acquired six lumber mills, a paper mill, and land + timber assets for $234M ($140M for the mills and $88M for inventory). Since then the price of lumber has been at or below breakeven for GreenFirst. Talk about bad timing. Subsequently, they’ve divested non-core assets for ~$125M and are now left with the more desirable 4x Ontario mills.

These mills should be worth ~$200M. Competitor Interfor owns 12% of shares and has expressed acquisition interest. GreenFirst also owns lumber duty, tax loss carry forward, and non-core land assets. At bottom of the cycle pricing they have several years of liquidity runway left. A mid-cycle environment should see them producing $20m+ of free cash flow.

• A Writeup from River Oaks Capital (Link).

Walker & Dunlop 🇺🇸
NYSE: WD • EV: $2.3B • Last Close: $51.16

Walker & Dunlop brokers loans on behalf of commercial real estate owners. They hold licenses with Fannie Mae and Freddie Mac, which are strictly controlled and limited in number. WD earns a ~1% fee on origination and an annual ~0.24% fee for servicing the loan. Servicing is non terminable and extends for the life of the loan, typically 5-10 years. This creates predictable recurring earnings that are insulated via a regulatory moat.

Shares of WD have struggled recently due to a slow housing market and by concerns over credit risk. Originators like WD are responsible for underwriting loan credit and are therefore required to retain credit exposure, typically structured as a 5% first loss with losses above that shared up to a maximum of 20%.

Even in a worst case scenario with exceptionally high levels of fraud, credit exposures would total a small amount of the EV. A recovery in real estate capital markets would drive strong earnings growth, which even during the current prolonged housing downturn has averaged ~7%. Shares are available today for an EV/EBITDA of 7.5x compared to their historical range of 10-20x.

• A Writeup from Feather Fund (Link).

• A Recent VIC Writeup (Link).

Ascent Industries 🇺🇸
NASDAQ: ACNT • EV: $78M • Last Close: $14.94

After a long history of mismanagement, Ascent Industries is now a pure play specialty chemicals manufacturer with an aligned and experienced management team at the helm.

All three of Ascent’s facilities are currently running below 50% utilization meaning very little incremental investment is required to scale up to ~$20M of EBITDA in the near term (80% utilization), compared to breakeven today. With $58M of net cash, that would put shares at an EV/EBITDA of 4x while. Peers trade at 8x+ and management having been buying back a lot of stock.

• A Writeup from River Oak Capital’s H2 2025 Letter (Link, pg. 10).

• A VIC Writeup (Link).

Check Out: Letter #8: Bob Iger (2005)

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

eDreams is a European online travel agency (OTA) utilizing a subscription business model. Through the Prime subscription, customers pay an annual fee in exchange for flight and travel discounts, exclusive deals, and premium perks. A Prime membership pays for itself after 2 or 3 bookings.

The share price has been hammered after 1) Ryanair cut its flight capacity from being listed on OTA platforms, and 2), eDreams changed its Prime subscription from annual to monthly or quarterly payments. Both of which came as a surprise and are weighing down earnings this year.

Shares now trade for 5x FCF, with a 100M buyback in place. A subdued Ryanair headwind and normalized working capital flows will see FCF materially higher.

• A Pair of Writeups from Exceptional Value Stocks (Link, Link).

• A VIC Writeup (Link).

DGL Group 🇦🇺
ASX: DGL • EV: USD $136M • Last Close: AUD $0.40

DGL is a founder led Australian specialty chemicals company that just came out of a 6 month trading suspension due to inventory accounting issues. They are vertically integrated across chemical manufacturing, distribution, and waste and recycling. This helps with scale efficiency and pushes asset utilization and margins higher.

Most investors have moved on after a prolonged chemicals downturn which was topped off by the accounting mess. Shares trade well under tangible book value and for an EV of 4x underlying EBITDA.

“I have asked the Board to reduce my salary significantly until the financial performance of DGL Group improves. As the major shareholder, I do not feel comfortable drawing a full CEO salary until we can demonstrate that we are adding value for all shareholders. I would like to ensure that my interests are aligned with the interests of all shareholders." - CEO, Simon Henry

• An Industry Primer and DGL Writeup from ToffCap (Link, Link).

• A Writeup from David Katunarić (Link).

