In 1999, Japan made a bet that it wouldn’t give up on an industry it once dominated.

NEC and Hitachi, rivals for decades, merged their memory divisions into a single company. Mitsubishi would join later. The name they chose said it all: Elpida, the Greek word for hope.

It was not just a name. In the 1980s, Japanese manufacturers dominated the global memory chip market. DRAM was a symbol of Japan’s industrial miracle. Precision, scale, and quality. Then came the Koreans, led by Samsung. With a greater appetite for risk and a willingness to lose money for years, they gradually took the market for themselves. By the end of the 1990s, Japan’s memory industry was cornered.

Elpida was born as a last line of defense. A national champion, built with pride, meant to restore a throne Japan once held.

It lasted twelve years.

DRAM, the memory found in every computer and smartphone, has a cruel characteristic. No one chooses the brand. No consumer pays extra for a specific memory chip. It is a commodity. The market sets the price. When supply exceeds demand, prices can fall below the cost of production. The only defense is having more scale and more cash than the next guy.

Between 2007 and 2011, everything went wrong at once. Global oversupply. The financial crisis. A strong yen that made every exported chip more expensive. And Samsung, which instead of pulling back during the downturn, doubled down. It used each crisis to put even more pressure on weaker competitors.

In February 2012, Elpida filed for bankruptcy with roughly $5.5 billion in debt. At the time, it was the largest bankruptcy of a manufacturing company in Japanese history.

Hope, it seemed, had failed.

In 2013, Micron, an American competitor that had also gone through difficult years, acquired Elpida for roughly $2.5 billion. It did not even have to pay the full amount upfront. Most of the purchase price would be paid over the following six years.

A single leading edge fabrication plant could cost more than that. Yet for $2.5 billion, Micron acquired the Hiroshima fab, a stake in a Taiwanese operation, an assembly and testing facility in Akita, and thousands of engineers. It bought decades of accumulated capability for less than it would have cost to recreate it.

At the time, no one treated it as a historic moment. The headlines were about bailouts, consolidation, and liquidation. Buying a bankrupt memory maker in 2013 was about as exciting as buying used office furniture.

In 2013, no one was talking about artificial intelligence. No one was talking about strategic memory shortages. No one saw Hiroshima as a critical asset for the next generation of computing.

Micron bought it anyway.

Fast forward to today. The Hiroshima facility is one of Micron's crown jewels. It is where the company produces its most advanced DRAM. It was also the site where Micron first introduced EUV lithography, the cutting edge of the industry.

The engineers who came over from Elpida became part of Micron's technical backbone. And the company is now investing billions of dollars in a second facility in Hiroshima, next to the original one, to produce the memory that powers artificial intelligence data centers.

The assets of a company that “ended” in 2012 became part of the foundation of one of the most important companies of the AI era.

The easy lesson from this story is that every failure is a success in disguise. It isn’t. Elpida’s shareholders lost almost everything. The failure was real. And for every Elpida, there are dozens of bankrupt companies whose assets were acquired for pennies on the dollar and still went nowhere.

The lesson, perhaps, is something else.

Companies die. What they build does not necessarily die with them. Factories, engineers, twenty years of accumulated knowledge. None of it disappeared in bankruptcy. It changed hands and kept evolving.

The facility that produced 25 nanometer chips in 2012 now operates at the frontier of lithography. The engineers who came over from Elpida helped develop generation after generation of new technology.

And while all of that was happening, the world was changing.

Computing became hungry for memory. What had once been a commodity became a strategic resource. The capabilities kept moving forward on one side. The world, eventually, moved toward them on the other.

Value emerged where those two paths crossed.

What does not seem valuable today may be indispensable tomorrow. The hard part is staying alive long enough to see tomorrow arrive.

No one tells you when tomorrow is coming, and waiting for it can be expensive. Elpida carried those assets through years in which they did little more than burn cash, and it failed before the payoff arrived. Holding an undervalued asset sounds easy in hindsight. In real life, it means surviving every day the market insists you are wrong.

There is another side to the story.

Micron bought at the bottom of the pessimism, when the seller had no alternative. The best time to buy an asset rarely coincides with the moment it feels obvious to buy it. In 2013, the acquisition looked like a footnote. The most important chapters of a story usually look like footnotes while they are being written. If they looked important, the price would be different.

In the end, the difference between Elpida and Micron was not foresight. Neither company predicted artificial intelligence. The difference was that one went bankrupt waiting for the future, while the other was alive when it arrived.

Elpida… The name meant hope. In a way, hope was fulfilled. Just on someone else’s balance sheet.

See you outside the index,

Leo Caroli

Two updates for today:

1) D-BOX Technologies (TSX: DBO)

D-BOX reported its FY2026 results. Let's take a look at what the company delivered and get to know the business a little better.

The numbers

• FY2026 revenue: CA$57.6M (+35% YoY) - record year

• Q4 revenue: CA$14.7M (+70% YoY) - strongest quarter in the company's history

• Adj EBITDA FY2026: CA$15.5M (+112% YoY), with a 27% margin (up from 17% in FY2025).

• Net new screens added: 189 (66 on Q4) - record

• Cash: CA$17.6M, debt: near zero

• NCIB in place

The flywheel

D-BOX designs, manufactures, and commercializes haptic motion systems for cinemas, sim racing, and training/simulation applications. Today, the company's primary focus is expanding its cinema footprint. Each new screen becomes a high-margin royalty asset, generating recurring revenue from every ticket sold with virtually no incremental cost. This year, total royalties grew 32% while the North American box office grew only 7%.

That gap kind of summarizes the thesis. The installed base is expanding so much faster than ticket sales can fluctuate that total royalties are outgrowing the box office by a wide margin. Growth is increasingly driven by screen count, not just by how many people went to see The Devil Wears Prada or whatever.

Exhibitors are treating D-BOX as a capital allocation priority to maximize yield per seat. In the CEO's words: "D-BOX is a highly deployable premium experience that can help maximize per-patron yield by delivering a uniquely immersive experience for audiences of all ages."

More screens → more royalties → more cash → more screens, with fixed costs diluted along the way. The model is self-reinforcing and the loop is working.

Changes in 2026

First, D-BOX announced in FY2026 its share repurchase program, up to ~9.5% of the float. The fact that management feels comfortable repurchasing stock while still adding screens at record pace says something about where they think the business is. The language in the release shifted to "disciplined capital allocation focused on creating long-term shareholder value".

Second, operating expenses grew 6% while revenue grew 35%. That's not cost-cutting. That's a business model revealing its true operating leverage. EBITDA margins went from 17% to 27% in one year.

What to watch

Not everything is clean. Sim racing and simulation both declined for a second straight year. Together they shrank from 44% to 28% of revenue. But that is somehow expected, as D-BOX is quietly undergoing a structural mix shift.

Q4 gross margins compressed 600bps because system sales (aka 'lower-margin hardware') dominated the mix, representing 53% of Q4 revenue versus ~12% a year earlier. Entirely justified by the strong pace of installations.

Looking into FY2027, I am following the pace of net screen additions, the growth of the royalty base relative to the box office, and the execution of the newly authorized buyback program, which has the potential to create meaningful shareholder value.

The bottom line

D-BOX underwent a complete management overhaul in 2025. Following a three-year activist campaign led by Daniel Marks, the board was replaced and a new CEO and CFO were appointed within months of each other. The strategic shift was decisive. Rather than prioritizing hardware sales, management refocused the entire business on its royalty engine.

The results have been hard to ignore. The key question is whether the record addition of 189 screens this year represents a new baseline or a peak. The opportunity suggests substantial runway ahead. North America alone has roughly 42,000 cinema screens. Globally, there are more than 200,000. D-BOX is installed in just 1,200 of them, which is less than 1% of the global market.

At ~13x ttmEBITDA, the market does not seem to be pricing a royalty engine compounding at 25%+. And there's also a tax shield, meaning D-BOX won't pay taxes for what I think is a little more than a couple years. Bad for the government, good for shareholders.

The most common concern is customer concentration. Cinemark appears to represent around half of all active screens and the bulk of recent additions. If that single relationship stumbles, the growth story stalls. I'm not very worried about that. D-BOX delivers strong ROIC for exhibitors, so theaters have real economic incentive to keep and expand the partnership. And the company is actively diversifying. On June 3rd, D-BOX announced a new partnership with B&B Theatres. Still, landing a second major chain like AMC would transform this from a concentrated relationship into a genuine platform. Until then, it's worth acknowledging the risk.

All in all, FY2026 was a strong year for D-BOX and further evidence that the transformation is working.

