Note that Part I of my upcoming HEICO Deep Dive will be published on Sunday, July 19.

Not many people want to hear what I’m about to say about KPG.

This may not make me popular.

At first glance, Kelly Partners looks like a classic “quality compounder”: a steady, recurring-revenue professional services business that grows through acquisitions in a large, fragmented market, posts attractive margins, and is led by a (seemingly) highly incentivized founder.

But once you look beneath the surface, a few things start to bother me. The issue isn’t the business itself. It’s the math and how the deal structures can make consolidated financials look far better than the economics that are actually attributable to shareholders.

The business model

Kelly Partners Group (KPG) was founded in 2006 by Brett Kelly. KPG is a programmatic acquirer of tax and accounting firms. The early focus was Australia (major metro areas), but more recently, the group has expanded into the UK and the U.S. (All figures in A$ unless noted.)

The market itself is structurally attractive. Tax and accounting services are high-retention and relationship-driven, especially for SMEs and private clients. Switching accounting firm is possible, but the switching costs are real: years of accumulated know-how would get lost. In practice, most business owners don’t wake up thinking, “Today I’m shopping for a new accountant.” If the work is solid and deadlines are met, inertia is powerful.

And believe me: sometimes the accountant knows more than the client’s own spouse. I’m speaking from experience—I spent several years in audit and at a small accounting firm. Once you sit close enough to the numbers, you realize how much sensitive information flows through the accountant-client relationship. In extreme cases, that can even include “undisclosed” children and related financial obligations the spouse is not aware of—and is not meant to be aware of.

The accounting market is also deeply fragmented with many small local accounting firms, and the total addressable market is large. KPG argues that across its three operating markets there are roughly 87,000 smaller firms (excluding the Big 4 and mid-tiers) that could face succession by 2029. That would be about ~54% of roughly ~162,000 existing accounting firms. In other words: plenty of targets, and plenty of willing sellers.

According to KPG, the market is roughly $12.5 billion in size. With about $130 million in revenue, KPG still has less than 1% market share, which suggests a long runway for growth.

In Australia, KPG is now the 11th-largest accounting firm, with an estimated ~3% market share (excluding the Big 4: Deloitte, EY, PwC, and KPMG).

The bird’s-eye view

From 30,000 feet, the following chart is impressive: a ~25% revenue CAGR over 17 years…

…paired with a nice-looking EBITDA trajectory (even though the EBITDA margin has been meaningfully lower over the last three years).

Add a founder who knows how to sell the story to the market. The investor deck hits all the right keywords:

How many IR decks include a “100-bagger checklist”?

If you stop here, the story sounds close to ideal: recurring revenue + fragmentation + programmatic M&A.

Now comes the uncomfortable part (that I care more about than many others and that keeps me from adding KPG to my portfolio despite the hefty drawdown).

The 51/49 structure

KPG’s twist is its 51/49 structure: it typically acquires only ~51% of the operating firms, while local equity partners keep ~49%. Management presents this as an alternative to traditional “everyone-equal” partnerships—meant to keep operators invested and accountable while still enabling centralized strategy and scale benefits.

But if the pitch is a huge market with a looming succession wave, why do sellers keep nearly half the equity so consistently? I understand minority rollovers—plenty of serial acquirers do that—but they usually buy 75–90% and leave 10–25% with the seller. At KPG, buying ~51% isn’t occasional. It’s the operating model.

KPG calls this the Partner-Owner-Driver® model and describes it as driving long-term strategic alignment and creating a foundation for long-term success and growth for the clients, people, and partners of Kelly Partners.

The listed parent (KPGH) owns 100% of the local holding companies (labeled “Subsidiary Company” in the diagram below). Those entities then acquire a narrow majority stake in the operating businesses that are usually structured as partnerships.

The consolidation optics

When KPG buys ~51% of a firm, it usually controls it—so it consolidates 100% of that firm’s revenue, costs, profit, assets and liabilities into the group financials. That’s normal accounting according to IFRS. But it creates a very abnormal investor trap:

Consolidated numbers can look excellent even when very little incremental shareholder value is created by the structure.

A quick detour into consolidation accounting: a group financial statement treats the parent and its subsidiaries as one economic unit. Once the parent controls a subsidiary (typically via >50% ownership), the subsidiary is fully consolidated. Revenue, costs, and profit show up at 100% in the income statement. The same applies on the balance sheet: assets and liabilities are consolidated at 100%.

The portion of profit and equity attributable to minority shareholders is reported as non-controlling interests (NCI) below net income in the income statement and within equity on the balance sheet.

A simple example illustrates this. Assume the Parent has no operating business of its own and only incurs $5.0 million of interest expense, while the Subsidiary generates all operating profit. Whether the Parent owns 100% or just 51% of the Subsidiary, the consolidated income statement looks identical from revenue down to net income (EAT), because control requires the Subsidiary to be fully consolidated. In both cases, the Group therefore reports $100.0 million of revenue, $40.0 million of EBITDA, and $16.0 million of EAT.

The difference appears only below EAT, where the Subsidiary’s profit is allocated between non-controlling interests and the Parent’s shareholders. In the 51% ownership scenario, $10.3 million of EAT is attributable to minority shareholders, leaving only $5.7 million attributable to the Parent’s shareholders after the Parent’s interest expense. The same economic distinction applies further up the income statement: although the Group reports $40.0 million of consolidated EBITDA, only $20.4 million, or 51%, is economically attributable to the Parent’s shareholders.

Most investors ignore NCI, and for most companies that’s fine because the amounts are immaterial. With KPG, it’s a mistake. If you focus on the fully consolidated numbers and mentally dismiss the minorities, you can end up valuing economics that don’t belong to shareholders of the parent company.

You can see it in the balance sheet: equity attributable to KPGH shareholders is only A$28.4 million, while equity attributable to NCI is A$38.1 million.

The same story shows up at the very bottom of the income statement (a line many investors never reach—they stop at EBITDA). Out of A$16.5 million of profit, only A$3.4 million (!) is attributable to the parent; A$13.0 million goes to NCI.

Read that again: the group reported A$16.4 million of statutory profit, but only A$3.4m was attributable to the parent entity under statutory accounting in FY25.

An EBITDA of ~A$40 million carries a very different “story value” in an investor deck than A$3.4 million of net income attributable to shareholders.

The chart below shows how the group’s profit has been split over time.

Summing the period FY2015/16 to FY2024/25, KPG reported total group profit of A$99 million. Of that:

• Attributable to shareholders: A$31 million (31%)

• Attributable to minorities: A$68 million (69%)

That’s substantial.

So why isn’t the split 51/49?

Because the listed parent incurs costs that the partners do not share—such as holding-company expenses, acquisition-related amortization, other purchase-accounting effects, and more. This is the same mechanism illustrated in our hypothetical example: although the minority owners hold only 49% of the subsidiary, 64% of reported Group (!) net income is attributable to NCI, while just 36% accrues to the Parent’s shareholders. The reason is that costs incurred at the Parent level are borne entirely by its shareholders and do not reduce the earnings attributable to NCI. As a result, the NCI share of consolidated net income can be materially higher than the underlying minority ownership percentage, particularly when the Parent’s standalone cost base is significant.

KPG provides the following reconciliation:

Again: this isn’t fraud. It’s just the deliberate outcome when you buy just 51% in a partnership-style structure. The only mistake is pretending the consolidated numbers describe shareholder economics.

The cash test: where the story gets real

The most important question to ask (and the one almost nobody asks):

How much cash is left for the parent shareholders after the partners get paid?

Unlike profit, cash flow cannot simply be attributed to shareholders and NCI based on their respective ownership percentages. Cash generated by a subsidiary initially remains within that legal entity and is not allocated to either shareholder group until it is distributed. The cash flow statement therefore follows the consolidated-entity perspective and reports the Group’s total cash movements, while only actual transactions with NCI and shareholders—such as dividends or changes in ownership interests—are separately identifiable. Consequently, consolidated operating cash flow may materially overstate the cash that is economically available to the Parent’s shareholders.

If the parent company owns 100% of a subsidiary, the parent can move cash upstream via dividends or cash pooling without any real friction. Those intra-group movements disappear in consolidated financial statements.

Once meaningful minority owners exist, that changes. Cash pooling can still happen, but it creates real intercompany balances that have to be repaid; the cash is not simply “free.” Dividends can be paid, but then minorities participate too.

For FY2024/25, KPG reports operating cash flow of A$31.3 million. Against EBITDA of A$39.3 million, that looks like normal 80% cash conversion.

But then you hit the line that matters:

In financing cash flow, KPG reports “Distributions paid to non-controlling interests.” In FY2024/25, that was A$19.1 million (!)—more than half of operating cash flow.

These are partnership distributions, comparable to dividend payments from the subsidiary to its owners (e.g. the parent company and minority shareholders). The parent company also receives its own share internally, but you don’t see that as an external outflow, as intra-group transactions are eliminated in consolidated numbers. The cash paid to minority owners leaves the group, which is why it shows up in financing cash flow. Either way, it’s obviously not cash that belongs to public shareholders—and ignoring it is costly.

Owners’ Earnings

KPG highlights a metric it calls “Owners’ Earnings.” It starts with operating cash flow and subtracts lease repayments and maintenance capex:

• Step one: subtract “repayment of lease liabilities” of A$6.4m. Under IFRS 16 (AASB 16 in Australia), office leases create a right-of-use asset and a lease liability. The right-of-use asset is depreciated (lifting EBITDA and depreciation), while cash lease payments are split between interest expense and principal repayment. That principal repayment is the A$6.4m reported in financing cashflow. Before IFRS 16, rent was simply an operating expense. Without this adjustment, operating cash flow can look optically higher for lease-heavy businesses. If you don’t correct for it, operating cashflow is overstated (I have written about this topic here).

• Step two: subtract maintenance capex. IFRS doesn’t define this line item, so it inevitably involves management judgment (not necessarily good or bad—just something to keep in mind).

On that basis, KPG’s “Owners’ Earnings” chart implies roughly ~24% annual growth:

But the ~A$24 million reported for FY25 are not, as the label “Owners’ Earnings” might suggest, cash that belongs to KPGH shareholders. A cleaner shareholder-level view subtracts the distributions to NCI. The chart then looks like this:

There’s still a positive trend—but the owners’ earnings attributable to shareholders are materially lower than what KPG discloses for the group. Far less impressive than growth in reported EBITDA and in Owners’ Earnings.

KPG says it aims to communicate transparently. Still, it’s hard to shake the feeling that this topic is kept in the background. In June 2025, I emailed Brett Kelly asking why distributions to NCI are ignored in the “Owners’ Earnings” narrative and equity story. After a couple of follow-ups, I still haven’t received a response.

To be fair, KPG does calculate Owners’ Earnings for the parent company somewhere in its investor presentations (hidden). It’s close to my estimate, but not identical. The problem is that the bridge isn’t explained anywhere, and the parent-level number isn’t put at the center of communication, although this metric reflects the true economics attributable to shareholders much better. The communication remains on the group-level optics, which is misleading.

The chart below shows how KPG has allocated capital over time, including distributions to NCI, which are too material to ignore.

From FY2023/24 to FY2024/25,

• EBITDA increased by $5.8 million, from $33.5 million to $39.3 million,

• while adjusted operating cash flow—defined as operating cash flow less IFRS 16 lease repayments—rose by $4.8 million, from $20.1 million to $24.9 million.

• However, distributions paid to NCI increased even more sharply, by $9.7 million, from $9.4 million to $19.1 million.

As a result, the owner earnings attributable to KPG’s shareholders declined from $10.7 million to just $5.8 million despite the improvement in reported EBITDA and cash generation.