• A VIC Writeup (Link).

Theravance Biopharma 🇺🇸
NASDAQ: TBPH • EV: $463M • Last Close: $16.81

Theravance recently announced that its lead development program had failed phase 3 clinical trials and they would now accelerate a strategic review while ceasing R&D and reducing G&A headcount by 50%.

The remainco is left with ~$400M in cash, $100M in 2026 milestones tied to a low-hurdle sales goal, ~$2.6B in Irish tax assets, and ~$60 - $70M per year in FCF coming from their royalty in YUPELRI, a growing drug commercialized by Viatris with patent protection until 2039.

• A Writeup from Laughing Water Capital’s Q1 2026 Letter (Link, pg. 8).

• A Writeup from Clark Street Value (Link).

• Andy Summers on the YAV Podcast Discussing TBPH (Link).

Re/Max Holdings 🇺🇸
NYSE: RMAX • EV: $546M • Last Close: $11.29

Re/Max announced a deal this week to be acquired by The Real Brokerage, technology-powered real estate brokerage, for $13.80 in cash. This looks very similar to the COMP/HOUS deal announced last year which did not get a second request from the FTC. The shareholder vote looks like it will easily pass as well. Shares trade at a 22% spread.

Check Out: Beyond the Sky - Jeffrey Yan

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

Meritage Hospitality is a Wendy’s franchisee with 355 locations in the US. Wendy’s is in crisis mode with same-store sales falling 11.3% in Q4 2025. Meritage is now in forbearance with their lenders and paying interest only.

“Recent financial performance represents decreased sales and margin compression resulting from a combination of three primary external factors which include, 1) an unusual winter weather (La Nina) cycle in the Southern U.S. that created highly disruptive sales impacts in many of our southern markets, 2) decreased marketing funds and extreme digital discounting by the prior Wendy’s brand management team, and 3) record U.S. beef prices culminating from a low herd, Mexican border shut down on beef imports due to disease and tariffs on the importation of beef from South America.” - Rob Schermer, CEO, 2026 Shareholder Letter

They did ($6.8M) of EBITDA this year vs their previous 5-year average of ~$40M. This was their first year of negative EBITDA in two decades. A turnaround is underway, insiders own 61% of the company, and shares trade for a value of ~$1.6M per location.

• Meritage Sidoti Microcap Investor Conference Presentation (Link).

Barry Callebaut 🇨🇭
SWX: BARN • EV: USD $12.2B • Last Close: CHF 1071.00

Barry Callebaut is responsible for 20% of all chocolate produced in the world. They source, process, and manufacture cocoa into into bulk and specialty chocolate. The company enters into long-term contracts with customers such as Nestle, Mars and Hershey, and is paid via a cost-plus formula. This creates strong visibility into future profits for which investors have historically awarded a ~15x EBITDA multiple.

In recent years, however, Barry Callebaut ran into a dual crisis of 1) a salmonella incident at their largest factory, and 2) a cocoa supply shock sending prices up by 5x. While profits were safeguarded, working capital exploded sending Net Debt/EBITDA over 6x.

Both crisis are now behind them. Looking out a year or two, EBITDA should exceed $1.1B+ with Net Debt/EBITDA falling back below 1x. This would translate to a share price of CHF 2800+ at their historical multiple.

• A VIC Writeup (Link).

• A 2024 Writeup from Turtle Capital (Link).

Reading International 🇺🇸
NASDAQ: RDI • EV: $208M • Last Close: $1.19

Reading International is a bit of a mess. Succession battles, entrenched dual share class structure, and questionable capital allocation. There are some positives though that point to a potential liquidation in the near future. The controlling family members are both 60+ and asset sales have been cleaning up the capital structure.

Today the EV sit’s at $208M which is approximately what their remaining real estate assets are worth (see Dylan Marrello’s writeup below). This leaves us with their cinema business thrown in for free, which should be capable of at least $25M in EBITDA. Comparables trade for up to 10x. At 8x, the equity is worth $18/share. Notably, the super voting Class B shares trade for $9/share with the same economic rights as the more liquid class A shares.

• A Great Writeup and Recent Updates from Raging Bull Investments (Link, Link, Link, Link).

Check Out: Why Is Canada’s Population so Concentrated?