2) Tasmea $TEA.AX

Tasmea is an Australian specialist industrial services company that owns and operates multiples services businesses across mining, energy, infrastructure, water, and defense. Think of it as a serial acquirer of essential maintenance and electrical contractors. A seriel acquirer of small, owner-led businesses bought at modest multiples and run under a decentralised model. Tasmea is a nice company led by its competent co-founder, Stephen Young.

Well, this nice company made two announcements in the same week, and together they tell a clear story about where it is heading.

On June 2nd, the company announced the acquisition of Maxim Group, a specialist electrical contractor, for up to A$254 million (~5.4x EBIT). Maxim brings direct exposure to three of the fastest-growing infrastructure markets in Australia: data centres, battery energy storage and major government infrastructure. The business has grown revenue at ~70% CAGR over the past two years and carries a A$1.3 billion identified pipeline. Post-deal, Tasmea’s Electrical segment alone is expected to generate approximately A$100 million in EBIT. Forecast EPS accretion of 31% and net leverage stays below 1.0x. Very good deal, if you ask me.

Two days later, Tasmea declared a fully franked special dividend of 10 cents per share, with its Dividend Reinvestment Plan (DRP) available at AU$6.85 per share. At the same time, Tasmea reaffirmed its guidance and emphasized that the special dividend would not compromise growth initiatives or acquisition capacity. The sequence is telling. Just days after announcing the A$254 million acquisition of Maxim Group, Tasmea returned capital to shareholders while maintaining its growth outlook.

I thought the Maxim acquisition was excellent, and the market appeared to agree, sending the shares up more than 15% on the announcement. The stock has already nearly doubled this year and is now trading around all-time highs. Even so, the valuation may not prove demanding if Tasmea can continue acquiring high-quality earnings at attractive multiples as it did with Maxim.

Share

CLIQUE AQUI PARA LER EM PORTUGUÊS

Website: https://www.smart-shooter.com/
Ticker: $SMSH (TASE)

The other day, the U.S. Army awarded another defense contract. Normally, that wouldn’t be particularly noteworthy. This one, however, caught my attention.

It’s a $10.7 million contract awarded to an Israeli company that most people have never heard of: SmartShooter, based in Kibbutz Yagur, near Haifa.

What makes it interesting is that this wasn’t a first order. It was a follow-on contract. The buyer and seller already knew each other. The product had already been purchased. It had already been tested. And now the customer wanted more.

That detail matters.

In the defense industry, repeat orders are one of the strongest endorsements a company can receive. Products that fail in the field rarely get a second chance. There is no marketing campaign more powerful than a military customer returning to buy again.

So what exactly does SmartShooter do?

SmartShooter specializes in intelligent fire-control systems for small arms. Its flagship product is the SMASH family of systems, an AI-powered smart sight that helps soldiers engage targets with far greater accuracy.

The concept is straightforward. The technology behind it is not.

The SMASH system can be mounted on virtually any rifle in a matter of minutes. Once installed, it uses computer vision and image processing to identify and lock onto a target. From that point on, the system automatically tracks both the target’s movements and the shooter's movements, automatically adjusting the aim in real time. When engagement is authorized, the system releases the shot at the optimal moment to maximize hit probability.

In tests involving moving targets at 100 meters, conventional red-dot sights delivered hit rates of roughly 20%, even in the hands of experienced operators. The SMASH system, by contrast, achieved hit rates of approximately 80%, with novice users performing nearly as well as seasoned professionals.

In practice, SMASH narrows the gap between an average soldier and a highly trained marksman. It’s no coincidence that the company’s motto is “One Shot, One Hit.” The technology is protected by patents and was developed over more than a decade in close collaboration with the Israel Defense Forces.

A Brief History

SmartShooter traces its origins to the Second Lebanon War in 2006. Michal Mor was then leading R&D teams in Rafael’s missile division, helping develop some of Israel’s most advanced military technologies. Yet when her husband, a reservist in the Golani Brigade, was called up and sent into Lebanon, he did so armed with nothing more than a standard infantry rifle.

Looking at that rifle, Mor had a realization. She had spent years at Rafael developing guided missile systems, where precision was everything. The question she kept coming back to was simple: why couldn't the same technology that guides a missile be used to guide a bullet?

Later that year, Mor left Rafael alongside Avshalom Ehrlich, a software engineer who had been part of her team. Over time, the two decided to turn the idea into a company of their own. In May 2011, they founded SmartShooter in Kibbutz Yagur, a small community in northern Israel.

Their first meeting with the Ministry of Defense featured a makeshift prototype: a laptop connected to an airsoft rifle. Three years of development condensed into a device sitting on a conference table, only to be met with a blunt verdict “it would never work”. But when Mor explained the concept to missile engineers at MAFAT, they saw potential.

Eventually, the Golani Brigade became the first unit to test the system. Soldiers initially disliked the idea of a computer deciding when their rifle could fire. But after seeing the system's accuracy firsthand, the skepticism quickly faded.

What followed were years of market education. There was no established market for smart weapon sights because the category itself did not yet exist. Before SmartShooter could sell its product, it first had to convince militaries that such a product should exist at all. The company spent years conducting demonstrations around the world, navigating lengthy evaluation programs, and steadily building credibility.

The turning point came with two events few could have predicted: Russia’s invasion of Ukraine in 2022 and the Israel–Hamas war in 2023. Both thrust the drone threat to the forefront of military planning. In Gaza, SMASH was deployed against attack drones in dense urban combat, giving the system the kind of real-world validation that no controlled test environment could provide.

SmartShooter's expansion accelerated. The company grew from a niche supplier into a business with operations spanning more than 25 countries, while establishing subsidiaries in the United States, Germany, and Australia. In March of this year, it went public on the Tel Aviv Stock Exchange, raising approximately NIS 200 million.

SmartShooter Today

SmartShooter is arguably in the strongest position of its history. Revenue grew an impressive 50% in 2025, with significant operating leverage as the business scaled. Following its IPO, the company is well capitalized, carries no debt, and generated positive cash flow. Meanwhile, new orders continue to arrive:

New customers have significantly diversified SmartShooter’s revenue base. In 2024, Israel’s Ministry of Defense accounted for 47% of revenue. By 2025, that figure had fallen to 22%. Over the same period, the United States emerged as a meaningful market, contributing 22% of revenue, while Europe accounted for 24% and Asia-Pacific for 13%. SmartShooter has evolved from a business heavily dependent on a single domestic customer into a geographically diversified operation spanning more than 25 countries, including:

• Americas: United States

• Europe: United Kingdom, Germany, the Netherlands, Sweden, and others

• Asia-Pacific: Australia, India, Japan, Singapore, and Thailand

• Middle East & Africa: Israel’s Ministry of Defense and several African countries

SmartShooter operates through two primary product lines, both leveraging the same underlying computer-assisted fire control architecture.

1- Small Arms Systems:

Designed for the individual soldier:

• SMASH 3000

• SMASH X4

These systems are designed to increase accuracy, lethality, and engagement effectiveness against moving targets and small drones.

2- Remote Weapon Systems:

Designed for fixed and remotely operated weapon platforms:

• SMASH HOPPER

• SMASH HOPPER MG

Designed for border security, critical infrastructure protection, and remote operations requiring high levels of precision.

In addition, the systems can be integrated into larger military architectures.

The $10.7 million contract mentioned at the beginning of this post is for the SMASH 2000L, a variant of the SMASH 3000. Delivery is scheduled for the third quarter of 2026 and follows a string of recent awards from the U.S. military:

1. U.S. Army (May 2025): PEO Missiles & Space selected a brigade-level integrated C-UAS solution as part of the Joint Counter-Small Unmanned Aircraft Systems Office (JCO) program.

2. U.S. Marine Corps (July 2025): Awarded a contract for a portable dismounted C-UAS solution integrated with remote weapon stations (RCWS), the TAK battlefield network, and radar systems.

3. U.S. Air Force (March 2026): Issued multiple orders for base protection and air-defense missions utilizing SmartShooter's C-UAS solutions.

The pace of these awards highlights the scale of the opportunity emerging in the counter-drone market. As inexpensive drones have become one of the most disruptive technologies on modern battlefields, militaries around the world have been forced to accelerate investments in C-UAS capabilities.

Some industry estimates suggest the market could exceed $10 billion by 2030. Regular readers know I’m generally skeptical of TAM projections, especially those stretching years into the future. In this case, however, I don’t need to believe the exact number. It is enough to recognize that both the direction and the velocity of demand in the C-UAS market appear highly favorable.

SmartShooter currently trades at a market cap of approximately $360 million. Adjusting for the post-IPO balance sheet, enterprise value is likely closer to $280 million. Against 2025 EBITDA of roughly $7 million, the stock does not immediately appear cheap.