Leverage

As a result of the inflated revenue and EBITDA lines, the headline net debt / EBITDA is understated. KPG reports gearing (Net Debt / Underlying EBITDA) of 1.42x for FY2024/25. It looks conservative, but does it reflect the underlying economics appropriately?

Remember: KPG’s fully consolidated EBITDA includes subsidiaries it owns only ~51% of. But most of the debt sits at the parent level. Putting parent net debt against fully consolidated EBITDA—of which shareholders effectively own only a bit more than half—can paint a dangerously rosy picture.

I’d make a few adjustments:

• Net debt: KPG excludes roughly A$7 million of contingent consideration. I would typically include it when calculating net debt.

• EBITDA: I would haircut EBITDA by 50% as a blunt proxy for the portion economically attributable to minorities.

• Leases: KPG’s leverage framing also downplays IFRS 16. Economically, offices aren’t free. You should count leases once—either include lease liabilities in net debt or adjust EBITDA to reflect rent, as long as leasing is not immaterial. I include lease liabilities in net debt.

On my net debt number alone, leverage moves to ~2.5x. If you then use only the portion of EBITDA that is actually attributable to the parent (and therefore to debt repayment), leverage rises to ~5.0x, vs. ~1.4x disclosed by KPG.

Excluding lease liabilities, net debt still comes in at ~A$65 million. Comparing that number with the ~A$5-10 million Owners’ Earnings attributable to the parent, it doesn’t look that conservative anymore. It would take 5-10 years of cashflow available to shareholders to pay down net debt.

Conclusion

Stepping back, here’s how I think the model really works: KPG typically buys just 51%. That gives it control—and with it, full consolidation—so the reported revenue and EBITDA lines look great, and the upfront cash outlay per deal stays relatively low (vs acquiring 100% right away). The trade-off is that a meaningful share of the economics never belongs to the parent in the first place. It shows up later as a structural cash outflow in the financing section—“distributions to NCI”—right where most investors don’t bother to look. These distributions to NCI will never show up in FCF, although it’s a significant “expense” for shareholders of the parent company KPGH.

This structure makes the company appear economically larger than it really is—and that can support a higher valuation if you just look at EV/EBITDA-multiples. The obvious question is: how many analysts and investors actually value the minority interests properly and include them in enterprise value or cut EBITDA in half to reflect only the portion that is attributable to shareholders of the parent company?

It also supports a low-looking leverage ratio that can be used to fund more acquisitions. The group keeps adding consolidated EBITDA—without generating much free cash flow for public shareholders—because distributions to NCI sit in financing cash flow and quietly drain the pool. Those payments are a structural outflow. They are not discretionary capital allocation.

With a market cap of roughly A$175m and net debt of about A$97m (including earn-outs and lease liabilities), enterprise value is around A$272m. Using my adjusted owners’ earnings attributable to shareholders of A$4.8m, and adding back A$6.5m of interest expense (which reduces operating cash flow), I get roughly A$11.3m of “adjusted owners’ earnings before interest.” That’s an owners’ earnings yield of about 4.2% on EV—even after a ~60% drawdown in the share price.

For me, KPGH is not investable, and the recent developments surrounding Brett Kelly have only reinforced my critical assessment—although I do not intend to discuss them further here.

New to Slow Compounding?

This Company Snapshot is the short version of the most extensive research project I have published so far.

The full HEICO Deep Dive was published in collaboration with Compounding Quality, one of the best-known investment newsletters on Substack with more than 550.000 readers.

The project is based on months of work, including going through more than 1,200 pages of earnings call transcripts, decades of annual and quarterly financials, aviation aftermarket reports, regulatory documents, PMA industry sources, and even a doctoral thesis on the aircraft parts value chain.

Slow Compounding is built for long-term fundamental investors who want deep, source-driven research on high-quality compounders and serial acquirers.

Subscribe now

A small plastic bag containing 1,000 metal bushings can cost an airline $50,000.

That price has little to do with the value of the metal. It reflects the extraordinary market power of OEMs, which can use their control over approved replacement parts to charge exorbitant prices and push through substantial price increases over time.

HEICO has built one of the world’s most successful compounding machines around this simple imbalance.

Since the Mendelson family took control in 1990, revenue has grown from $26 million to approximately $4.5 billion, while net income has increased from $2 million to almost $700 million. Along the way, HEICO has developed thousands of aircraft parts, completed more than 100 acquisitions, and expanded into a decentralized portfolio of highly specialized aerospace and electronics businesses.

Yet the company’s success is not simply the result of selling cheaper parts or acquiring more businesses.

It is the result of a system in which regulatory barriers, customer trust, recurring aftermarket demand, entrepreneurial autonomy, and disciplined capital allocation reinforce one another.

In our Quality Growth framework, HEICO checks almost every box:

🛩️ Recurring demand from aircraft that require maintenance for decades

🧱 High barriers created by regulation, operating history, and customer trust

💰 A strong value proposition based on lower prices and better availability

🔬 A repeatable organic growth engine driven by new product development

🤝 A decentralized acquisition model with a reputation as a permanent home

👨‍👩‍👦 Meaningful family ownership and a genuinely long-term culture

📈 A multi-decade runway supported by fleet growth and rising aftermarket demand

This Company Snapshot provides the guided tour: what HEICO does, why its business model is so powerful, how its two operating groups differ, and where the opportunities and risks lie.

Please do not click "Download"—it links to the Teqnion Deep Dive. A bug in the Substack editor is currently preventing me from deleting it.

Download

👔 Company Name: HEICO Corporation (“HEICO”)

🔎 ISIN: US4228061093 (Common Stock) and US4228062083 (Class A)

🔧 Business model: Aerospace and electronics serial acquirer

🌍 Geographic exposure: global (~62% U.S. and ~38% international sales across approximately 130 countries)

📈 Stock Price: HEI Common Stock: $339 | HEI.A Class A Common Stock: $253

💰 Market Capitalization: ~$40 billion

👨‍💼 Number of CEOs since foundation: 2 (one transition — Laurans Mendelson to Eric and Victor Mendelson)
since 1990

👨‍👩‍👦 Founder-/Owner-operator: Yes — owner-operated by the Mendelson family

📅 CEO tenure: Since 1990

🥇 Insider ownership: ~21.0% of shares

📊 10Y EPS CAGR: ~16%

🔁 Reinvestment profile: High

💸 Capital intensity: Low — capex-light, but inventory-intensive

🏰 Moat: Regulatory approvals, customer trust, installed-base scale, price and availability, culture

🧨 Main risks: Valuation, acquisition execution, aviation cyclicality, OEM countermeasures

🌳 Slow Compounding fit: Yes – high-quality (niche) serial acquirer

Business Model in a Nutshell: HEICO is a decentralized aerospace and electronics serial acquirer. Its Flight Support Group provides lower-cost aircraft parts, repairs, and distribution services, while its Electronic Technologies Group supplies highly engineered components for aero-space, defense, space, medical, and other de-manding markets. Growth comes from new product development, market-share gains, and disciplined acquisitions of niche businesses that retain signifi-cant operational autonomy.

Subscribe now

If you are new to the world of serial acquirers, please check out my fundamental piece before you continue with Heico:

A World-Class Compounding Flywheel Riding a Multi-Decade Tailwind

A $50,000 Bag of Bushings

Imagine I handed you a small plastic bag containing 1,000 metal bushings. They are unremarkable to look at—small, round pieces of metal that most people would probably describe as washers or spacers. Unless you happened to know exactly what you were looking at, you might expect to find something similar in a hardware store.

How much do you think the bag costs?

Perhaps $100? Maybe $500? Even $2,000 would sound expensive for a collection of such simple-looking components.

The original equipment manufacturer (OEM) charges $50,000.

The bushings are installed in a CFM56 aircraft engine, one of the most widely used commercial jet engines in the world. Yet the fact that they ultimately end up inside an engine does not make the price any less striking. The bag contains 1,000 small pieces, implying an average price of $50 per bushing. A set of approximately 400 small Teflon washers can also cost $50,000. So can 80 drain plugs. In the latter case, the customer is paying more than $600 for each plug.

At first glance, these prices seem almost impossible to justify. The raw materials are clearly not worth tens of thousands of dollars, and the physical manufacturing process alone cannot explain the difference. Yet the airline is not really paying for the metal, the Teflon, or the plug itself. It is paying for something much more valuable: the right to install that component on an aircraft and continue operating the aircraft safely and legally.

That distinction is the key to understanding HEICO.

A commercial aircraft can be worth well over $100 million, but it generates revenue only while it is flying. If one small component is worn, damaged, or unavailable, the entire aircraft may remain on the ground. Passengers may have to be rebooked, connections can be missed, crews and airport slots may be lost, and the disruption can spread across the airline’s network. The economic damage caused by an aircraft sitting idle for another day can therefore be many times larger than the price of the part needed to return it to service.

The obvious response would be to buy the component somewhere else. After all, there are thousands of industrial manufacturers capable of producing small metal parts. An airline could presumably take the original bushing or washer to one of them, ask for an identical copy, and avoid paying the OEM’s extraordinary price.

In aviation, however, that is not an option.

A part cannot be installed on a commercial aircraft merely because it looks identical, fits into the same space, or is made from the same material. Its design must be approved, it must be manufactured within an accepted quality system, and its origin and production history must be documented and traceable. The manufacturer must also be capable of showing that the part performs its intended function under the demanding conditions in which an aircraft operates.

A machine shop may be perfectly capable of manufacturing an identical-looking component. That does not make the component an approved aircraft part.

When the original aircraft, engine, or component manufacturer is the only approved source, the airline therefore has very little bargaining power. It can purchase the part at the price requested by the OEM, wait for the OEM to deliver it, or keep the aircraft on the ground. This combination of regulatory barriers, limited alternative supply, and the urgent need to return expensive assets to service explains why the price of a small replacement part can bear so little relationship to its raw-material value.

HEICO was built around offering airlines another choice.

The easiest way to understand what the company does is to think about the difference between branded and generic medicine.

When a patient visits a pharmacy, there may be two versions of the same treatment. The first is the branded drug sold by the pharmaceutical company that originally discovered and developed it. That company financed the research, carried the risk that the development program might fail, completed the necessary clinical work, obtained regulatory approval, and established the product in the market. The branded drug therefore reflects not only the cost of manufacturing the tablet itself, but also the original development burden and the value attached to the brand.

Over time, other manufacturers may be permitted to offer an approved generic alternative. The name on the packaging is different, but the generic product must satisfy strict regulatory requirements and provide the same intended therapeutic effect. Because the generic manufacturer did not finance the original discovery process and does not carry the same brand premium, it can normally offer the medicine at a substantially lower price.

The analogy is not perfect in every technical or legal respect, but it captures the economic idea remarkably well.

The OEM aircraft part is the branded drug. A PMA part is the regulated generic alternative.

PMA stands for Parts Manufacturer Approval, an authorization granted by the U.S. Federal Aviation Administration. It allows an independent company to develop and manufacture a new replacement part that can be installed in place of the corresponding OEM component. A PMA part is not second-hand, refurbished, counterfeit, or an unofficial copy. It is a factory-new aviation product manufactured under an FAA-approved design and production system and required to comply with the applicable airworthiness standards.

In other words, HEICO does not simply produce a cheaper piece of metal that happens to fit into the same opening. It performs the engineering work, obtains the regulatory approval, establishes a controlled manufacturing process, and provides the documentation and traceability that allow an airline to install the part with confidence.

That is what turns an apparently ordinary component into a legitimate alternative to the OEM product.

HEICO has become the world’s largest independent provider of such approved alternatives. Historically, the company has introduced new PMA parts at prices approximately 30–40% below those charged by the OEM.