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

Superior Plus is an over the road propane and compressed natural gas distributor, delivering over 1B gallons annually throughout North America. They’ve struggled with quality of service and operational execution in recent years but the tide is beginning to turn with new management making steady progress on the cost to serve and customer growth.

It’s expensive and unpractical to expand natural gas distribution to rural areas, these customers instead rely on transportable fuels. Route density creates meaningful costs advantages for scaled players while the majority of customers lease their propane tank from Superior creating pricing power and low churn due to removal fees and pump out costs.

The industry is highly fragmented, however most major operators are publicly traded master limited partnerships (MLPs) with minimum quarterly distribution targets. This comes at the cost of re-investment. Superior Plus is not an MLP and as one of the most advanced players, can continue to achieve meaningful synergies on acquisitions. Small and midsize distributors lack the tank monitoring penetration, scheduling and routing technology, and pricing sophistication that Superior has.

In 2020, one of Brookfield’s private equity funds (which target 20%+ returns) made a $260M investment through perpetual preferred securities with a 7.25% coupon that can convert into 30M shares at $12/share. Emeth Value expects Superior will generate more than its current market capitalization in free cash flow over the next 5 years.

• A Great Writeup from Emeth Value’s H2 2025 Letter (Link, pg. 5).

• Superior’s Jan 2026 Investor Deck (Link).

Salesforce 🇺🇸
NYSE: CRM • EV: $162.9B • Last Close: $171.31

Salesforce is a customer relationship management and enterprise resource planning software provider. They’re applications are centered around collecting, analyzing, and utilizing customer data. Switching from Salesforce’s core systems like ERP, financials and HR is difficult, especially when these complex interconnected systems are deeply integrated with customer data.

An experienced founder led management team has successfully navigated a constantly changing technology market for 27 years now and are harnessing AI to make the product better. Shares trade for 13x FCF with a $50B repurchase program in place.

• Some Recent Updates from Summit Stocks (Link, Link, Link).

• A Recent VIC Writeup (Link).

• A Writeup from Drew Cohen (Link).

Old Market Capital 🇺🇸
OTCMKTS: OMCC • EV: $15M • Last Close: $4.50

Old Market Capital, formerly Nicholas Financial, wound down their previous business and sold their subprime auto loan book in 2024. They then acquired majority ownership of Amplex Internet for ~10x free cash flow. Amplex CEO Mark Radabaugh continues to own 30%.

Amplex is building out fiber in rural towns. They received a 20 year, $21M loan from the USDA at a 2% interest rate to aid them in this. Amplex owns the trucks and a local service team as opposed to the big internet providers who outsource the work and have little interest in small towns. They should generate ~$6m of FCF compared to $4M at the time of acquisition and OMCC recently down listed to the OTC Markets which should cut overhead costs to less than $1M.

OMCC also has $50M NOL’s and $15M of cash that they will use to increase their ownership of Amplex (they have options to acquire up to 80% of the business) and to buyback shares. There remains plenty of fiber passing for Amplex to continue building out into the future as well.

• A Writeup and Recent Update from River Oaks Capital (Link, pg. 38, Link).

Medical Facilities 🇨🇦
TSE: DR • EV: $133M USD• Last Close: $16.80

Since an activist campaign in 2022, Medical Facilities has been selling assets at 8-10x EBIT and is now left with 51% stakes in two surgical hospitals in the US. These hospitals are of equal quality to those already sold and are doing a combined $54M of EBIT.

Both hospitals are grandfathered in by the Affordable Care Act and rank in the top 5% nationally for patient outcomes. Better yet, management has been aggressively buying back under valued shares with sales proceeds.

• Recent Writeups from Maaiz Khan, Keir Reynolds, and Old Rope Research (Link, Link, Link).

• River Oaks Capital Medical Facilities Pitch (Link)

• A VIC Writeup (Link).

Reckitt Benckiser 🇺🇸
LON: RKT • EV: $38.2+B • Last Close: $50.66

Reckitt Benckiser owns a portfolio of consumer health and hygiene brands including Lysol, Durex, and Mucinex. They recently divested their non core Essential Home business but shares have been under pressure due to litigation at their baby formula subsidiary. Notably, Reckitt has not faced any product recalls unlike their competitors. Management expects this segment to also be divested in the near term.