That, however, is a rearview-mirror assessment. The pace of recent contract awards and the company's operating leverage suggest it is not unreasonable to envision a 2026 revenue base of $55 million with EBITDA margins in the 20%-25% range.

More importantly, very few investors appear to be paying attention. SmartShooter only recently went public and remains listed exclusively on the Tel Aviv Stock Exchange. It has no meaningful sell-side coverage and little institutional ownership.

Yet the company already operates across more than 25 countries. Its products have been tested in real combat environments, validated through repeat orders from the U.S. military, and positioned within one of the fastest-growing segments of the defense industry.

Whether the shares are attractive ultimately depends on how much of that future is already reflected in the price. What seems harder to argue is that SmartShooter is just another lousy micro-cap. It definitely is not.

To wrap up

SmartShooter appears to be one of the rare companies that has effectively created a new product category. The concept of a "computerized rifle sight" had been attempted before (TrackingPoint being the most notable example), but SmartShooter's approach is different: an external, adaptable system for small arms that combines target tracking with onboard fire-control logic. More importantly, it is a solution that has demonstrated real-world adoption on the battlefield.

Behind the company is a team drawn from Israel's leading defense contractors, including Rafael, Elbit, and Israel Aerospace Industries. SmartShooter benefits from intellectual property protection, operates in a market with favorable long-term tailwinds, and benefits from a period in which defense spending is rising across much of the world.

At the same time, investors should not overlook the risks. SmartShooter is a newly public company with limited liquidity, a short profitability record, and the ever-present possibility that future technological advances alter the competitive landscape.

What do you think?

See you outside the index,

Leo Caroli

Share

Good afternoon, everyone!

The past few days in the market have produced some remarkable stories. Perhaps none as striking as Allbirds.

Allbirds was Silicon Valley’s favorite sneaker. Eco-friendly footwear, a 2021 IPO, a $4 billion valuation. The kind of hype we’ve seen many times before. Even Leonardo DiCaprio invested in this Titanic.

Four years later, it turned to dust. Down 99.5% from its peak. Stores closed, operations dismantled, and the brand itself sold for $39 million to American Exchange Group.

Up to that point, a fairly standard story of excess and correction.

But then comes the twist.

The company rebrands. NewBird AI. Yes, AI. It raises $50 million to buy GPUs. The new long-term vision: become a GPU-as-a-Service company. And, as if that weren’t enough, the stock surged more than 800% in a single day on the news, with trading volume 875 times above normal.

A delirious reaction that would make Fama proud.

Not every corporate transformation arrives with that kind of immediate impact. Some are quieter. Small changes that compound over time until, before you realize it, the company no longer resembles what its industry classification suggests.

That’s the case with Mestek, a manufacturer of heating, ventilation, and air conditioning (HVAC) equipment that, as we’ll see, has built one of the most unusual balance sheets in the market.

Website: https://www.mestek.com/
Ticker: $MCCK

Mestek is headquartered in Massachusetts and operates twelve manufacturing facilities. Its two main business segments are HVAC equipment and metal forming equipment.

Mestek’s story is deeply intertwined with the Reed family. Beginning in the 1950s, John Reed steadily built the business through a series of acquisitions across the HVAC sector. That long expansion reached a turning point in 1986, when Reed National Corporation capitalized on the bankruptcy of Mesta Machine Company to acquire its most profitable divisions, laying the foundation for what would become Mestek, Inc.

Mestek was once listed on the NYSE but deregistered in 2005 after spinning off its subsidiary Omegaflex. The move seemed part of a plan to take the company private, but that never fully materialized.

John Reed passed away in 2013, at which point his son, Stewart Reed, took over. It didn’t take long for Stewart to leave his mark on the business. In 2015, Mestek began building a commodities investment portfolio, starting with platinum. In Stewart’s own words, the trading and commodities investment account was “the last vestige of my teenage lone ranger years”.

Seeing a newly anointed CEO (and the founder’s son) deploy company cash to speculate in commodities, echoing his teenage “lone ranger” days, is hardly a reassuring signal for investors. That concern was especially acute among long-time shareholders, who still remembered Carecentric, Mestek’s ill-fated attempt under John Reed to diversify into healthcare IT.

But at the end of the day, the core business remained profitable and cash-generative. And with Stewart as controlling shareholder, CEO, and chairman, investor opinion didn’t seem to keep him up at night.

Starting from a small base, the commodities portfolio expanded quickly, reaching meaningful scale within a decade. The company also holds cash and U.S. Treasuries. In total, Mestek’s financial assets now exceed $400 million, roughly twice its total liabilities. Balance sheets like that are rare.

Meanwhile, the HVAC business generated nearly $40 million in operating profit last year. A historically profitable operation dating back to the 1940s, now paired with a balance sheet that increasingly resembles an investment vehicle.

In the end, Stewart has built something unusual. At $65 per share, Mestek trades at roughly book value. With just under 8 million shares outstanding, its market cap is about $500 million. Stripping out the commodities portfolio, what remains is a long-standing HVAC business valued at roughly 4x EBIT.

The right mental model today is to view two businesses under one roof: a 10-year-old commodities trading company and an 80-year-old HVAC company. On a consolidated basis, the performance of one can easily obscure the other. In recent years, rising gold prices have tilted the balance toward the commodities side. As a result, Mestek has built an NCAV of roughly $370 million.

The usual caveats apply. Despite sitting on a large pool of financial assets, Mestek has no track record of dividends or share buybacks.

Stewart Reed controls roughly 70% of the company and has a clear appetite for commodities speculation. The stock is genuinely illiquid, and the company has filed delay notifications for its last two annual reports, which is never a particularly comforting sign.

Still, this is a cheap small cap backed by substantial real assets, with a long-standing and consistently profitable core business, and largely off the radar of major institutions.

The Reed family has built something peculiar. Possibly brilliant. Clearly risky. Undeniably unusual.

Best,

Leo Caroli

Share

Good morning, folks!

It’s been a while since we last talked about Hong Kong. Since 2024, to be precise. Longtime readers might remember when we ignored the buzz around one of those flashy “Super Wednesdays” to introduce a toy manufacturer instead. That turned out to be a good call. Our old friend Dream International nearly quintupled—though it later gave back a good chunk of those gains.

Today I’m heading back to the Hong Kong exchange. I’m not suggesting lightning strikes the same place twice. But the truth is that, over the past year, I’ve been seeing more interesting opportunities popping up across Asia. And if that’s the case, why not write about Asia? Perhaps guided by a mild dose of superstition, I figured I’d return to Hong Kong.

The company we’re looking at today is listed there, yet it rarely comes up in investor conversations. Still, it has more than forty years of history.

And right now, it may be entering its best phase yet.

Meet Kwan Yong Holdings.

The business

Website: https://www.kwanyong.com.sg/
Ticker: $9998.HK

Founded in 1984 by Kwan Mei Kam and Tay Yen Hua, Kwan Yong began as a small local contractor in Singapore.

Over the years, the company evolved into a specialist in institutional construction. Its core projects include schools, hospitals, and nursing homes.

In 2020, Kwan Yong went public on the Hong Kong Stock Exchange to increase its visibility, gain access to international capital, and support its expansion. Today, the company is on track to reach around SGD 280 million in annual revenue, with key clients including Singaporean government agencies such as the Ministry of Education and the Ministry of Health.

The story of Kwan Yong since going public can be summarized in five acts:

2020–2021: The COVID shock

The company’s IPO coincided with one of the worst possible moments for the construction sector. The pandemic halted projects, drove up material costs, and restricted the flow of foreign labor into Singapore. During this period, Kwan Yong played defense. The company cut costs, preserved cash, and relied on government subsidies to weather the crisis.

2022–2023: Gradual normalization

As restrictions eased, the sector began to recover. Projects gradually resumed and the industry pipeline started to regain visibility. Kwan Yong reached its lowest revenue levels during this period, but managed to restore its gross margins.

2023: Recovery

Construction activity in Singapore accelerated, driven by a strong pipeline of public projects. Kwan Yong took advantage of the momentum. Revenue resumed growth, profits inflected, and the company’s backlog nearly doubled.

2024: An operational stumble

Despite the favorable sector backdrop, one specific project did not go as planned. Kwan Yong had to recognize a provision of roughly SGD 10 million, which pressured margins—serving as a reminder of the operational risks inherent in this type of business.

2025: Boom

Construction activity accelerated again in Singapore, and Kwan Yong rode this new expansion cycle. Revenue grew 80%, margins doubled, and cash more than doubled, leading the company to announce a dividend of HKD 0.02 per share.