In some cases, the difference becomes considerably larger over time. OEMs have frequently used their strong aftermarket positions to raise replacement-part prices substantially, whereas HEICO has generally followed a much more restrained pricing approach. As the two prices develop along different paths, an initial discount of 30–40% can eventually widen to 50%, 60%, or even approximately 70%.

These savings matter particularly in an industry that has historically struggled to earn attractive returns on capital. Airlines operate with thin margins and have limited control over many of their largest expenses. Fuel prices are largely determined by global markets, labor is operationally essential, aircraft ownership and leasing costs are contractually fixed, and airport and air-traffic-control charges cannot simply be negotiated away. Maintenance, repair, and overhaul (MRO) is therefore one of the relatively few major cost categories in which an airline can actively reduce spending without flying fewer routes, removing seats, or weakening the passenger experience. A dollar saved on maintenance can flow directly into operating profit without the airline having to sell another ticket.

For airlines, those savings are not limited to one unusual purchase. Aircraft maintenance creates frequent and recurring demand. Components wear out, fail, become damaged, reach mandatory inspection limits, or have to be replaced after a specified number of flight hours or takeoffs and landings. A commercial aircraft may remain in service for 20–30 years, while the underlying aircraft or engine platform can continue generating demand for replacement parts and repairs for 35–40 years or longer.

A single approved part may therefore be purchased repeatedly across hundreds or thousands of aircraft over several decades. Once HEICO has completed the engineering and regulatory work, the same product can continue generating revenue whenever another component needs to be replaced.

This is what makes the model so powerful. HEICO is not merely helping an airline save money on one bag of bushings. It is gaining access to a long stream of recurring demand within an installed fleet that continues to fly, wear, and require maintenance year after year.

The $50,000 bag therefore represents much more than an example of excessive aerospace pricing. It illustrates the economic gap on which HEICO was built: a captive OEM aftermarket on one side, and airlines desperate for safe, approved, available, and more affordable alternatives on the other.

HEICO sits between them.

Sharing the Savings

The remarkable feature of HEICO’s model is not only that the company can sell an approved alternative for less than the OEM. It is how much of that difference management has historically chosen to leave with the customer.

Suppose an OEM part costs $100 and HEICO can develop and manufacture an approved alternative at a much lower cost. HEICO could charge $90 or $95. The airline would still save money, and HEICO would retain most of the economic benefit created by the alternative. Given the customer’s limited options and the importance of keeping the aircraft in service, such a strategy might remain profitable for a considerable period.

HEICO has deliberately avoided that approach. Instead, it may offer the part for $60 or $70, creating a saving large enough to matter to the airline and to justify the technical and administrative work required to approve another supplier. This is not an indication that HEICO lacks pricing power. It is a deliberate decision to reinvest part of that pricing power in the customer relationship.

The airline lowers its maintenance costs, while HEICO can still earn an attractive margin because it does not carry the development cost of the entire aircraft, engine, or larger system. Lower prices encourage greater adoption, support higher volumes, and make customers more willing to evaluate additional HEICO products. Those volumes, in turn, allow HEICO to invest further in engineering, product development, inventory, manufacturing, quality control, and distribution. The stronger the platform becomes, the more products it can develop and the more value it can offer to customers.

The approach resembles Nick Sleep’s concept of scale economies shared, most famously associated with Costco. Costco does not retain every benefit created by its purchasing power and operating scale. It returns a meaningful share to its members through lower prices, thereby strengthening customer loyalty and increasing volume. The additional volume improves Costco’s bargaining power and cost position, allowing it to offer even more value and reinforcing the cycle.

HEICO follows a similar principle in a very different industry. As the company’s catalog, engineering base, purchasing power, manufacturing knowledge, customer relationships, and distribution capabilities have expanded, it could have used those advantages primarily to increase prices. Instead, it has preserved a substantial discount to the OEM. Customers save money, HEICO earns attractive returns, and the resulting trust and volume support further product approvals. The savings shared with customers reinforce the scale advantages that made those savings possible.

HEICO also exercises restraint in the amount of market share it pursues. Management has historically been satisfied with an attractive minority position—often around one-third of the market for an individual product—rather than trying to eliminate the original supplier. The OEM can continue supplying most of the volume at a substantially higher price, while HEICO serves the portion of the market willing and able to adopt an alternative.

This creates the possibility of rational coexistence. The airline receives meaningful savings, HEICO earns attractive economics, and the OEM retains most of a profitable aftermarket. An OEM considering an aggressive response must weigh the cost of reducing prices across its much larger installed base against the benefit of winning back the minority share held by HEICO. As long as HEICO does not threaten the entire market, the economics of a price war may be unattractive for the incumbent.

HEICO has reinforced that balance by generally avoiding the most valuable life-limited engine parts, which represent some of the crown jewels of the OEM aftermarket. Its restraint therefore operates in both directions: the company leaves substantial savings with customers and substantial volume with the original supplier. Rather than maximizing the economics of a handful of products, it has created a model that can be repeated across thousands of parts.

The resulting flywheel is straightforward. More approved products create more opportunities for customers to save. Greater savings encourage adoption and deepen the commercial relationship. Higher adoption produces volume, product ideas, operating data, and additional resources for engineering and inventory. Those capabilities allow HEICO to develop more products, which create further savings and strengthen the relationship again.

That is a much more durable proposition than simply selling a cheaper component.

Why Price Alone Is Not Enough

A discount of 30–40% sounds compelling, but aviation customers do not choose parts on price alone. The employee responsible for approving an alternative may generate substantial savings for the airline, yet that individual also bears the perceived professional risk if the product later creates a problem. Choosing the OEM is institutionally easy: it is the incumbent supplier, its part was installed originally, and few people are criticized for remaining with the established source. Choosing an alternative requires someone to make an active decision and accept responsibility for it.

This is why Victor Mendelson, one of HEICO’s current Co-CEOs and the long-time leader of its Electronic Technologies Group, has described it as extremely difficult to convince people to use an alternative part on an airplane. The financial benefit belongs to the airline, while the perceived downside may sit with the engineer, maintenance executive, or procurement professional who approved the switch. A lower price can begin the conversation, but credibility determines whether the sale actually happens.

HEICO has spent more than three decades reducing that perceived risk. It has built specialized engineering and regulatory capabilities, established quality systems across a broad manufacturing network, accumulated extensive operational experience, and developed long-standing relationships with airlines and maintenance organizations. The HEICO Parts Group has shipped more than 90 million parts without a Service Bulletin, Airworthiness Directive, or in-flight shutdown attributed to one of its products.

A new competitor may be able to hire engineers, purchase machinery, and obtain approval for an individual product. It cannot purchase a comparable operating history. Every year of successful performance adds to a record that can only be built over time, and every approved product creates further evidence that HEICO’s processes work beyond a single component.

The partnership with Lufthansa Technik played a decisive role in establishing that credibility. When Lufthansa Technik acquired a minority interest in HEICO’s aircraft-parts operations in 1997, one of the world’s most respected maintenance organizations placed its technical reputation and capital behind the company. Lufthansa could help identify parts that were expensive, difficult to source, or commercially attractive, while also providing operating knowledge and anchor demand once HEICO had completed the development and approval process.

The first HEICO part an airline approves may require considerable analysis, internal discussion, and technical documentation. Once that part has operated successfully, the company is no longer an unknown supplier. The customer has already reviewed HEICO’s quality systems, documentation, engineering capabilities, and support organization. Each subsequent approval can build on the credibility established by the products that came before it.

Catalog breadth magnifies this advantage. An airline may be reluctant to invest time in qualifying a small PMA manufacturer that offers only one or two products and therefore creates limited potential savings. HEICO can offer thousands of opportunities across engines, airframes, hydraulic and pneumatic systems, interiors, avionics, wheels and brakes, electromechanical equipment, and other categories. The technical and administrative effort required to establish the supplier relationship can be spread across a much larger pool of potential savings.

HEICO already serves virtually every major airline. Its remaining organic opportunity is therefore not primarily about adding customer names. It is about expanding the number of products each existing customer is willing to purchase. Every successful part increases confidence in the next one, while every newly approved product expands the portion of the customer’s maintenance spending in which HEICO can participate.

Availability adds another layer to the proposition. A lower-priced part is useful only if it can be delivered when required. When an aircraft is grounded, the economic cost of waiting can quickly exceed the price difference between competing components. HEICO therefore carries meaningful inventory and has built distribution capabilities that allow approved alternatives to be supplied at short notice. In periods of severe supply-chain disruption, immediate availability can become more important than the discount itself. A customer may first approve a HEICO part because the OEM cannot deliver, then continue buying it after the alternative has performed successfully and the higher OEM price becomes difficult to justify.

The moat is therefore not a single patent, approval, factory, or customer contract. It is the accumulated system of regulatory knowledge, engineering capability, quality control, operating history, product breadth, inventory, distribution, customer relationships, and institutional trust. Each element supports the others, and the system becomes progressively harder to reproduce as it grows.

From One Aircraft Part to Approximately 20,000

The HEICO recognized by investors today began taking shape when the Mendelson family assumed control in 1990. At the time, the company generated $26.2 million in revenue and $2.0 million in net income. Its aerospace operation was centered on essentially one core PMA product: a replacement combustion chamber used in a Pratt & Whitney aircraft engine.

That product demonstrated that an independent manufacturer could overcome the technical, regulatory, and commercial barriers required to compete with an established aerospace supplier. It did not yet prove that the opportunity could become a scalable business. HEICO still needed to expand its approval capabilities, gain the trust of conservative airline customers, establish repeatable manufacturing processes, hold inventory, and show that it could support products reliably over long periods.

The Mendelsons proceeded gradually. HEICO added further engine components, expanded into additional aircraft systems, developed repair capabilities, and used acquisitions to obtain technical expertise, customer relationships, manufacturing processes, and entry into adjacent product categories. Over time, the portfolio moved beyond its original engine focus into fuel, hydraulic, pneumatic, electromechanical, wheel and brake, airframe, interior, structural, and avionics products.

The acquisition of Wencor in 2023 materially increased the scale of the platform. Before the transaction, HEICO offered approximately 12,000 PMA parts and Wencor roughly 7,000, creating a combined portfolio approaching 20,000 approved products. HEICO now adds approximately 400–550 new PMAs each year and also offers thousands of DER-approved repairs, which allow damaged components to be restored rather than replaced with expensive new OEM products.

The number of products is important, but the process that continually creates them is the more valuable asset. Customers identify components suffering from high prices, poor availability, or an absence of competitive supply. HEICO evaluates the size of the market, the remaining life of the aircraft or engine platform, the technical complexity of the product, the likely customer demand, the incumbent price, and the development and inventory investment required. Promising candidates move through engineering, regulatory approval, manufacturing setup, quality validation, and commercial introduction.

Not every new product becomes important immediately. Airlines may require time to complete their own technical reviews, and adoption can build gradually as the part establishes an operating history. Product cohorts can therefore ramp for several years, while mature products continue generating recurring revenue from the installed fleet. HEICO’s organic growth reflects the layering of many such cohorts: older products continue selling, newer products gain adoption, and hundreds of additional approvals enter the portfolio each year.

This also explains why the opportunity can remain large despite the company’s long history. New aircraft and engine platforms enter service, existing platforms mature and move beyond initial warranties, OEM prices rise, supply problems emerge, technologies change, and customers identify additional components or repairs for which they want another source. The addressable opportunity continually renews itself.

What began with one combustion chamber has consequently become the world’s largest independent portfolio of FAA-approved jet-engine and aircraft-component replacement parts. Yet describing HEICO simply as a PMA company now understates the breadth of the business.