What’s left is a durable #1 consumer branded business trading at 14x 2027 earnings with a 4.3% dividend yield. This is the lowest valuation in over a decade and management is leaning into buybacks.

• A VIC Writeup (Link)

Check Out: The Passive Index Bubble: Myth or Reality?

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

Cooper Companies makes contact lens’ and various fertility products. Regulation, technical complexity, and customer loyalty have created a high barrier to entry oligopoly in the contact lens industry. Cooper is the primary supplier of private label business and has consistently taken market share from JNJ and Alcon. Cooper is the largest player in the in the fertility and women’s health industry. Both of these industries are consistent growers in the high single digits range.

Shares trade for 15x earnings compared to their longer term average of ~24x. Management compensation also recently changed to include total shareholder return and free cash flow incentives.

• A VIC Writeup (Link).

Scotts Miracle-Gro 🇺🇸
NYSE: SMG • EV: $6.1B • Last Close: $61.46

Scotts is a leading manufacturer of consumer lawn and garden products in North America. Brands include the likes of Scotts, Turf Builder, and Miracle-Gro.

The company over-levered to enter the cannabis industry and set almost $2B on fire rolling up companies in the indoor and hydroponic gardening space. This along with a tough period post a surge in demand from COVID has shares trading for 15x earnings. A return to pre COVID margins would see that drop closer towards 10x.

“Our valuation is completely unfair, but I get it - we’re in the penalty box… This is a very unique consumer franchise that is not being properly valued.” - CEO Jim Hagedorn

• A Writeup from TSOH Investment Research (Link).

• A Short VIC Writeup from 2024 (Link)

• An Article on the Harthorne Debacle (Link)

Check Out: The Seth Godin interview: What I learned about the long view from sci-fi legends

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

GetBusy is a small UK owner of 2 software companies. SmartVault is a USD $18M ARR vertical SaaS product for SMB accounting an tax professionals with impressive KPI’s. The controlling family of GetBusy has indicated their intent to sell this business. Workiro is a profitable collaboration platform integrated into Oracle’s NetSuite ERP product with £9.3M of ARR. They allow document management, task automation, secure file sharing, e-signatures, client/supplier portals, etc., and NetSuite’s 41,000 customer base provide a large TAM to sell into.

Management have been buying in the open market and are incentivized to return cash to shareholders with incentive plans that start to payout after £70M in gross cash distributions to shareholders or a USD $90M+ sale of their SmartVault business. Incentives max out at £150M in gross cash distributions and a $250M sale of SmartVault.

• A Recent VIC Writeup (Link).

• A Writeup from Compound & Fire (Link).

Compass 🇺🇸
NYSE: COMP • EV: $7.8B • Last Close: $7.31

After the recently completed merger with Anywhere Real Estate, Compass is by far the largest real estate brokerage in the US. They’re trading at an EV/EBITDA of 7x which is at the bottom of where they’ve historically traded despite the housing market being at a cyclical low.

COMP’s scale has allowed them to invest in creating the best technology stack, an advantage that real estate agents love. Their scale amongst a highly fragmented market also allows for exclusive listings and the muscle needed to pushback against housing portals like Zillow. Achieving expected deal synergies and a stabilized housing market would drop the valuation to 5x (~18% FCF yield).

• Commentary from Lake Cornelia (Link, Link).

• A VIC Writeup (Link)

Card Factory 🇬🇧
LON: CARD • EV: £301M • Last Close: £0.64

Card Factory is the leading greeting cards retailer in the UK and is the only vertically integrated operator. They are capital light with a negative working capital cycle and all 1,103 stores on lease. Despite selling into a flat TAM, CARD has grown revenues through steady market share gains. Management has guided to medium term PBT growth of mid-to-high single digits.

Shares trade for 4.5x earnings with an 8% dividend yield. A key operational risk is managements push into online retailing.

• A Writeup from Kostadin Ristovski (Link).

• Closed Position Updates from Pernas and Mallard Research (Link, Link).

• An Interview with CEO Darcy Willson-Rymer (Link).