In 2025, Kwan Yong also completed Kempas Residences, its first public housing project. In recognition of the successful execution, the company received the HDB Construction Award.

That same year, the company signed the largest contract in its history: a SGD 201 million project with the Ministry of Health to develop a polyclinic and a nursing home.

On the one hand, the IPO came during a difficult period for the sector. On the other, it allowed Kwan Yong to weather the storm well capitalized. With the recent upturn, the company has reached a record level of cash. It’s worth noting, however, that this cash position has been largely driven by advances from newly signed contracts. In other words, a good portion of this money will be used to fund the execution of projects. In the meantime, Kwan Yong benefits from this working capital dynamic, earning financial income along the way.

The 1H26 results (the half-year ended December 2025) continued to reflect these favorable tailwinds. Cash climbed to SGD 149 million, while the backlog reached SGD 715 million.

With a market cap of HKD 440 million (~SGD 72 million), Kwan Yong trades at roughly half of its cash balance and, by my estimates, around 4× FY2026 earnings—a combination of growth, profitability, and valuation that is hard to ignore.

Management

Kwan Yong is essentially a family-run business, led by its founders for decades.

The company is headed by Mr. Kwan Mei Kam, its Chairman and CEO. He founded the business a few years after graduating in Building from Singapore Polytechnic, and also holds a degree in Construction Management from Central Queensland University in Australia.

Working alongside him is his wife, Ms. Tay Yen Hua, who oversees administrative operations and human resources. Now in their early seventies, the couple has worked together since the company’s early days. Through the holding company Ideal Smart Ventures, they control roughly 75% of Kwan Yong.

Another family member involved in the business is the couple’s daughter, Ms. Kwan Shu Ming. A graduate in Accounting from Curtin University of Technology and a member of CPA Australia, she leads the group’s finance and accounting functions.

Rounding out the executive team is Mr. Jacob Wong San Ta, a civil engineer from the National University of Singapore. With the company since 1995 and prior experience in the public sector, he oversees the Tender and Contract, Purchasing, and Project departments.

Risks

The main risk facing Kwan Yong is tied to the industry it operates in. As with any cyclical business, the question is not if the market will slow down, but when that slowdown will arrive.

For now, the trend still looks healthy. The Building and Construction Authority (BCA)—Singapore’s government agency responsible for regulating and developing the construction sector—estimates that demand for construction contracts in 2026 will range between SGD 47 billion and SGD 53 billion.

In 2025, about SGD 50.5 billion in contracts were awarded, up from SGD 44.6 billion in 2024. This growth has been driven primarily by institutional and public housing projects.

Over the medium term, between 2027 and 2030, projected annual demand is expected to range between SGD 39 billion and SGD 46 billion.

Despite the optimistic outlook, the sector will always remain sensitive to macroeconomic factors. Raw material costs, labor availability, financing conditions, and potential geopolitical tensions affecting global trade can all weigh on the pace of new projects.

An additional risk concerns leadership succession, as Mr. Kwan is already in his seventies.

To wrap things up

Kwan Yong is another example of the small construction companies benefiting from the current strength of Singapore’s construction sector. After surviving the most difficult years, the company now finds itself with a growing backlog, a healthy balance sheet, and a rather modest price.

That modest price is not exactly a mystery. This is a small, controlled, illiquid company with virtually no analyst coverage. Being listed in Hong Kong also limits access for many retail investors, as some brokers simply don’t offer the market. On top of that, it operates in a naturally cyclical industry.

Perfect investments are rare—perhaps even nonexistent.

In the end, the investor’s job is to decide which risks are worth taking, and at what price.

Which brings us to the question: is Kwan Yong’s price attractive enough to compensate for its risks?

As always, thank you for your time!

Best,

Leo Caroli

Share

Ticker:$BEOB
Website: beobank.com

Good morning, folks!

Today we’re going to talk about a bank. Yes, a bank. It might be the first time we’ve done that around here. And no, this isn’t another deep dive into Banco do Brasil’s earnings. Relax. As you know, we prefer simple things.

And when it comes to banks, simplicity is a virtue. A good bank is a boring bank. It takes deposits here, lends money there. Earns its spread, keeps a safety cushion, and goes to sleep. The next day, it does it all over again. No complicated derivatives, no attempts at financial genius.

Today’s bank fits that description. Practically a provincial bank. Meet Bank of Eastern Oregon.

Bank of Eastern Oregon (BEO) was founded in 1945, just as World War II was drawing to a close. While American troops were preparing to cross the Rhine, a different problem was unfolding in rural Oregon: the community lacked a bank. In Gilliam County, a group of farmers decided to solve the problem themselves. Each put up a bit of capital. No grand vision. No Wall Street ambition. Just a practical need and a local solution. That’s how Gilliam County Bank was born. A few years later, the name changed to Bank of Eastern Oregon.

BEO serves rural communities across eastern Oregon, eastern Washington, and western Idaho. It’s safe to say that large Wall Street banks aren’t lining up to serve these regions. Today, BEO operates just over 20 branches. In some counties, it’s the only bank in town.

That’s the quiet beauty of a community bank. Small on the balance sheet, large in importance. Without institutions like this, many small towns would simply be underserved. And that local presence becomes a competitive advantage. What large national bank wants to stretch across the country to chase a small rural market? Who wants to start a rate war with a bank that’s been embedded in the community for eight decades? A bank whose name is tied to the grandfather’s farm loan? There are easier and more glamorous battles to fight.

Bank of Eastern Oregon, as expected, has predominant exposure to the agricultural sector. Its loan portfolio shows balanced duration and decent repricing flexibility, a valuable trait when navigating shifts in interest rate cycles.

Concentration in agriculture naturally brings the macro risks we know well in Brazil. Unpredictable weather, commodity price volatility, rising input costs... It’s an environment shaped by forces that management simply cannot control.

That said, BEO has been operating in this landscape for more than eighty years, a track record that’s hard to dismiss. And judging by the quality of its assets, management appears to know what it’s doing. BEO closed 2025 with an excellent non-performing assets to total assets ratio of just 0.33%:

It appears that BEO went through a period of higher stress between 2017 and 2019. Since then, the bank has executed a cleanup process. Non-performing assets have declined steadily, and net charge-offs have turned negative. In recent years, the loan portfolio seems to be under control. Even so, the bank hasn’t become complacent. It maintains a tier 1 capital ratio of 10.7%, a good cushion for a bank of its size. This is good practice. Prudence, in banking, compounds quietly over time.

Deposits have grown at a 10.9% annual rate over the past eight years. The period between 2019 and 2021 was particularly strong, likely driven by pandemic-related relief programs as well as expansion efforts. In 2019, for instance, BEO expanded its footprint by acquiring branches from Umpqua Bank, reinforcing its presence in rural communities across eastern Washington

Expanding into adjacent rural markets appears to be a sensible strategy for BEO if it aims to pursue growth. The demographic realities of eastern Oregon itself impose natural constraints: an aging population, modest economic growth, and median household income at roughly 80% of the state average. In other words, it’s not exactly a territory known for explosive organic growth.

Banks like BEO are built on long-standing customer relationships, and that makes a tangible difference in the stability of their deposit base. Roughly 30% of deposits are noninterest-bearing, which helps keep funding costs low. Historically, BEO has demonstrated good control over its funding, an advantage that comes from operating in niche markets where rate wars are largely unnecessary.

BEO operates with a solid 5.2% NIM and has demonstrated the ability to navigate different interest rate cycles while preserving a healthy spread. Putting the pieces together — deposit stability and cost control, loan mix, balanced duration, repricing capacity, low NPA levels, and adequate capitalization — what emerges is a niche bank with a resilient franchise and capable management.

Anyone looking to invest in Bank of Eastern Oregon will, in practice, be buying its holding company, BEO Bancorp. This structure is common among community banks and can sometimes conceal a few hidden traps. My preference is to look at the core banking operations through the subsidiary’s call report. Still, investors shouldn’t ignore the holding company. Leverage and complexity can create risks that don’t show up at the bank level. Here, BEO Bancorp is relatively straightforward. Bank of Eastern Oregon is its only operating subsidiary, and the holding carries roughly $8 million in subordinated debentures.

Decent operational execution — reflected in an efficiency ratio in the low 60s — has translated into a solid profitability profile for this small Oregon bank. BEO generates a 1.73% return on assets and a ROE above 16%, compelling quality markers for an agro-focused community bank. Over the past eight years, the bank has compounded tangible book value at roughly 12% annually.