Two Operating Groups, One Economic Logic

HEICO operates through two principal segments. The Flight Support Group, or FSG, is the larger platform and the historical core of the company. It designs and manufactures PMA parts, repairs and overhauls selected aircraft components, distributes proprietary and third-party products, and manufactures specialized aerospace equipment. Depending on the customer’s situation, HEICO may provide a new replacement part, restore the existing component through an approved repair, source a needed OEM product through its distribution network, or combine several of these capabilities.

This broader offering matters because the customer’s underlying problem is not necessarily a desire to buy a PMA part. The customer needs an aircraft or component returned to service safely, quickly, and at a reasonable cost. By supplying parts, repairs, inventory, distribution, and technical support, HEICO can address that problem through several routes rather than forcing every situation into a single product category.

FSG generated approximately 70% of HEICO’s fiscal 2025 revenue and operating income. Its scale, catalog breadth, regulatory capabilities, and direct customer relationships make it the crown jewel of the company and the clearest expression of the HEICO model.

The Electronic Technologies Group, or ETG, receives less attention but provides a second growth engine. Its companies manufacture highly engineered components and subsystems used in aircraft, defense equipment, missiles, satellites, medical devices, telecommunications infrastructure, and other technically demanding applications. The portfolio includes products such as power supplies, antennas, microwave components, shielding, sensors, connectors, and other specialized devices that are often small relative to the value of the larger system in which they operate.

ETG reaches attractive economics through a somewhat different mechanism. FSG frequently challenges a highly priced incumbent by creating an approved alternative. ETG businesses often become the incumbent by having their component designed into a customer’s product or platform. Once the product has been selected, tested, qualified, documented, and incorporated into the larger system, replacing it may require redesign, additional testing, new documentation, customer approval, and regulatory or technical requalification. The possible cost saving from changing suppliers is often insignificant compared with the expense, delay, and risk involved in replacing a product that already performs reliably.

The two groups therefore appear diverse at the product level but share a common economic logic. HEICO focuses on small yet important products used in much larger systems, where the cost of failure is high and the component’s selling price is modest relative to the value of the aircraft, engine, missile, satellite, medical device, or other platform. Customers care about reliability, availability, documentation, qualification, and technical support far more than the cost of the underlying raw material.

This is why Laurans Mendelson did not primarily describe HEICO as an aerospace or electronics company. He regarded it as a vehicle for generating and reinvesting cash. The industries and products provide the setting, but the underlying objective is consistent: acquire or develop businesses with strong margins, durable demand, defensible customer positions, and opportunities to reinvest capital at attractive returns.

The Acquisition Flywheel

Organic product development forms one side of HEICO’s compounding model. Acquisitions form the other.

The company generally seeks specialized businesses with strong customer relationships, capable management teams, attractive margins, defensible niche positions, and opportunities for continued growth. Many targets are founder- or family-owned companies that have developed valuable products and technical expertise but remain too small to attract widespread attention.

After an acquisition, HEICO typically preserves the identity, management, employees, and operating autonomy of the business. Decisions about customers, products, engineering, manufacturing, and personnel remain largely with the people closest to the operation. Headquarters focuses on capital allocation, financial discipline, incentives, acquisitions, and the balance sheet rather than attempting to run dozens of specialized companies through a centralized corporate bureaucracy.

Sellers may retain a minority interest and continue managing the business. This allows an entrepreneur to achieve personal liquidity and diversify wealth without necessarily abandoning the company, employees, and customer relationships built over many years. The seller can continue participating in future value creation while gaining access to HEICO’s capital, acquisition expertise, public listing, and broader network.

HEICO also presents itself as a permanent owner. It does not generally acquire businesses with the intention of improving reported metrics for a few years and selling them to the next financial buyer. For founders who care about the future of their employees, the independence of the business, and the preservation of its identity, that distinction can matter as much as the highest immediate purchase price.

The seller proposition has become more credible with time. A founder considering HEICO can speak with entrepreneurs who sold their businesses many years earlier and ask whether management honored its promises. Each successful acquisition therefore becomes a reference for the next one. The larger and longer the track record becomes, the harder it is for a new buyer to reproduce.

This can generate proprietary or less competitive deal flow and allow HEICO to acquire businesses that might not choose a conventional strategic buyer or private equity firm. It does not eliminate competition, and HEICO will not win every attractive transaction, particularly when another buyer is prepared to pay a price that management considers excessive. However, it expands the factors on which the company can compete beyond valuation alone.

The parallels with the customer proposition are striking. HEICO does not attempt to capture every available dollar from airlines, and it does not attempt to extract every degree of control from acquired entrepreneurs. In both cases, management willingly leaves meaningful economics or autonomy with the other party because the resulting relationship may endure for decades.

HEICO’s long-standing refusal to maximize each individual transaction strengthens confidence in both components of its growth engine. Customers continue bringing the company expensive or difficult-to-source products they would like HEICO to develop. Entrepreneurs continue considering HEICO as a permanent home for the businesses they created. The acquisition pipeline and the organic product pipeline are therefore both supported by a reputation accumulated over many years.

This is particularly valuable because there is frequent, recurring demand for HEICO’s products. Aircraft continue flying, components continue wearing, maintenance events continue occurring, and customers repeatedly return for replacement parts, repairs, distribution, and technical support. Preserving the relationship is more valuable than maximizing the margin on one order. In a business built around repeated transactions over long product lives, retention is not merely a commercial metric. It is central to the economics.

Longevity is the cornerstone of compounding.

An Internal Flywheel Riding a Multi-Decade Tailwind

HEICO’s internal compounding system operates within one of the world’s most durable secular growth markets. Commercial air travel has expanded over many decades as populations, incomes, tourism, trade, and international connectivity have increased. The path has never been smooth. Recessions, terrorist attacks, wars, fuel-price shocks, and the COVID-19 pandemic have all caused severe temporary disruptions. Yet passenger demand has repeatedly recovered and resumed its long-term growth.

The current environment offers additional support to the aftermarket. Aircraft manufacturers and their suppliers have struggled to deliver enough new aircraft, forcing airlines to keep older models in service for longer than originally planned. Older aircraft generally require more maintenance, while the high utilization of the available fleet accelerates the accumulation of flight hours and cycles that trigger inspections, repairs, and replacement demand. Long OEM lead times also increase the value of an approved second source capable of delivering a part immediately.

These near-term conditions are favorable, but the longer-term installed-base development is more important. The global commercial fleet is expected to expand substantially over the coming decades as air travel grows, particularly in emerging markets where trips per capita remain far below developed-market levels. Each aircraft entering service creates a potential stream of parts, repairs, inspections, and inventory demand extending far beyond the original delivery.

New aircraft initially remain more closely tied to OEM warranties, service agreements, and approved maintenance networks. As those aircraft age, warranties expire, maintenance events become more frequent, and operators become more motivated to reduce costs. The addressable market for independent aftermarket suppliers therefore grows not only as more aircraft enter service but also as each generation of aircraft matures.

HEICO benefits from both time horizons. In the near term, delayed deliveries, aging fleets, high utilization, and parts shortages support demand for repairs and readily available alternatives. Over the long term, a larger installed fleet creates a broader base of aircraft and systems requiring maintenance for decades.

The financial record reflects the combination of these industry tailwinds with HEICO’s internal execution. Between fiscal 1990 and fiscal 2025, revenue increased from $26.2 million to approximately $4.49 billion, representing annual growth of roughly 16%. Net income attributable to HEICO rose from $2.0 million to approximately $690 million, equivalent to about 18% annually.

Growth has come from a combination of new products, increased adoption of existing products, market-share gains, acquisitions, and entry into adjacent niches. Importantly, HEICO has generally achieved this without relying on aggressive price increases or substantial equity issuance. Acquisitions have predominantly been financed with internally generated cash and debt, resulting in relatively limited dilution for long-term shareholders.

The business is also modestly capital-intensive in the conventional sense. Physical capital expenditure represented approximately 1.6% of fiscal 2025 revenue. HEICO does, however, commit meaningful capital to inventory because immediate availability is a central part of the customer proposition. It also expenses research and development through the income statement, meaning current earnings already bear the cost of product-development efforts that may generate revenue for many years.

The Mendelson family’s significant ownership reinforces the long-term orientation. Its financial outcome is primarily linked to the value of HEICO’s shares rather than the maximization of executive compensation or short-term transaction volume. Subsidiary managers and employees also participate through retained minority interests, performance incentives, and ownership of HEICO shares.

The result is a rare combination of organic product development, recurring aftermarket demand, programmatic acquisitions, attractive margins, modest physical capital requirements, aligned ownership, and a secular growth runway extending over multiple decades. Few businesses possess even several of these characteristics. HEICO has assembled them within one compounding system.

Understanding HEICO Requires Understanding the Industry

This Deep Dive is intentionally structured differently from a conventional company report. When I began researching HEICO, I did not yet have a deep understanding of the aerospace industry, and I quickly realized that the company could not be properly analyzed in isolation. Understanding why HEICO’s products are valuable—and why its competitive position is so difficult to replicate—first requires an understanding of the broader aviation industry, the aerospace value chain, aircraft maintenance, and the regulatory system governing every part installed on an aircraft.

I therefore expanded the research well beyond HEICO itself. The result is not only a company Deep Dive, but also a broader reference guide to the aerospace industry that can be revisited when analyzing other companies operating across aircraft manufacturing, aerospace supply, maintenance, repair, distribution, and the aftermarket.

HEICO itself is therefore not examined in detail until Part III. The first two parts establish the industrial, economic, and regulatory system in which the company operates.

• Part I (published on July 19) begins with the broader aviation industry. It examines long-term passenger growth, the current aircraft shortage, production backlogs, delayed retirements, fleet utilization, and the expected expansion of the global installed fleet. It then moves through the aerospace value chain, explaining how a modern aircraft is assembled from highly specialized systems and components and why aircraft and engine manufacturers rely on a deep network of suppliers. Finally, it establishes the regulatory framework governing design, production, maintenance, and continuing airworthiness.

• Part II turns to the aircraft aftermarket. It explains why maintenance demand exists, how maintenance, repair, and overhaul developed into a global industry, how aircraft checks and component repairs work, and where control and bargaining power sit across the value chain. It then examines PMA parts and DER-approved repairs in detail, including the regulatory approval routes, the airline economics supporting adoption, the price and availability proposition, and the practical barriers that still limit penetration.

• Only after that foundation has been established does Part III turn directly to HEICO. It covers the company’s development under the Mendelson family, the creation of the modern Flight Support Group, the customer proposition, the acquisition of Wencor, HEICO’s current market position, and the remaining organic runway.

• Part IV examines the Electronic Technologies Group and the HEICO Operating System, including ownership, incentives, decentralization, acquisition criteria, and the reasons entrepreneurs may choose HEICO over competing buyers.

• Part V brings the analysis together through HEICO’s acquisition history, financial development, capital intensity, competitive advantages, risks, and valuation.

Parts I and II are therefore not an extended detour before reaching the company. They are necessary to understand why the company exists. Without a working knowledge of aircraft lifecycles, certification, continuing airworthiness, maintenance economics, OEM aftermarket power, airline incentives, and the PMA approval process, it is difficult to appreciate why a small bag of apparently ordinary components can cost $50,000—or why offering the airline an approved alternative for substantially less can become the foundation of an extraordinary business.

Most company reports begin with HEICO’s revenue, margins, acquisitions, and valuation multiples. Those figures describe what the company has achieved, but they do not fully explain why it has been possible.

To my knowledge, no other publicly available HEICO Deep Dive develops the underlying aerospace ecosystem in comparable depth.