Baltic Classifieds Group 🇬🇧
LON: BCG • EV: £818M • Last Close: £1.83

Baltic Classifieds Group consists of 14 online classifieds portals in the Baltic region, the majority of which are the dominant player by a wide margin. Classifieds portals are two sided marketplaces with high margins, zero capex requirements, and no working capital. Also of note is that BCG is under monetized compared to competitors and the Baltic region is the fastest growing in the EU.

Temporarily depressed vehicles sales in Estonia from a new tax and AI disruption fears have shares trading at 14x earnings, well below their 28x average since the 2021 IPO. All free cash flow is returned to shareholders through dividends and buy backs.

• A Writeup from Johan Lunau (Link).

• A Series of VIC Writeup’s (Link, Link, Link).

Swiss Marketplace Group 🇨🇭
SWX: SMG • EV: $3.5B USD • Last Close: CHF 27.40

Another classifieds business, this time in Switzerland. SMG was formed in 2021 after the two long standing classifieds competitors in the country combined their auto and real estate portals. They currently trade for 24x FCF with room to expand monetization and recently cleared a regulatory overhang with the Swiss Price Supervisor body.

• FY 2025 Results Presentation (Link).

• A VIC Writeup (Link).

Check Out: College Admissions Essays

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.

• A new Ashtead Technology Writeup from James Emanuel (Link).

Atlas Engineered Products (AEP)

Shares of Atlas are hovering near 52 week lows as the housing downturn continues to weigh on new builds.

• A New Writeup from Canadian Value Investors (Link).

BFF Bank (BFF)

BFF Bank is now trading for half of book value after an investigation into their financial reporting standards by the Italian regulators came to light.

Basic Fit (BFIT)

Basic Fit has performed well this year as membership numbers at mature clubs track higher than expectations. They also made a move into the franchising model through an acquisition in Germany.

• A Recent Writeup from Casteleyn Partnership (Link).

Burford (BUR)

Shares of Burford got crushed after their case against YPF was overturned. The market appears to have over reacted with shares trading well below the value of their business ex the YPF claim.

CAB Payments (CABP)

CAB Payments received a USD $1.15 per share (84p per share) offer to be taken private which was rejected by the board on the basis of undervaluing the Company.

• The Offer Announcement from CABP (Link).

Dexelance (DEX)

Being at the crossroads of home furnishings and luxury brands, Dexelence has gotten hammered this year. Shares now trade for an EV/EBITDA of 4x.

F.I.L.A. (FILA)

• A Bloomberg article from November going over some options for FILA’s DOMS stake (Link).

Floor & Decor (FND)

Floor & Decor is getting interesting at these levels with shares near all time low valuations. Charlie Munger has called them an “inevitable”.

• Bristlemoon Capital’s Floor & Decor Writeup from their Q4 2025 Letter (Link).

Helios Towers (HTWS)

Helios continues to execute on all KPI’s.

• A New Writeup from Jussi Askola (Link).

Lifecore Biomedical (LFCR)

Lifecore shares are down sharply after disappointing the market with weaker than expected revenue guidance. Shares trade for under 16x EBITDA with capacity at ~20% while peers have transacted for at least 18x.

PAR Technology (PAR)

PAR has gotten caught up in the SaaS beat down with shares down over 80% from their peak. The lack of a major tier 1 customer win has spooked investors as well.

Six Flags Entertainment (FUN)

Six Flags continues to struggle after tough weather conditions and a harder than expected integration with Cedar Fair. Activists are firmly focused on the company and own a material amount of stock.

• Some FUN Commentary from the 13D Activist Fund’s Q4 2025 Letter (Link, pg. 5).

Offshore Drillers (VAL/RIG/NE/SDRL/BORR)

Transocean and Valaris announce a transformative merger that bodes well for the entire industry.

• A Nice Writeup from Hugo Navarro and Mihail Stoyanov (Link).

Vistry (VTY)

Vistry has traded down to ridiculous levels after announcing disappointing guidance for 2026. Shares trade for 4x of currently depressed PBT and just 1.4x of their long term PBT guidance.

Watches of Switzerland (WOSG)

• A New WOSG Report from Broyhill Asset Management (Link).

Disclosure: This newsletter does not provide investment advice. Information presented is for informational purposes only and should not be considered a recommendation to buy or sell securities. The author may or may not own the securities discussed.