Today, Bank of Eastern Oregon trades at ~8x earnings and about 1.3x tangible book value. These are reasonable multiples for an institution with the profile we’ve just examined. BEO doesn’t promise to revolutionize banking. It doesn’t talk about disruption or reinventing money. Still, the value created hasn’t gone unnoticed. Since the beginning of last year, the stock is up nearly 50%. Over the past decade, it has delivered a CAGR above 18%. Not bad for a rural bank that sticks to the basics. Basics overseen by Jeff Bailey, who has been with the bank since 1996 and has served as CEO since 2009.

BEO has a market cap of ~$133 million, with approximately 21% insider ownership. That said, the shares are extremely illiquid, making position building an exercise in patience and introducing an additional layer of risk.

In any case, it wouldn’t be prudent to build a large, concentrated position in a bank like BEO. With community banks, it often makes more sense to think in terms of allocation and assemble a basket of conservative institutions with strong asset quality, reasonable valuations, and diversified risk exposure.

But, as always, each investor knows where their own constraints lie.

PS: By the way, if you’re interested in banks, it’s worth checking out the Dirt Cheap Banks substack. I came across it recently and found it excellent.

Best,

Leo Caroli

Share

*This piece was originally published in portuguese back in 2022, when almost nobody was reading Negócios Pelo Mundo.

Every now and then, YouTube’s algorithm sends you on one of those unexpected detours, the kind that pulls you toward videos you’d never have discovered on your own. That’s exactly what happened to me. A few minutes into dutifully following the recommendation chain, I landed on something that seemed to have nothing to do with where I started. Or so I thought.

It was a somewhat stiff reporter doing a lighthearted segment about Rio de Janeiro during the very first Rock in Rio. 1985! I found myself genuinely entertained, watching something that felt both familiar and distant at the same time. The same elements, just in a completely different context.

And that gave me an idea for how to wrap up my Sunday: pull up the IPO prospectus of a high-growth tech company. Also in a completely different context.

That’s how I ended up reading the Microsoft 1986 IPO prospectus. The link is here:

Microsoft - Prospecto IPO 1986

247KB ∙ PDF file

Download

Download

The first surprise is that the prospectus is only about 50 pages long. Early on, it lays out a handful of milestones. It’s fascinating to look back at them:

1980/1981:
They began developing systems for IBM, which started selling PCs running MS-DOS.

1982:
International expansion into Europe.

1983:
Launch of Microsoft Press to publish and sell computer-related books.
Launch of the Microsoft Mouse.
Launch of Microsoft Word, with a fun highlight in the document describing it as “graphics-based.”

1985:
Release of Microsoft Excel for the Apple Macintosh.
Release of Microsoft Windows, described as “a graphical operating environment that runs on the Microsoft MS-DOS operating system.”

Microsoft was founded in 1975, which means the company was already 11 years old at the time of its IPO. It already had products, revenue, profitability, and cash generation when it decided to go public. Take a look at this passage from the prospectus:

“Since its inception, the Company has funded its activities almost entirely from funds generated from operations.”

Bill Gates gave an interview to Harvard in 2013 where he addressed this exact point. I’ll include the transcript below:

David Rubenstein: “When you started (Microsoft) how did you avoid what so many other entrepreneurs are unable to avoid, which is taking venture capital money and diluting themselves, you didn’t really take venture capital money, how did you avoid that?”

Bill Gates: “Software, if you can sell it in volume, has extremely favorable economics because the costs are quite finite, and you can scale the volume, in our case, to sell millions of copies – every Apple II Computer, Commodore PET, IBM PC had copies of our software, so our income went up a lot. We didn’t have to build any factories, we were cash positive, and we didn’t have a year where we lost money.

The first few years Paul (Allen) and I worked for free, but after that we were just generating cash. We eventually gave away, or sold, 5% of the company for a million dollars at a 20 million dollar valuation, just to get a venture capital company to join our board and give us some adult advice about various things, which was quite helpful. We picked one in the valley, a guy named Dave Marquardt came on our board and did a fantastic job.

That money sat in the bank, and it’s still in the bank today, so it was not for anything to do with capital, but rather just to join the team... he’s still on the board of Microsoft and is still extremely helpful as a lot of important decisions get made.”

Quite a difference from what we’ve gotten used to in recent years, isn’t it? And just for the record, Dave Marquardt stepped down from Microsoft’s board in 2014.

But back to the prospectus, let’s look at a few numbers:

Microsoft was genuinely a healthy company when it went public. Since 1982, its growth had compounded at roughly 80%. Growth had indeed slowed from the triple-digit pace of earlier years, but even ~40% is hardly something to dismiss. At the same time, margins kept improving as the revenue mix moved further away from hardware and toward software.

On top of that, Microsoft was extremely well capitalized. It went public with zero debt, and because it was generating substantial cash, it was essentially piling money onto its balance sheet:

In 1986, Microsoft grouped its revenue into three main product categories:

1. Systems (54%)

This included its operating systems (MS-DOS, XENIX, and MSX-DOS) as well as programming-related products such as Microsoft’s own compilers.

At the time, Windows had only just been released and was still a major unknown. The prospectus even included the following remark:

“It is too early in the life of Microsoft Windows to determine what level of acceptance it will attain in the marketplace.”

It’s amusing to read that now, knowing everything we do about what Windows eventually became.

It’s also interesting to see how Microsoft encouraged developers to build applications for Windows, emphasizing the practicality of the graphical interface for users.

2. Applications (38%)

This category included Microsoft software such as Word, Chart, File, and — to the delight of (almost) all of us — the newly released Microsoft Excel.

3. Hardware and Books (8%)

This category consisted mainly of the Microsoft Mouse and books covering Microsoft products, software development, and computing in general. After all, no one had Google back then to look things up.

So if you wanted to learn more about the IBM PC, you had to buy a book like “The Peter Norton Programmer’s Guide to the IBM PC.” And Microsoft actually published and sold materials like that.

Since this was also a time before the cloud existed, Microsoft distributed its products through OEMs and through retail channels.

OEMs were manufacturers such as IBM, Compaq, HP, AT&T, and Olivetti. Microsoft had arrangements with them in which the OEM generally paid the company based on the number of copies of the product it produced or the number of microcomputers it shipped, with an initial minimum commitment fee. In addition to that, Microsoft worked with independent distributors that sold the company’s products and provided sales support and training to large customers.

Anyone considering this IPO also had to be fully comfortable with the people in charge. Bill Gates and Paul Allen were still major shareholders, holding a large stake in the company:

How much would you be willing to pay for a software company with more than a decade of operating history, 40% growth, no leverage, strong cash generation, 30% operating margins, an ongoing international expansion, no customer concentration, and a 30-year-old fanatic founder like Bill Gates at the helm?

Well, this company’s IPO priced at just $21 per share, roughly 3.7x sales and about 15x fwd P/E. Quite a contrast with the tech IPOs we’ve seen over the past few years, isn’t it? Not to mention the SPACs…

In the end, it’s the same element — a tech IPO — simply placed in a completely different context: time.

Thanks for reading!

PS: One historically interesting detail in the prospectus is the reference to the famous lawsuit brought by Seattle Computer Products against Microsoft over the $50,000 purchase of Q-DOS, the operating system that would eventually give birth to MS-DOS.

Share

Good morning, folks!

I suggest everyone reads today’s post. Of course, I suggest everyone reads all posts. But today’s, in particular, brings one of those rare episodes that, paradoxically, happen more often than they should.

I also took the chance to share a few reflections I don’t normally bring into the Radar. However, these are reflections that deserve to surface when circumstances call for them. And in one of our cases today, the circumstances demanded it.

Anyway, here you go:

1) Trubar Inc $TRBR.V

Who remembers the Trubar case? In short, a fast-accelerating CPG with the optionality of a takeout on the horizon.

Regarding the acceleration, our attention was fixed on signs of heavy reliance on aggressive subsidies. As we wrote in Radar 16:

“The growth engine continues to be marketing and aggressive incentives. Wherever we look—vendor discounts, gross margin, marketing expenses—we are left with the impression that expansion is only sustained through strong subsidies.”

As for the possibility of a sale, our impression was that management was not only open to an exit but actively preparing the ground for it. And that it could come with a nice premium:

“Furthermore, Trubar even posted this table with recent M&A activities in the sector. A clear demonstration of management’s willingness to seek a future exit.” Radar 12

“There is still clear room for TRUBAR to capture a good premium in an exit, but until that happens, it would be interesting to see the company manage to better balance growth and profitability.” Radar 16

Well, the sale happened!

Trubar announced its sale agreement to the Turkish giant ETİ Gıda for C$1.64 per share in cash. A 64% premium over the last closing price and over 100% since our last update. The expectation is that everything will be done by the first quarter of 2026, resulting in a profitable outcome for shareholders.