That context ultimately reveals what makes HEICO unusual. It is not simply an aerospace supplier, a PMA manufacturer, an electronics group, or a serial acquirer. It is a carefully constructed system that converts customer cost pressure, regulatory complexity, recurring maintenance demand, technical credibility, entrepreneurial autonomy, and disciplined capital allocation into a flywheel capable of compounding over decades.

Stay tuned for Part 1 of the upcoming deep-dive!

One of the hardest things in investing is recognizing productive reinvestment before it becomes visible in the reported financials. In real time, it rarely looks clean. Revenue may still be growing, but margins compress. Operating income may stall or decline. Analysts lower near-term estimates. Investors begin to question whether management is being disciplined or simply spending too much. The market often struggles to distinguish between a company that is spending defensively to protect a weakening position and a company that is deliberately sacrificing near-term profits to strengthen its long-term competitive position.

This distinction is especially difficult in e-commerce. A platform can spend on shipping, fulfillment, vouchers, technology, seller tools, and customer benefits for years before the payoff becomes obvious. Some of that spending may create durable value by increasing frequency, loyalty, density, and monetization. Some of it may simply subsidize transactions that would disappear as soon as incentives are reduced. The P&L alone rarely tells you which is which at the beginning.

So let us begin with a thought experiment.

The thought experiment

Imagine a young e-commerce company that started from a very small base and then grew at extraordinary rates. Just three years after its founding, the company went public. At the time, it was still tiny, generating only $148 million in net sales. Within a decade of its founding — and seven years after going public — net sales had grown to almost $7 billion. The business was still mostly first-party: it sold products directly to consumers, carried inventory, invested heavily in fulfillment capacity, and absorbed the operating complexity of building a large online retailer from scratch.

Gross profit increased from almost nothing to roughly $1.6 billion over the first eight reporting periods, while the gross profit margin moved from the high-teens into the low-to-mid twenties.

However, almost all of this expanding gross profit was being consumed below the gross profit line. Fulfillment, technology, marketing, and administrative expenses remained heavy because the company was still building the infrastructure required to support a much larger business. This is why the company could show clear gross profit progression while still reporting substantial operating losses. EBIT remained negative for several years, and the operating margin only turned positive once the revenue base had grown large enough to absorb the investments made in earlier periods.

By year 6 after its IPO, the company turned EBIT positive. By year 8 after its IPO, the company generated approximately $440 million of EBIT on $6.9 billion of net sales, equivalent to a 6.4% operating margin. For a business that many investors had previously dismissed as structurally unprofitable, this looked like an important proof point.

There was, however, one important cost item that already hinted at how management thought about the business: shipping.

The company was willing to absorb a negative gross profit contribution from shipping in order to make the customer proposition more attractive. Shipping revenue increased over time, but outbound shipping costs grew even faster. Beginning in the early 2000s, shipping became a clear gross profit headwind, with the gap between what customers paid for shipping and what the company incurred in outbound shipping costs widening materially.

This is an important detail because it shows that the company was not treating shipping as a pass-through cost to be fully recovered from customers. Management was already using shipping economics strategically. The company was effectively accepting lower reported gross profit in exchange for a better customer experience, higher conversion, and greater purchase frequency: Use logistics and shipping convenience as a way to deepen the customer relationship, even if that meant depressing near-term profitability.

At that point, a short-term investor might have expected management to let EBIT-margins expand. The company had spent years proving the model. The obvious next step seemed to be harvesting operating leverage.

Instead, management launched a membership program.

The customer proposition was extremely attractive: pay a fixed annual fee and receive a materially better shipping experience. From the customer’s perspective, this made the platform more useful. From the company’s perspective, it created a new cost burden. Shipping became even cheaper and easier for the customer precisely because the company absorbed more of the cost itself.

The following year, management added another major logistics initiative, this time designed to allow third-party sellers to place inventory inside the company’s fulfillment network. The two initiatives reinforced each other strategically, but they also made the reported financials more difficult to interpret.

The P&L immediately reflected the reinvestment.

Operating margins declined from 6.4% to 3.6% within two years. Operating income declined in absolute dollars from roughly $440 million to approximately $389 million, even though revenue continued to grow from $6.9 billion to $10.7 billion.

This is exactly the kind of situation the market often dislikes. Revenue was still growing, but margins were contracting as management reinvested to accelerate the flywheel and strengthen the moat. The company had only recently demonstrated profitability, and management was already choosing to give some of it back. Investors had to decide whether this was disciplined long-term investment or another example of an e-commerce company sacrificing profitability for growth.

The stock reflected that uncertainty. During this investment period, the share price fell sharply, with a maximum drawdown of roughly 45% from peak to trough. To a skeptical investor, the conclusion was easy: the company had reached a local margin peak, competition and shipping costs were pressuring the model, and management was once again prioritizing growth over profits.

Yet the market’s interpretation changed quickly once the financial evidence began to improve. As operating margins recovered and revenue growth accelerated, the share price rebounded from roughly $2 to almost $5 at the peak in the following year. What had looked like margin-destructive spending began to be reinterpreted as the upfront cost of building a stronger customer relationship, a denser logistics network, and a more powerful marketplace flywheel.

Would you have bought the stock?

The membership program was Amazon Prime, launched in early 2005. The seller logistics initiative was Fulfillment by Amazon (FBA), launched in 2006. In hindsight, both look obvious. At the time, they did not. They looked like expensive reinvestments in a business that had only recently convinced the market it could earn money at all.

Two decades later, the scale of what emerged from that supposedly expensive reinvestment cycle is difficult to reconcile with how uncertain the setup looked at the time. In 2025, Amazon generated $716.9 billion in net sales, of which $128.7 billion came from AWS. The remaining $588.2 billion came from Amazon’s non-AWS businesses — a broad mix of first-party retail, third-party marketplace services, advertising, subscriptions, physical stores, and other consumer-facing activities. Consolidated operating income reached $80.0 billion, equivalent to an 11.2% operating margin. On a gross merchandise value basis, Marketplace Pulse estimates that Amazon and its sellers processed roughly $830 billion of GMV in 2025.

Prime has become one of the central pillars of this ecosystem. Amazon does not disclose Prime membership regularly, but the last official figure was already more than 200 million members globally several years ago. In the U.S. alone, CIRP estimates that 201 million customers used a Prime membership, while eMarketer had previously projected roughly 90 million U.S. Prime households by 2025 (~7 out of 10 households). From the end of 2007, when the Prime/FBA flywheel was beginning to show up more clearly in the numbers, through the end of 2025, Amazon’s split-adjusted share price compounded at roughly 24% per year.

The combination of Prime and Fulfillment by Amazon had started to reinforce the broader marketplace flywheel. Prime increased purchase frequency and made Amazon the default starting point for more customers. This habit formation was central to the economics of the model. In e-commerce, the most valuable customer is not necessarily the one who places the largest single order, but the one who starts more shopping journeys on the platform. Once a marketplace becomes the default starting point, it no longer has to win each transaction from scratch. More frequent customer visits generate more data, improve personalization, increase advertising opportunities, raise seller value, and create the order density required to make fulfillment more efficient. Prime was powerful because it moved Amazon closer to that position: not just a place where customers occasionally bought products, but the place they instinctively checked first.

Need something? You check Amazon first. That sounds simple, but for an e-commerce platform it is one of the most valuable positions imaginable. It means the company has moved upstream in the purchase journey. It is no longer merely competing at the moment of transaction; it is shaping where the transaction begins.

Additionaly, FBA improved the seller-side proposition by allowing third-party merchants to offer a better delivery experience. → Better delivery improved the customer experience. → A better customer experience made Prime more valuable. → More Prime members created more order volume. → More order volume increased fulfillment density. → Greater density helped justify further logistics investment. → The flywheel starts to kick in.

This is the historical pattern that matters for Sea Limited today.

Not because Sea is Amazon.

Not because Shopee VIP is guaranteed to become Amazon Prime.

And not because Southeast Asia or Brazil should be treated as identical to the United States.

The comparison is useful as it shows how an e-commerce platform can deliberately accept near-term margin pressure in order to build a more valuable customer relationship, a denser logistics network, and a stronger marketplace flywheel.

The question is whether Shopee VIP can play a similar strategic role inside Sea’s ecosystem.

Sea’s Current Setup: A Profitable Platform Re-Entering Investment Mode

To answer that question, it is important to start with the setup Sea finds itself in today.

Shopee has already gone through one major profitability reset. After the post-pandemic boom, the market’s central concern was whether the company could ever turn its enormous GMV base into a sustainably profitable business. For several years, Shopee had been valued primarily on growth, scale, and market leadership. Then, as capital became more expensive and investors demanded profitability, the question changed. Growth alone was no longer enough. The company had to prove that the model could work economically.

Sea responded by heavily cutting incentives, rationalizing spending, improving monetization, and demonstrating that Shopee could generate positive earnings.

As a result, growth slowed, but the business crossed into profitability, albeit at a level still well below what a mature Shopee could potentially earn.

As competition from TikTok Shop became a more prominent concern, Sea chose not to maximize near-term margins. Rather than harvesting operating leverage, management elected to reinvest more aggressively into growth initiatives (again).

The company is investing logistics, fulfillment, and Shopee VIP. This is why the current moment is so interesting. Shopee is no longer an unproven e-commerce platform burning cash to acquire users. It is a much larger, more mature marketplace that has already demonstrated profitability and is now deliberately choosing to spend part of that profit pool to strengthen the ecosystem.

The market’s concern is understandable. Investors had only recently become comfortable with Shopee’s ability to earn money. A renewed investment cycle naturally raises the fear that margins were temporarily inflated, competition is intensifying, and the business now requires a higher level of spending simply to defend its position.

That is the defensive spend view.

There is, however, another interpretation. Sea may be using its scale and profitability to move from a transactional marketplace toward a more habitual commerce platform. In that view, Shopee VIP is not merely a discount program and Fulfilled by Shopee (FBS) is not merely a logistics cost center. Together, they are an attempt to increase purchase frequency, improve fulfillment reliability, create more order density, and make Shopee the default shopping destination for a larger share of users.

This is the offensive reinvestment view.

The distinction matters because the two scenarios can look almost identical at the beginning. In both cases, margins compress. In both cases, near-term earnings estimates may come down. In both cases, investors worry that management is giving up profitability just as the business was starting to scale. The difference only becomes visible later. Defensive spend preserves the existing business at a lower margin. Offensive reinvestment expands the future profit pool by making the platform more valuable to customers, sellers, and advertisers.

That is the same analytical problem investors faced in the earlier case study. The Prime membership product and the FBA buildout initially made the financials look worse, even though they eventually helped the platform become more valuable. The question for Sea is whether Shopee VIP and FBS are creating a similar improvement in customer behavior and ecosystem density, or whether they are simply subsidizing transactions in a more competitive market.

What Shopee VIP Is Trying to Change

The strategic importance of Shopee VIP is not that it gives users another set of discounts. E-commerce platforms have always used promotions, vouchers, and free shipping to stimulate demand. The more important question is whether VIP changes the structure of the customer relationship.

A normal marketplace relationship is transactional. A customer opens several apps, compares price, delivery time, seller reputation, and available vouchers, and then chooses where to buy. In that world, the platform has to win every transaction again and again. Even if the marketplace is large, customer loyalty can remain shallow because users are constantly comparing alternatives.

A membership product tries to change that behavior. It gives the customer a reason to start the shopping journey inside one ecosystem. If the benefits are used frequently enough, the platform becomes more habitual. The user does not necessarily stop comparing forever, but the starting point changes. That is the key.