Now, a small confession. I tried a Trubar bar a few days ago. I didn’t like it. But I’ve long noticed that my palate is a bit like a film critic who’s completely out of touch with the audience.

So I’ve learned my lesson. My personal taste deserves very little weight when I analyze this kind of business. It gets in the way. It creates noise. What matters far more are things like distribution, velocity, and consumer uptake. The streets tell truths my palate insists on hiding from me.

Anyway, case closed here, and congratulations to all Trubar shareholders!

2) Intellego Technologies $INT.ST

I don’t know if Sweden produces soap operas the way Brazil does. But if they don’t, they could easily start by using Intellego as inspiration. Intellego in 2025 was pure dramaturgy, the kind that makes the audience laugh, cry, and faint all at once.

There was euphoria, collapse, a shootout between bulls and bears, and, as in any respectable tragedy, police banging on the door and handcuffs gleaming in the light of a major scandal. A wild saga, great to follow.

I mentioned Intellego for the first and only time here back in early April. After that, the stock more than tripled. And, defying the gospel of social-media engagement (the one that preaches repeating booming tickers ad nauseam), I chose to leave it alone. My only other comment came now, in November:

And when I said “far from over,” I confess I had no idea what was waiting just around the corner. Less than a week after that tweet, Intellego’s CEO was arrested.

Yes, arrested!

In light of this latest development, it’s worth retracing the recent history of a name which, whether we like it or not, ended up becoming one of the main characters of the year for global small-cap investors.

So let’s walk through what happened, month by month:

APRIL (+95.2% in the month)

• Preliminary Q1 2025 Results:

  • Revenue > SEK 200 million (+150% vs Q1 2024)

  • EBIT > SEK 100 million (+138% vs Q1 2024)

• EBIT target of SEK 160 million for 2025 reached in just 4 months.

MAY (+7.12%)

• Q1 2025 Report released:

  • Revenue: SEK 201 million (+152% vs Q1 2024)

  • Net Profit: SEK 94 million (+165% vs Q1 2024)

  • CFO: SEK 28 million (vs SEK 34 million Q1 2024)

• Guidance Revision:

  • Revenue: > SEK 600M (vs previous 265M)

  • EBIT: > SEK 250M (vs previous 102M)

  • In other words, +126% revenue and +145% EBIT YoY.

JUNE (-5.2%)

• New CFO: Hans Denovan, starting in September. Goal is to prepare Intellego for accelerated internationalization and a possible uplisting.

• Warrant exercise ~SEK 34.6m.

JULY (+69.9%)

• Preliminary Q2 2025 Results:

  • Revenue > SEK 200 million

  • EBIT > SEK 100 million

• Revised EBIT target of SEK 250 million for 2025 already surpassed at the beginning of Q3.

AUGUST (+53.9%)

• Q2 2025 Report released:

  • Revenue: SEK 217 million (+297% vs Q2 2024)

  • Net Profit: SEK 112 million (+747% vs Q2 2024)

  • CFO: SEK 67 million (vs SEK 9.9 million Q2 2024)

• New Guidance Revision:

  • Revenue: > SEK 700M (vs previous 600M)

  • EBIT: > SEK 400M (vs previous 250M)

SEPTEMBER (-26.93%)

• EBIT target (the one revised for the second time) of SEK 400 million for 2025 already surpassed.

• Warrant exercise ~SEK 22.8m.

OCTOBER (-44.73%)

• Preliminary Q3 2025 Results:

  • Revenue > SEK 290 million (+300% vs Q3 2024)

  • EBIT > SEK 172 million (+700% vs Q3 2024)

  • Buyback recommendation of up to SEK 100 million.

• Additional clarification, in a separate release, regarding cash position. Issued on the same day as the preliminary results.

• Subsidiary Yuvio receives an order of EUR 22 million from a German customer.

• Additional clarification regarding the Yuvio press release. Days after announcing the €22M order, Intellego clarified that the client is MoveoMed GmbH and that the equipment was sold but not yet delivered.

• New COO: Henrik Resmark, starting in January.

• Board issues statement rebutting accusations, reinforcing strong financial position, robust governance (Deloitte audit, IFRS, bank due diligence), defending strategic partnerships (LiKang, Henkel), warning of possible market manipulation by third parties, and affirming focus on transparency, growth, and preparation for future main market listing.

• Nasdaq Stockholm demands rectification of the previous statement.

• Share issuance to Heights Capital Management (HCM), raising SEK 210 million in immediate cash.

• Announcement that they will anticipate the Q3 report release by two weeks.

NOVEMBER (-43.78%)

• Q3 2025 Report released:

  • Revenue: SEK 294.2 million (+296% vs Q3 2024)

  • EBIT: SEK 171.6 million (+698% vs Q3 2024)

  • Net Profit: SEK 81 million (+559% vs Q3 2024)

  • CFO: SEK 36 million (vs -SEK 5 million Q3 2024)

  • Buyback: Still wants to do it, but not right now after the deal with HCM.

• 2026 Target:

  • Revenue: SEK 2 billion

  • EBIT: SEK 1 billion

• CFO Change. Hans Denovan left, and COO Henrik Resmark took on the CFO position in addition to his current role.

• Board issues statement reaffirming that the business remains healthy and decides to propose a buyback of up to SEK 250 million (up to 10% of shares) given the drop in share prices.

• Company issues statement informing that LiKang increased purchases of dosimeters after strong market feedback.

• CEO Arrest. CEO Claes Lindahl, just one day after the LiKang statement, was arrested on suspicion of gross fraud. Swedish authorities believe the company misled the market through its 2025 press releases and quarterly reports. A freeze of SEK 100 million of the company’s cash was ordered.

• Board member Jacob Laurin takes over as interim CEO.

• Trading suspended and Nasdaq Stockholm initiates delisting process. The exchange informed that, following recent events, it saw grounds to remove Intellego from public trading and granted the company until December 5, 2025, to present its defense or counterarguments before the final decision.

• EKN and SEK filed a police report against Intellego on suspicion of fraud.

• Hiring of KPMG to conduct an independent forensic investigation into its financial statements.

Loading…

Conclusion

The reading of these events is enough to make an investor raise an eyebrow. Explosive growth, even more explosive guidance revisions, a CFO who comes and goes in a matter of months, press releases that need a press release to explain the press release, an overdose of optimism, a buyback, then dilution, then another buyback…

These are the old, familiar red flags. And Intellego, let’s be honest, didn’t have just one or two. It had a whole parade.

I won’t go through the full catalog of red flags here. Plenty of investors have already done that far more thoroughly than I ever could. Andrew’s YAV post is a great starting point if you want to go deeper. He even recounts a rather revealing call he once had with the then-CEO, Claes Lindahl.

What really captures my attention are the reactions to the red flags. There was no shortage of intelligent, competent people defending the thesis tooth and nail. People whose work I genuinely respect. Every red flag found its explanation, its excuse, its “this is just part of the playbook.”

The issue with red flags is that any single one, in isolation, doesn’t say much. It’s like a headache on a Monday. It could be a hangover, it could be a bad night’s sleep, it could be any of life’s minor annoyances. But when it’s followed by fever, nausea, and body aches, it becomes a diagnosis.

Companies aren’t much different. When the warning signs stack up, the most sensible approach is to keep your distance. The market is vast, and we don’t need to play every game it invites us to.

The question that lingers (and returns every so often) is why truly great investors held onto their optimism despite the mounting red flags?

The most heated bears will say it’s incompetence, that “these people never did any real analysis.” Nonsense. That’s just the emotional outburst of someone who spent months in a narrative war and now feels vindicated. Very good people make very ordinary mistakes. We’re all human.

For me, the answer doesn’t lie in analytical skill but in cognitive biases. Which is why those red-flag explanations weren’t analysis. They were rationalizations.

The thesis felt almost irresistible. Astronomical growth, low multiples and a founder with meaningful skin in the game. The presence of other smart investors and the arrival of Deloitte as auditor added authority and legitimacy. The soaring stock price fueled resulting and FOMO. The army of optimists offered that comforting sense of belonging against the bears.

In truth, the tribal war against the bears went well past healthy debate. It turned into months of heated arguments, sarcasm, and visceral blows. Nothing inflames conviction like this kind of attack. Public declarations cement a position. No one wants to look inconsistent. And in a such polarized environment, any legitimate critique from “the other side” gets dismissed outright.

There are many biases. And I am sure many of these investors know them all by heart. But as Kahneman taught us, knowledge doesn’t grant immunity. Recognizing biases is one thing, but escaping them forever is impossible.

What remains, then, is to consider what we can actually do about it. A bias checklist, perhaps. It may sound almost embarrassing, as if we need to write down the obvious. But if airplane pilots, with their own lives at stake, rely on checklists to avoid forgetting the essentials, why shouldn’t investors, who battle the constant threat of self-deception?

Or, at the very least, let’s embrace a simple, almost naive rule:

If I need to justify too many red flags, I am no longer analyzing.
I am rationalizing.

And when that happens, the wisest move is to step back and move on to the next one!

Share

Good morning, folks.

Hope you haven’t forgotten about me, because I sure haven’t forgotten about you. And to prove it, I decided to write this piece right in the middle of earnings season.

While the giants like Meta, Apple, and Amazon are dropping their numbers, I thought it’d be appropriate to share with you a company worth… six million dollars.

Yes, you read that right. Six million.

What can I say? As Forrest Gump once put it, “Negócios Pelo Mundo is like a box of chocolates, you never know what you’re gonna get.”

Today’s chocolate: Delivra Health Brands.

Delivra

Ticker:$DHB.V
Website: delivrahealthbrands.com

The beginning

Once upon a time, in Canada, there was a woman tormented by post-surgery pain. On one side, the fear of opioid addiction. On the other, a digestive system that couldn’t handle oral medication.

Her husband, Canadian scientist Dr. Joseph Gabriele, a pharmacologist and, apparently, a determined romantic, set out to relieve her pain through the skin. After years of research, he came up with a topical delivery system capable of sending therapeutic molecules directly to the pain site. He called it Delivra™.

The system worked. And in 2007, Delivra Corp. was born. The company focused on expanding applications of its delivery platform, most prominently through the LivRelief™ brand: a line of natural creams for muscle, joint, and nerve pain, as well as varicose veins, cuts, and mild burns.

The cannabis rush

Delivra went public in 2016 and quickly dove headfirst into the cannabis wave. It struck partnerships and licensing deals with Kalytera, Canopy, and Avanti, to develop cannabis-based therapeutic products.

Then came 2018, when Canada entered full-on cannabis euphoria. The new Cannabis Act turned the sector into an investor playground. Money poured in, stocks skyrocketed, and anything with “cannabis” in the name looked like a ticket to instant wealth. In the end, the carriage eventually turned into a pumpkin.

Delivra was one of the passengers on that wild ride. In 2019, it was acquired by Harvest One Cannabis for about $19 million. The new owner viewed the deal as a way to fast-track its entry into the wellness and cannabis-infused topical therapy market. But when the bubble burst, Harvest One had no choice but to undergo a major restructuring.

Turnaround

Harvest One went through a strategic review to streamline its operations. It moved away from the capital-intensive model of cannabis cultivation and processing, repositioning itself as a pure consumer products company (both with and without cannabinoids). The company sold non-core assets and strengthened its balance sheet, gaining room to reinvest in its two flagship brands, Dream Water and LivRelief. Costs were cut, management changed, and a leaner structure took shape. Along the way, the company also reclaimed a familiar name: Delivra Health Brands.

Let’s take a look at the timeline of this transformation:

• 2020 – The reset
  Sells cultivation arm United Greeneries and other non-core assets. Begins strategic review and refocuses on CPG with Dream Water, LivRelief, and Satipharm (European CBD). Revenue up 40%, but a CAD 80M loss due to goodwill impairment. CEO Grant Froese and CFO Aaron Wong leave. Gord Davey takes over.

• 2021 – The CPG pivot
  Completes the sale of Satipharm, ending exposure to cultivation and pharmaceutical. Fully focused on health and wellness, emphasizing sleep (Dream Water) and pain relief (LivRelief™). Partners with The Valens Company for the production and distribution of infused LivRelief topicals. Revenue stable, operating expenses down 58%.

• 2022 – Rebranding and expansion
  Full transition to CPG. Rebrands as Delivra Health Brands. Expanded distribution through major retailers such as Walmart, Shoppers Drug Mart, Rexall, Sobeys, and Amazon, while also securing international partnerships across the Caribbean, Latin America, Africa, and the United States. Cuts G&A and headcount by 60%.

• 2023 – Signs of inflection
  Positive EBITDA for the first time. Revenue up 20% YoY, gross margin reached 49% (from 9% in 2020). Sells Lucky Lake facility to strengthen working capital. Expands Dream Water in the U.S. and partners with Canopy Growth for infused LivRelief distribution.

• 2024 – Profit and growth
  First annual profit. Revenue up 26%, record 52% gross margin. Launches Dream Water Immunity Shots, LivRelief 1:1 THC/CBD, and Cooling Cream. Expands across the U.S. and Canada, increasing marketing investment. Cash reaches CAD 4.2m.

• 2025 – Consolidation and global expansion
  Revenue hits CAD 13.4m (+8% YoY) with gross margin above 50%. EBITDA positive for the third straight year. Dream Water expands into Saudi Arabia, Kuwait, and QVC U.S. Reverse split 10:1. Preparing for US CBD entry once regulations allow.

Today

Delivra currently sells three main brands:

• Dream Water® – sleep aids in liquid, powder, and gummy form, featuring melatonin, GABA, and 5-HTP.

• LivRelief™ – topical creams using a proprietary transdermal delivery system for muscle, nerve, and joint pain, varicose veins, and wound healing.

• LivRelief Infused – creams infused with CBD and/or THC.

As CEO Gord Davey said on the latest call: “We’re not a turnaround story anymore. We’re a growth story.”

To support that growth, Delivra defined five strategic pillars:

1. Distribution expansion: stronger presence across Circle K, 7-Eleven, Casey’s, Hudson News, WH Smith, Loblaws, Sobeys, Shoppers Drug Mart, and Rexall.

2. Internationalization: replicate Middle East success in Latin America and Europe.

3. Innovation: new Dream Water lines (Immunity, Beauty, Enhanced Sleep) and scientific advances in LivRelief.

4. Selective M&A: acquiring profitable or near-profitable wellness brands that fit Delivra’s structure.

5. Marketing and brand awareness: campaigns like “Shush Your Mind” (Dream Water) and “Quiets Chronic Pain” (LivRelief), plus influencer partnerships.

For those interested, the video below is quite recent and features a short interview with the CEO Gord Davey.

Numbers

The transformation Delivra has undergone in recent years is obvious in the numbers and impossible to ignore. Today, it’s a different company. Healthier. Leaner. And under control.

But when the CEO says Delivra is now a ‘growth story’... well, that might be a bit open to interpretation.

Yes, the company is growing. Just not the kind of explosive growth that “growth story” usually implies. At least, not yet. Davey sounds convinced about this growth potential. Now it’s up to the numbers to prove him right.

The old dependence on financings and dilution is gone. The current balance sheet can sustain operations, as long as no major acquisition shows up along the way (which, to be fair, can’t be ruled out, since it’s one of the company’s five strategic pillars).

As for valuation, it doesn’t offend. Quite the opposite. I’d say roughly 8x 2026 EBITDA, being conservative. But keep in mind that Delivra is a nanocap, just CAD 8.5 million in market cap. With the stock trading at C$0.27, any sneeze can move the needle by double digits.

Naturally, a company this small isn’t for everyone. The risks are there, both in the stock and in the business itself. It’s illiquid, operates in a regulated and competitive market, and growth usually demands heavy marketing spend.

On the other hand, it’s a transformed company. And the management team deserves credit. They executed the turnaround remarkably well, and today’s Delivra is nothing like the one that existed a few years ago. It’s that same team that now carries the optimism for what comes next.

What do you think? Is it worth sharing that optimism?

From where I stand, I’d say it’s at least worth sharing this post.

Best,
Leo Caroli

Share

1974 was a year of unrelenting drama. Not the trivial gossip of local affairs, but real, high-stakes upheaval. We’re talking Watergate, conflict in the Middle East, the collapse of Bretton Woods, American stagflation, and the oil crisis. Who enjoyed every moment of this? The news sellers, of course! Misfortune was abundant...

No disaster is truly complete without a respectable bear market. Say no more! Between 1973 and 1974, the Dow Jones and S&P 500 plummeted nearly 50%. Newsweek's cover from that time vividly captured the bear’s fury, and this month marks its 50th anniversary. Like revisiting a battlefield, we will travel back to that era of uncertainty—not just for nostalgia, but to learn. Chaos doesn’t only serve the news sellers; it serves us all, offering fertile ground for timeless lessons.