For Shopee, this matters because the company operates in markets where competition is intense, average order values can be relatively low. TikTok Shop, Lazada, Tokopedia, Mercado Livre in Brazil, local players, social commerce channels, and offline retail all compete for consumer attention and wallet share.

Shopee VIP is designed to move the business in that direction. Free shipping with no minimum spend reduces friction for smaller, more frequent purchases. Daily vouchers create a reason to return to the app more often. The membership structure gives users a recurring relationship with the platform rather than a purely transactional one. If this works, VIP can increase order frequency, improve retention, and deepen Shopee’s position in the consumer’s shopping routine.

The Early Evidence: Small Penetration, Outsized GMV Contribution

Shopee VIP is still early, and the available data should be interpreted carefully. Some of the numbers rely on estimates, and the time period is short. It would be a mistake to treat the current data as proof that VIP will become a Prime-like product. At this stage, the data is better understood as a set of early signals.

Those signals are nevertheless highly interesting (and promising).

Sea first piloted the VIP membership program in Indonesia in the first quarter of 2025, targeting some of Shopee’s most loyal users with a paid membership product that included more generous free shipping benefits, larger discount vouchers, and priority customer service, as Indonesia is Shopee’s largest and most strategically important market in Southeast Asia, with enough scale to test whether a membership product could change behavior in a meaningful way.

The early response was strong enough for management to expand the program quickly. By the end of the second quarter, Shopee VIP had been rolled out to Thailand and Vietnam as well. Over the rest of 2025, the program was expanded further across Shopee’s Asian markets. By year-end, subscribers had surpassed 7 million, and management noted that VIP members were consistently showing a double-digit spending uplift after joining.

After proving the concept in Asia, Sea launched Shopee VIP in Brazil in April 2026, extending the membership model into its most important Latin American growth market.

Shopee VIP members increased from approximately 1.5 million in Q1 2025 to around 10.0 million in Q1 2026. That is a rapid adoption curve for a program that is still in the early stages of scaling — and notably, the Q1 2026 figure does not yet include any contribution from Brazil, where the rollout only began afterward.

By comparison, Amazon Prime appears to have reached roughly 10 million members only by the end of 2012, around seven years after its 2005 launch (although Amazon did not officially disclose Prime membership at the time, so this figure is based on third-party estimates). Shopee reached that same milestone in roughly one year.

However, the penetration is still low relative to Shopee’s overall buyer base. Based on estimates, Shopee had ~418 million active buyers in Q1 2026. Against that base, 10 million VIP members would imply only around 2.4% penetration.

The GMV contribution appears far more significant than the membership penetration would suggest.

“By the end of March, total subscribers across our Asian markets surpassed the 10 million, up more than 40% from the previous quarter, with strong program retention averaging above 80%. Across all markets, our Shopee VIP members have consistently demonstrated double digit spending uplift after subscribing by as much as 30%-40% in some markets. Shopee VIP members now contribute around 20% of GMV across Asia. Building on this success, we have rolled out our Shopee VIP program in Brazil in April.” — Forrest Li (CEO), Q1 2026 earnings call

In other words, a very small share of buyers may already be driving a disproportionately large share of marketplace activity.

This is probably one of the most important charts in the article, but it also requires the most careful interpretation. VIP members are almost certainly not average Shopee users. Heavy users are more likely to join a membership program in the first place because they can extract more value from free shipping and vouchers. Therefore, part of the apparent outperformance reflects selection bias. VIP is not necessarily turning average users into power users overnight; it is likely attracting Shopee’s existing power users first. In that sense, the early adoption curve may partly reflect the low-hanging fruit: the users for whom the value proposition was most obvious from day one because they were already ordering frequently enough to justify the membership.

However, that does not make the data unimportant. In fact, it may be strategically desirable. A marketplace’s most valuable customers are its high-frequency, high-trust, high-retention customers — the classic Pareto principle applied to marketplace behavior. If VIP allows Shopee to identify, retain, and further monetize these users, that alone has value. The stronger version of the thesis is that VIP not only captures existing power users, but gradually increases the frequency of users who would otherwise have ordered less often.

My estimates suggest that quarterly GMV per VIP member is dramatically higher than GMV per non-VIP buyer. The exact ratio should not be treated as a precise metric, because the calculation relies on a handful of undisclosed metrics. But the directional message is difficult to ignore: VIP members are economically much more valuable than the average non-VIP buyer.

Sea’s own disclosure provides a useful sanity check for this estimate. On the Q3 2025 earnings call, management noted that Shopee VIP members in Indonesia “bought three times more frequently and spent five times more than non-subscribers” during the quarter, while accounting for about 10% of total GMV in Indonesia.

Why Frequency Is A Key Variable

Frequency determines how often the customer relationship is “refreshed”. A user who buys once per year has limited strategic value. A user who buys several times per month is very different. That user generates more GMV, more data, more opportunities for advertising, more seller value, and more logistics density. Over time, frequency can become one of the most important drivers of marketplace quality.

This is why membership programs can be so powerful when they work. They are not merely monetization tools; they are behavior-shaping tools. A well-designed membership product reduces the friction of placing incremental orders and gives users a reason to return more often. The more frequently the customer returns, the more natural it becomes to start future shopping journeys on the same platform.

For Shopee, this is particularly important because the company does not operate in a winner-take-all environment. Consumers can compare prices, search on other marketplaces, discover products through TikTok, buy through social channels, or return to offline retail. Shopee’s challenge is to become the most convenient, trusted, and habitual commerce destination.

Shopee VIP directly targets that challenge. Free shipping with no minimum spend lowers the barrier for small orders. Daily vouchers increase the incentive to open the app regularly. If those benefits create a loop in which users check Shopee more often, buy more frequently, and build a stronger preference for the platform, then VIP can improve the quality of Shopee’s GMV, not just the quantity.

That distinction is important. Subsidized GMV is low quality if it disappears once incentives are reduced. Habitual GMV is much higher quality because it reflects a deeper customer relationship.

This is also where management’s disclosure on post-subscription behavior becomes important. Sea noted that across markets, Shopee VIP members have consistently shown a double-digit spending uplift after subscribing, reaching as much as 30–40% in some markets. That does not eliminate the selection effect — VIP is still likely attracting Shopee’s existing power users first — but it suggests that the program is not merely identifying high-value customers. It appears to be increasing their spending after they join.

VIP and FBS: The Buyer-Side and Supply-Side Flywheel

Shopee VIP should not be analyzed in isolation. Its strategic value depends partly on how it interacts with Shopee’s logistics and fulfillment investments.

The buyer-side logic is straightforward. VIP makes the platform more attractive to consumers by reducing shipping friction and increasing the perceived value of frequent purchases. This can increase order frequency and create more predictable demand.

The supply-side logic comes through FBS. As more items are fulfilled through Shopee-controlled infrastructure, Shopee can improve delivery reliability, service consistency, and potentially cost efficiency. Better fulfillment improves the customer experience, which makes VIP more valuable. More VIP members create more order volume, which improves logistics density. Greater logistics density can justify more fulfillment investment, which further improves service quality.

This is the potential flywheel:

Amazon Prime created the customer demand for fast, reliable, low-friction shipping. FBA helped sellers participate in that promise. The two initiatives reinforced each other. Prime without fulfillment capability would have been difficult to scale. Fulfillment investment without Prime would have had less demand density behind it.

Shopee VIP and FBS may develop in a similar direction, although the economics will differ by region. Shopee’s markets generally have lower average order values and more fragmented logistics environments than Amazon’s early U.S. market. That makes the execution challenge harder, but it also makes density more valuable. If Shopee can aggregate demand, increase fulfillment control, and improve delivery reliability in markets where logistics are structurally difficult, that can become a meaningful competitive advantage.

Prime vs. VIP: Where the Analogy Holds

I’m not making the point that Shopee VIP will become as large or as powerful as Amazon Prime. The point is that both programs are attempts to form the consumer relationship from occasional and transactional to (even more) frequent and (even more) habitual.

Several similarities stand out.

Both programs were launched or scaled after the platform had already established meaningful market position. Neither was the first step in building the marketplace. Instead, each came after the company had already accumulated scale and could begin using that scale to improve the customer proposition.

Both programs created near-term pressure on reported profitability. Customer benefits, free shipping, fulfillment investment, and logistics infrastructure are expensive. The strategic logic may be strong, but the accounting impact appears first through the P&L.

Both programs were connected to fulfillment. Amazon Prime and FBA reinforced each other. Shopee VIP and FBS may do the same. A membership promise becomes more powerful when the platform can control delivery quality, and fulfillment infrastructure becomes more valuable when there is enough recurring demand to support it.

Both programs aim to increase frequency. A membership product that does not increase frequency is far less valuable. A membership product that makes the platform the first place customers check can change the long-term economics of the marketplace.

And both programs were introduced during periods when the market was uncertain about the payoff. In Amazon’s case, investors had to underwrite Prime and FBA before the flywheel was obvious. In Sea’s case, investors now have to decide whether VIP and FBS are defensive spending or offensive reinvestment.

Early Trajectory

Viewed through that framework, Shopee’s early trajectory is striking. On several of the early indicators we can observe, Shopee appears to be moving faster than Amazon did at a comparable stage. VIP reached roughly 10 million members in about one year, while Amazon Prime appears to have taken roughly seven years to reach a similar level based on third-party estimates. Shopee is also scaling VIP alongside FBS at a time when its GMV growth remains strong and the business has already demonstrated profitability.

Where the Analogy Could Break

The comparison also has clear limitations.

The first and most important difference is the nature of the value proposition. Amazon Prime was primarily built around convenience, especially fast and free shipping. Shopee VIP appears to rely more heavily on a combination of free shipping and daily vouchers. That matters because convenience-driven loyalty is usually more durable than discount-driven loyalty. A customer who stays because the platform is faster, more reliable, and easier to use is more valuable than a customer who stays only because today’s voucher is attractive.

The second difference is market structure. Amazon’s early Prime years benefited from rising U.S. e-commerce penetration, improving household internet adoption, and a relatively attractive logistics backdrop compared with many emerging markets. Shopee operates across Southeast Asia and Brazil, where average order values, payment methods, delivery infrastructure, geography, and consumer income levels can vary widely. Free shipping is harder to monetize when order values are low, and fulfillment density has to be earned market by market.

The third difference is competition. Amazon competed with traditional retailers and other e-commerce players, but Shopee also faces social commerce such as TikTok Shop. That changes the competitive battlefield. Shopee’s advantage must come from trust, selection, value, transaction reliability, fulfillment, and habit. VIP can help reinforce those advantages, but it does not eliminate the threat from entertainment-led commerce.

The fourth difference is the maturity of the broader ecosystem. Amazon eventually layered Prime on top of a business that would also benefit from a rapidly growing third-party marketplace, advertising, AWS, and an expanding logistics network. Sea has its own ecosystem assets, including Shopee, SeaMoney, and Garena, but the shape of the long-term monetization stack will not be identical.

What the Market May Be Missing

The market often struggles with reinvestment cycles because the costs are visible before the benefits are measurable. The expense shows up immediately in margins. The benefits show up later in retention, frequency, fulfillment efficiency, merchant demand, advertising monetization, and long-term earnings power.

Sea’s current valuation appears to reflect considerable skepticism. Investors see TikTok Shop competition, renewed spending, and lower near-term margins. They may be less willing to underwrite the possibility that Sea is reinvesting from a position of strength rather than weakness.

This distinction matters. Sea today is not the same company it was during the most painful part of the post-pandemic reset. Shopee has already demonstrated that it can generate profits. The marketplace is larger, more developed, and more deeply embedded in its core regions. Management now has a stronger base from which to invest.