1974

Amid the stifling tension of the market, as the stock exchange plummets, an old newspaper lies calmly on the table. Surrounded by apocalyptic reports and blood-red charts, this paper stands untouched by the chaos. It doesn't scream scandals, delve into politics, or attempt to predict the uncertain future of the economy. Instead, week after week, it does only one thing: it introduces a new company to the reader, providing a quiet glimmer of hope amid the storm.

The reader picks up this paper, 'Negócios pelo Mundo'. In elegant, fine print, the company's numbers unfold:

Its track record as a public business is modest, almost timid. The company only started trading on the NYSE in 1972, with its fiscal year closing on January 31st. Yet behind this brief public history lies a business with three decades of operation. Margins are thin, almost to the edge, but don’t be fooled: this is a well-managed business. It generates cash—something many only dream of. It dilutes sparingly and is solidly capitalized. Growth? Outstanding! Not through flashy jumps, but methodically, brick by brick. And believe it or not, in 1974, when the entire world seemed to be selling off everything down to their last shirt, the company started paying dividends. Amid the market drama, this company seems untouched.

But is it really?

Its stock crumbled like a sandcastle. In 1973, it hemorrhaged 61%. And in 1974, as if life weren’t bitter enough, it dropped another 29%. A staggering decline of over 70% in just two years! Not even the beaten-down indexes suffered this much. Perhaps there was solace in a tempting valuation. But maybe not even that. With indexes trading at 7x earnings, its PE of 13x still taunted investors like a cruel joke.

So now, what’s left to do?

Is it a buy or a sell?

The desperate investor searches frantically for answers. But ‘Negócios pelo Mundo’, that obscure little paper, seems to exist solely to humble us. It offers no recommendations, no guidance, not even a clue! Instead, it forces us to think. Think? How ironic! Because in the chaos of 1974, no one is thinking anymore. They’re just reacting—desperately. After all, a market collapse is ultimately a crisis of faith.

Is it a buy or a sell?

The exhausted investor tosses the newspaper aside. In that moment, he’s forced to reflect. Amidst the cold numbers of that company and his own emotional struggle in this turbulent market, he asks himself once again:

Is it a buy or a sell?

2024

Fifty years have passed, and our imaginary DeLorean has brought us back to September 2024. Like a specter from the past, a familiar name storms back into the headlines. With silver hair and wide glasses, this Arkansas-born woman is the new chief in town. Alice Walton once again claims the title of the richest woman on the planet.

Somewhere in the world, our investor friend, who has also journeyed through time, sits absorbed in an article about Alice Walton. In a world captivated by wealth, her face is everywhere. With a blend of nostalgia and quiet satisfaction, he devours the piece, feeling, in some strange way, like a distant sibling to Alice. No, they’ve never met. But their fates became intertwined, like stars bound to the same orbit. Their connection was forged in the chaos of 1974, when, holding onto his sanity for just a brief moment, our investor friend made a decision that would forever alter the course of his life:

It’s a buy! $10,000 in Walmart, BUY!

Today, the United States breathes the omnipresence of Walmart. So much so that 90% of Americans live within 10 miles of one of its stores. To many, it seems like an obvious move, a play anyone could have predicted. Our investor friend listens to this with a subtle, knowing smile—the kind only those who’ve fought in the arena possess. He knows all too well that any fool can play the prophet after the fact. But he also knows, and this is what truly matters, that no one can simulate the gut-wrenching fear that comes in the heat of chaos.

Our investor friend contemplates Walmart’s 2024 surge—a staggering 50%—with the calm detachment of someone who’s seen it all. The wide-eyed wonder of novices is long behind him. Market crashes, which once jolted his soul, don’t even cause a flinch now. He stands like an oak in the face of the storm, numbed by five decades of a narrative rollercoaster—stories that come with grandeur, yet prove hollow. Not even Amazon, with its ominous, overpowering presence, has ever cost him a night’s sleep.

What started as a modest $10,000—almost naive in its simplicity—has swelled into an astonishing $150 million. Yes, you read that right, one hundred and fifty million. And if not for the irresistible allure of life’s luxuries—a beachfront villa here, an Italian sports car there, the indulgence of unforgettable travels—that figure would have easily crossed the $200 million mark. The Walmart of 1974 wasn’t just a buy; it was a generational buy, the kind that shapes legacies and echoes through time as the hallmark of true foresight.

Our investor friend slowly pulls away from the article on Alice Walton, a woman whose ‘buy’ in Walmart came as naturally as birth. With a nearly solemn gesture, he places his glasses on the table and drifts into thought. What an extraordinary journey it has been! An odyssey that began with a simple, almost mundane decision: to buy. And then came the hardest choice of all—the one that demands patience, unwavering resolve, and an ironclad belief in your decision: to hold.

When asked how he accomplished such a feat, he hesitates—not because he lacks answers, but because the complexity of his journey—from the highs of euphoria to the depths of despair, from certainty to fear—makes it difficult to know where to begin.

It’s only fair to begin with Sam Walton. Yes, Sam—a fanatic. One of those rare, almost mythical figures whose mere presence alters the course of history for anyone fortunate enough to cross his path. The rational, cautious man knows better than to stand in the way of such a force, knowing he’d be crushed by a will that bends to no ordinary logic. But the wise man—ah, he walks alongside him. He recognizes the relentless, burning intensity of obsession and understands that within that madness lies the key to something far greater.

Sam Walton wasn’t just a businessman; he was a force of nature. His obsession with growing the business, his almost religious devotion to the customer, and his relentless drive to be the best, the cheapest, the most efficient made him unstoppable. When you encounter someone like that—rare as a meteor—there’s only one rational move: latch on and hold tight. Wherever he goes, wealth is bound to follow.

The quality of the business, my friend, was undeniable—blindingly obvious to anyone bold enough to see it. There was no mystery, no illusion. Three decades of an almost outrageous track record in efficiency, born from a modest shop in Bentonville and swelling like an avalanche into dozens of stores, all while still under the constraints of a private company.

And the business model? Simple, almost insultingly so. No miraculous technologies, no flashy innovations to make the mediocre marvel. Just the same relentless mantra: offer more for less. Those chasing something more 'sophisticated' ironically overlooked the brilliance. The secret—the only secret—lay in ruthless execution, in the unwavering consistency of every move, every brick Sam Walton laid with his own hands. The lesson for the sharp investor was painfully clear: simplicity, when fueled by obsession, is worth more than any technological miracle.

Now, let’s add the explosive ingredient: the sheer size of the opportunity. What lay ahead was a vast expanse of growth, a promise that stretched across decades of achievements. It was here that the painful, almost resigned acceptance of paying the market’s price settled in—because true greatness demands a premium. And Munger, the prophet of patience, quietly revealed a truth that few grasp: over time, an investor’s returns dance to the rhythm of the business itself, indifferent to slightly higher or lower multiples. Time, that ruthless teacher, reduces the initial price to a mere footnote in the face of five decades of compounded business exuberance.

And finally, courage. Or rather, the kind of courage born from rationality. Not the blind bravery of fools, but the cold composure of someone who understands the moment. Like firefighters running into the flames, the true stock accumulator charges forward when the market bleeds—because that’s where salvation lies. It’s the age-old commandment: buy on the sound of cannons! Despite the price demanded by the market, that stock was collapsing—down 70% in a storm of collective panic. And it’s in that very abyss where greatness is found, for those brave enough to take the leap.

Yes, that’s exactly what our investor friend would say if asked about his achievement—if anyone were curious enough to ask about his journey. He doesn’t want his story to be seen as a lesson in paying high multiples or as an example of reckless behavior. No! What he truly wants is for us not to hastily see elevated multiples as an inevitable tragedy. Because that’s the lesson he learned from the best of stocks…

At nearly 80 years old, that's exactly what our investor friend would say, if only there were someone to hear. But no one listens. In a world captivated by wealth, all eyes are on Alice Walton, the richest woman on the planet.

Our investor friend adjusts his glasses with a resigned sigh and turns on his computer—a modern tool he’s begrudgingly accepted. He types into the browser the name of an old acquaintance that has also succumbed to the digital age:

‘Negócios Pelo Mundo’

Does it ring a bell? That obscure publication that doesn’t scream scandals, delve into politics, or attempt to predict the uncertain future of the economy.

The computer screen lights up, revealing a surprise with the headline:

“A tale of two stocks, p.1”

It was his story—his own tale now revealed for the world to see!!!

A story he thought no one cared about.

Except for that obscure publication.

Thank you for reading,

Leo Caroli

PS 1: Here’s the link to Walmart’s 1974 annual report. Back then, annual reports were just 18 pages long... It’s a piece of history worth reading:

1974 Annual Report For Walmart Stores Inc

3.09MB ∙ PDF file

Download

Download

Share