If Sea were merely increasing subsidies to protect share, skepticism would be justified. But if the company is using its scale to build a stronger frequency and fulfillment loop, the market may be valuing the investment cycle incorrectly.

The potential upside does not require Shopee VIP to become Amazon Prime. It only requires VIP to improve the slope of Shopee’s core variables: frequency, retention, logistics density, and monetization. Even a modest improvement in those variables can have a large impact when applied to hundreds of millions of buyers and a very large GMV base.

What Needs to Be Proven

The coming quarters should be evaluated less on one single margin number and more on whether the strategic indicators continue to improve.

The first metric to watch is VIP penetration. A program with roughly low-single-digit penetration still has significant room to grow. If VIP penetration continues to increase while maintaining strong engagement, the program becomes more relevant to the overall Shopee ecosystem.

The second metric is VIP GMV share. If VIP members continue to account for a disproportionate share of GMV, Shopee may be concentrating more activity among its highest-value users. The composition of that GMV will matter, but the direction is important.

The third metric is frequency. This is the most important behavioral indicator. VIP only becomes strategically powerful if it increases order frequency and keeps users engaged over time.

The fourth metric is retention. A membership product can look strong early if it attracts heavy users, but the long-term value depends on whether those users remain members and continue to consolidate spending inside the platform.

The fifth metric is FBS penetration. VIP and fulfillment should ideally scale together. If more orders are fulfilled through Shopee-controlled infrastructure, the company may gain more control over delivery reliability, speed, and service consistency.

The sixth metric is logistics cost per order. The long-term case depends on density. If more frequent ordering improves cost efficiency, then near-term logistics investment can become a source of future operating leverage.

The seventh metric is competitive position. If Shopee continues to grow faster than the broader e-commerce market while TikTok Shop growth moderates, the defensive-spend bear case becomes less convincing.

Together, these indicators should help answer the central question: is Shopee VIP merely subsidizing transactions, or is it changing the structure of the customer relationship?

Subscribe now

Valuation: back to IPO-level pessimism

Measured on forward EV/Sales, the stock is now trading close to the lowest level in its public market history. EV/Sales is the more useful long-term valuation lens because Sea was not sustainably profitable before 2023, making a historical EV/EBITDA comparison less meaningful across the full period.

The contrast is striking. During the 2020–2021 growth bubble, Sea traded at more than 15x forward sales. Today, the multiple is closer to 1.5x, roughly back to the lowest levels seen since the IPO (also seen early 2024).

The market is no longer paying for a heroic growth story. It is assigning Sea a valuation that implies much lower expectations, despite Shopee being larger, more efficient, more entrenched, and closer to sustainable profitability than it was during earlier parts of its history.

If Shopee VIP helps increase order frequency, deepen customer loyalty, improve retention, and shift a greater share of spending into Shopee’s ecosystem, the market may eventually have to reassess not only the growth profile, but also the quality and durability of that growth.

At today’s valuation, the stock does not need a return to 2021-style optimism to work. It needs evidence that Shopee can compound profitably from here.

Conclusion

Shopee VIP is still early, and the Amazon Prime analogy should be treated as a framework rather than a forecast. Sea is not Amazon. Shopee VIP is not Prime. Southeast Asia and Brazil are not the United States. The logistics, competitive, and consumer behavior realities are different.

Still, the early signs are worth paying attention to.

VIP adoption has grown quickly. Penetration remains low. VIP members appear to account for a disproportionate share of GMV. The program is directly connected to frequency, shipping, vouchers, fulfillment, logistics density, and merchant monetization. Most importantly, it is being scaled at a moment when the market is worried that Sea’s reinvestment is defensive.

That creates the potential for misunderstanding.

If Shopee VIP proves to be mostly a discount program, the market’s skepticism will likely be justified. But if VIP becomes a habit-forming product that increases frequency, retention, and logistics density, then Sea’s current investment cycle could be much more attractive than the near-term margin debate suggests.

In the earlier case study, the membership product and the logistics initiative made the financials look worse before the strategic payoff became obvious. Margins compressed, operating income declined, and investors had to underwrite a flywheel that was not yet fully visible. Once the evidence began to turn, the market’s interpretation changed quickly.

Sea may be approaching a similar type of moment.

The right question is not whether Shopee VIP will become as important as Amazon Prime. That is unknowable and probably too ambitious as a direct comparison. The better question is whether Shopee VIP can change the slope of Shopee’s frequency, loyalty, fulfillment, and monetization curve.

If it can, the market may be looking at Sea’s reinvestment cycle from the wrong angle.

What looks like lower margins today may eventually prove to be the cost of building a more durable and more profitable platform.

Since 2008, Amazon’s stock has compounded at a mid-20s annual rate.

But the harder question is what Amazon looked like before it became obvious.

Around 2008/2009, Amazon was already a large company. It had billions in revenue, a dominant position in U.S. eCommerce, and a founder-CEO with a very long-term mindset. But many of the value drivers that define Amazon today were still either small, invisible, or not fully appreciated by the market. AWS was still early. Advertising was not yet the profit machine it later became. The marketplace was still maturing. Logistics looked expensive before it became a moat. And profitability was not yet the clean, obvious story investors can see in hindsight.

Platform companies often look messy. They invest heavily, profitability is delayed, and the market struggles to decide whether losses are value-destroying expenses or value-creating reinvestments. Only later does it become clear whether the company was simply burning cash or building infrastructure, trust, customer behavior, and network effects that would matter for decades.

Sea Ltd. is not Amazon. And Southeast Asia is not the United States.

But in some ways, Sea reminds me of Amazon around 2008/2009: already large, already relevant, but with much of its long-term earnings power still hidden behind investment, complexity, and market skepticism.

Is Sea just a volatile internet company with hard-to-forecast earnings? Or is it an earlier-stage platform compounder in a region that is still moving through a long digital consumption cycle?

I have followed the company from a distance since around 2020. Back then, Sea Ltd. was almost impossible to ignore. Garena was printing cash, Shopee was scaling at breathtaking speed, SeaMoney was viewed as a powerful fintech option, and the market was willing to dream very far into the future.

Then the story collapsed. The stock fell sharply, Garena normalized, Shopee had to prove that it could become profitable, and Sea was forced to move from growth-at-any-cost to self-funded growth.

I never fully forgot the company. But I also never really put it into the typical Slow Compounding bucket.

Sea is not the classic “boring” compounder I usually like to study. It does not have decades of smooth earnings growth, steadily expanding margins, or a proven playbook that has already worked through multiple cycles. The range of outcomes is wider. The competitive environment is tougher. The moving parts are more complex.

But sometimes the market’s discomfort is exactly where the opportunity begins.

Today, Sea consists of three very different businesses:

🛒 Shopee, the leading eCommerce platform in Southeast Asia and Taiwan (and Brazil), and the core of the long-term investment thesis.

💳 Monee, the fintech layer on top of the Shopee ecosystem, with digital payments, credit products, and financial services.

🎮 Garena, the historical cash flow engine, still profitable and still relevant, even if it is no longer the center of the equity story.

The key question is whether these pieces can compound together.

Shopee already has massive scale, but its mature profitability is still not fully visible. Monee is growing quickly, but brings credit risk. Garena remains highly profitable, but gaming is always hit-driven. None of this is as simple as buying a niche industrial compounder and watching EBITA margins slowly expand for the next decade.

That is also why this is not a full Slow Compounding Deep Dive yet. It is a Company Snapshot. Some Substackers might already call an article like this a Deep Dive, but for Slow Compounding, a real Deep Dive means going much deeper.

So think of this article as the first serious look.

We will walk through:

🌏 Why Southeast Asia matters — and why the region is still structurally underpenetrated

🛒 Why Shopee is the core asset — and why its take rate, GMV growth, and margin potential matter so much

💳 Why Monee could become a second earnings engine — but also the segment where risk management matters most

🎮 Why Garena still matters — even though Sea is no longer mainly a gaming story

📉 Why the stock has been so volatile — and why the market still seems scarred by the 2021/2022 collapse

💰 What the current valuation might imply — especially for Shopee’s long-term earnings power

The more I look at the company, the harder it becomes to ignore. Sea is not boring. It is not easy. But it sits at the intersection of eCommerce, fintech, digital entertainment, and the long-term rise of Southeast Asian (and Latin American) consumption.

Subscribe now

1. Why Sea Ltd. is interesting today

Sea Ltd. is not the kind of classic “boring” compounder I usually look for. There are plenty of reasons not to like Sea at first glance. The company does not have a decades-long history of steadily rising margins or exceptionally smooth earnings development. The stock has behaved more like a roller coaster in recent years, making Sea one of the most volatile large internet stocks of the past few years.

During the last years, the market has placed the company into almost every possible bucket:

Over the past few years, the market has pushed Sea through almost every possible narrative. It was first celebrated as a hypergrowth story, then dismissed as a cash burner. At the bottom, investors worried about insolvency risk, a massive capital raise, and potentially heavy dilution. Then came the turnaround phase. As profitability returned, the narrative shifted again: Sea was no longer just a survival story, but a platform with renewed earnings momentum. Today, the debate is changing once more.

Many investors have therefore already mentally filed Sea away as “too risky,” “too volatile,” or “too difficult to understand.” This is probably reinforced by a simple geographic disconnect: Sea is listed in the U.S., but much of its real story is playing out in Southeast Asia, a region many U.S.-based investors do not follow closely.

That is exactly where the opportunity might lie today.

The share price still carries the scars of the earlier euphoria and the collapse that followed.

Under the surface, Sea has changed significantly over the past few years. What started as a growth story funded by the gaming business has increasingly become a more diversified consumer internet platform. At its core today is Shopee, the leading eCommerce platform in Southeast Asia and Taiwan (and Brazil). At the same time, Monee is emerging as a digital financial services business closely tied to the Shopee ecosystem and could become a second major earnings driver over time. Garena, the gaming business, is no longer the sole engine of the equity story, but it remains an important part of Sea’s history and cash flow profile.

That is what makes the investment case unusual. Sea still offers substantial revenue and earnings growth potential, while the valuation still does not seem to fully reflect the company’s long-term earnings potential. As I will argue toward the end of this article, the setup looks quite compelling to me: the downside appears low, while the upside could be meaningful.

The company is now much larger, more mature, and more profitable than it was during the earlier hypergrowth phase, yet the stock still reflects a lot of skepticism. The market does not seem to value Sea as a stable long-term compounder. Instead, it still looks at the company through the lens of the past: volatile gaming revenues, heavy eCommerce losses, aggressive competition, abrupt strategy shifts, and extreme valuation cycles.

That is what makes Sea interesting to me. The company does not yet look like a perfect quality compounder. It is in its earlier phase of that journey. Shopee addresses a huge eCommerce market that is still underpenetrated and that grows strongly. Monee can build on top of this ecosystem. And Southeast Asia itself provides structural tailwinds from rising purchasing power, increasing digitalization, young demographics, and growing online penetration.

A cautious comparison would be Amazon around 2009. Even then, Amazon was no longer a small company. In 2009, Amazon generated $19.2 billion in revenue, still mostly from eCommerce. Growth was high, but profitability was not yet what later investors would look back on as obvious. Many of today’s value drivers — AWS, advertising, marketplace density, logistics advantages, and structurally higher margins — were already there in some form, but not yet fully visible. Investors who judged Amazon only by its near-term margins or its seemingly high multiple underestimated the long-term economics of the platform.

Since 2008, Amazon’s stock has compounded at roughly 25% per year.

With revenue of $16.6 billion in 2025, Shopee has therefore reached a similar revenue scale to Amazon in 2008/2009, while its margin profile still reflects a platform that is far from fully mature.

Sea is not “the next Amazon,” and Southeast Asia is not the United States. That comparison would be too simplistic.

The parallel lies in the nature of the question: Is this a volatile internet stock with hard-to-forecast earnings, or an early platform whose long-term earnings power is still not fully understood by the market?

Put differently: Investors who missed Amazon around 2010 may not get the same story again with Sea Ltd. But they may get a similar type of opportunity: an already large, yet still not fully mature platform in a structurally growing market, where the long-term profit potential remains hidden behind short-term volatility.

Sea Ltd.’s most important asset is Shopee, the leading eCommerce platform in Southeast Asia and Taiwan, as well as an important platform in Brazil. Alongside it stands Monee, formerly SeaMoney, as a growing digital financial services platform. Garena is the historical origin of the company and was the cash flow engine for a long time. Today, however, it is more of an important stabilizing pillar than the core of the long-term equity story.

Before turning to Sea Ltd. itself in more detail, we will first look at the market to understand why the opportunity is so compelling.

2. Southeast Asia: A fragmented but massive opportunity

Southeast Asia (“SEA”) is the geographical southeastern subregion of Asia, consisting of the areas south of China, southeast of the Indian subcontinent, and northwest of Australia. SEA includes eleven countries. It is commonly divided into Mainland Southeast Asia, comprising Cambodia, Laos, Myanmar, Peninsular Malaysia, Thailand, and Vietnam, and Maritime Southeast Asia, comprising Brunei, East Malaysia, East Timor, Indonesia, the Philippines, and Singapore.

Southeast Asia is not one single, homogeneous market. It is a cluster of countries with very different income levels, languages, regulations, logistics structures, and consumer habits. This fragmentation makes the region operationally more difficult than China or the United States. At the same time, however, it creates a natural advantage for regional platforms that can solve local complexity better than global one-size-fits-all platforms.

The following chart shows the eleven countries of Southeast Asia by population, area, gross domestic product (GDP), and GDP per capita. The most important takeaway is simple: Southeast Asia is not a niche market. The region is home to 700 million people with compelling demographics, and several economies that are already relevant on a standalone basis. Indonesia is by far the largest market, but the Philippines, Vietnam, Thailand, and Malaysia are also large enough to support independent platform ecosystems.

For Sea, this heterogeneity is both a curse and an opportunity. It is a curse because each country has its own requirements when it comes to payments, logistics, regulation, and consumer behavior. But it is also an opportunity because this exact complexity creates a barrier to entry. Anyone who wants to build a pan-regional platform in Southeast Asia has to do more than build software. They need to develop local infrastructure, merchant networks, payment rails, logistics processes, and user trust country by country.

The second chart puts Southeast Asia into a global population context. The comparison with China and India is particularly important. Southeast Asia is smaller than both giants, but large enough to form its own digital consumer region. At the same time, the region remains far less visible to many Western investors than China or India.

Another important point is demographics. Southeast Asia is not only large; in several key countries, it is still young.

The following chart from Asia Partners shows the 15–35-year-old population by country within Southeast Asia. This age cohort matters because it represents the core consumer base for many digital platforms: people who are mobile-first, open to online shopping, comfortable with digital payments, and more likely to adopt new consumer internet habits early.

The chart also shows how different the region is internally. Indonesia remains by far the largest demographic engine and is expected to see its 15–35-year-old population peak only around 2032. The Philippines is even more striking, with its young consumer base expected to continue growing until around 2055. By contrast, Vietnam, Thailand, Malaysia, and Singapore have already passed their respective peaks.

This matters for Sea because the company is not exposed to one single demographic curve. Shopee operates across a region where some markets are already more mature, while others still have many years of young consumer growth ahead. Indonesia and the Philippines, in particular, remain highly important long-term markets because they combine large populations, rising incomes, and still-growing young consumer cohorts (and they are the largest markets for Sea).

The next chart shows the region’s economic catch-up process. Southeast Asia has grown from around $648 billion of nominal GDP in 2000 to more than $4.2 trillion in 2025. Further growth is expected through 2030. The comparison with China, the United States, and India makes clear that Southeast Asia does not reach the absolute scale of the world’s largest economies, but it is still highly relevant as a standalone growth region.

For consumer internet companies, this nominal growth matters a lot. Platforms like Shopee do not only benefit from more users. They also benefit from rising incomes, larger basket sizes, growing digital advertising markets, better payment infrastructure, and increasing merchant professionalization. This is the macroeconomic soil in which platform compounders can emerge.

Subscribe now

3. The “Golden Age” at the consumer level

Southeast Asia is not just a large and growing region. What makes it much more interesting is the specific stage of economic development it is currently going through.

Asia Partners, a private equity firm focused on Asian technology companies, describes Southeast Asia as a region currently moving through the “Golden Age” of per-capita income. What they mean is a particular development phase in which real, inflation-adjusted income per capita reaches a level where consumer behavior begins to change structurally.

This is not just about people becoming a little wealthier. It is about a broad part of the population starting to have meaningfully more discretionary income. And historically, this has often been the phase in which large local consumer internet and platform companies emerge.

The three precedents of China, Korea, and Japan show that each country went through a period in which real per-capita income entered the Golden Age zone of roughly $4,000 to $6,000. When countries move into this income band, the probability increases meaningfully that the future technology and platform winners are founded, scaled, or brought to the public markets during that window.

Why does this happen?

Because gross savings rates, used as a rough proxy for discretionary income after basic needs have been covered, tend to rise meaningfully during the Golden Age. Once people reach a certain income level, the portion of income they can allocate more flexibly begins to grow.

That is exactly what matters for companies like Sea.

Platforms like Shopee do not thrive because people merely cover their basic needs. They benefit when consumers start to:

• shop online more frequently,

• look for greater selection,

• value convenience,

• use digital payments,

• adopt delivery services,

• and increasingly manage financial services digitally.

When households have little room to spend beyond basic necessities, the addressable market for these services is naturally limited. But once discretionary income starts to rise, consumer habits can change very quickly. Convenience, selection, speed, and digital user experiences suddenly become much more important.

And now, guess what: Southeast Asia has only recently been moving through this Golden Age.

On a real per-capita income basis, Southeast Asia is currently in the Golden Age zone, lagging China by ~15 years. The region is now wealthy enough for mass digital consumption to emerge, but still early enough that many digital categories, especially eCommerce, digital payments, and consumer credit, are far from mature.

That combination is exactly what makes Southeast Asia so interesting for platform companies right now.

4. Sea Ltd.

Sea Ltd. was founded in Singapore in 2009 and has evolved into one of the most important consumer internet companies in Southeast Asia. Today, the group consists of three core businesses: Garena in Digital Entertainment, Shopee in eCommerce, and Monee in digital payments and financial services.

Historically, the story started with Garena. The gaming business was Sea’s original engine and, for a long time, the main source of cash flow. Garena gave the company the financial base, user relationships, and entrepreneurial confidence to expand beyond gaming and build larger consumer internet platforms across Southeast Asia.

But the center of gravity has clearly shifted.

Today, Sea’s equity story is increasingly about Shopee and the financial services ecosystem around it. Shopee has become the group’s most important asset: the leading eCommerce platform in Southeast Asia and Taiwan, and an increasingly relevant platform in Brazil. It is the business that defines Sea’s long-term opportunity, because it sits directly on top of the region’s rising digital consumption, growing merchant base, improving logistics infrastructure, and increasing online penetration.

Monee, formerly SeaMoney, is the natural extension of that ecosystem. As more consumers and merchants transact through Shopee, the opportunity to offer digital payments, wallet services, consumer credit, and merchant financing grows with it. In that sense, Monee is not a random financial services add-on. It is deeply connected to the eCommerce platform and could become a meaningful second earnings engine over time.

Garena remains important, but its role has changed. It is no longer the sole driver of the Sea investment case. Instead, it is best understood as a profitable and cash-generative business that helped fund the group’s expansion and still provides stability to the broader platform.

Sea is not a pure eCommerce company, not a pure gaming company, and not a pure fintech company. For many investors, that lack of focus alone is enough to make the company look too complicated, too messy, or simply too hard to categorize. Sea is a regional consumer internet platform built around three interconnected pieces: entertainment, commerce, and financial services.

5. Garena: The cash cow

Garena is Sea’s oldest segment and the origin of the entire company. The business gives users access to mobile and PC-based online games, which Garena develops, curates, licenses, and localizes for individual markets.

The business model is relatively straightforward. Most games are free to play and monetized through in-game purchases. Players do not necessarily pay for access to the game itself. Instead, they spend money on digital items, upgrades, skins, characters, or other virtual content. As a result, revenue is not only driven by the number of players, but even more by how active and engaged those players are and by what share of the user base actually spends money inside the game.

The most important title remains Free Fire. Free Fire was developed in-house and launched in 2017. For Sea, this was a decisive turning point. Before Free Fire, Garena was mainly a regional publisher and operator of licensed games. With Free Fire, Sea became the developer of a globally successful mobile game. This materially improved the quality of the business model, because self-developed games offer much better long-term control over the product, updates, monetization, and brand value than purely licensed content.

Garena has played an enormous role in Sea’s history. For many years, Shopee required substantial capital to fund growth, logistics, marketing, and international expansion. Garena was the highly profitable cash flow engine that helped finance this expansion. That also explains why Sea was previously viewed much more as a gaming company. During the pandemic, user numbers, engagement, and bookings exploded. Afterwards, the business normalized significantly.

In 2025, however, Garena staged a strong comeback. For the full year, Garena’s bookings increased by 37.3% to $2.9 billion, GAAP revenue grew by 26.1% to $2.4 billion, and Adjusted EBITDA increased by 38.1% to $1.7 billion. In the fourth quarter of 2025, Garena had 633.3 million quarterly active users and 58.0 million quarterly paying users.

The following chart shows Garena’s revenue development:

And the next chart shows Garena’s adjusted EBITDA development:

Subscribe now

7. Shopee: The core of the growth story

Shopee is the growth engine that will largely determine whether Sea can become a true platform compounder over the long run.

Shopee operates the largest eCommerce platform in Southeast Asia and Taiwan and is also one of the leading eCommerce platforms in Brazil. At its core, Shopee is a regional marketplace that connects buyers, sellers, logistics providers, payment providers, including Monee, and increasingly advertisers. In 2025, Shopee connected roughly 400 million active buyers with around 20 million sellers.

The business model is essentially a marketplace model. Shopee brings buyers and sellers together without owning most of the goods itself. The value of the platform comes from enabling transactions, building trust, simplifying payments and logistics, and giving sellers access to millions of potential customers. Sea monetizes Shopee mainly through transaction-based fees, advertising revenue, various value-added services, and logistics-related services.

The simplified arithmetic of an eCommerce platform looks like this:

Number of orders × AOV = GMV

AOV stands for average order value. GMV, or Gross Merchandise Value, refers to the value of orders for products and services on the Shopee marketplace. Sea’s calculation of GMV for its eCommerce platform includes shipping and other charges.

GMV × Take Rate = Marketplace Revenue

The platform provider keeps a percentage of GMV as its take rate. This is the core revenue source of the marketplace.

In addition, platform operators can also act as merchants themselves by buying goods from suppliers, holding them in inventory, and taking full control over processes and pricing. This is called the 1P model, or first-party model. The classic marketplace model, by contrast, is called 3P, or third-party.

Shopee has now grown to around 4 billion orders in Q4 2025:

AOV declined steadily from roughly $18 in early 2017 to around $8–10, where it has stabilized since 2022:

This results in the following GMV development:

GMV increased from $4.1 billion in 2017 to $127.4 billion in 2025. That equals a growth CAGR of roughly 54% over eight years.

Unfortunately, Sea does not disclose the GMV split by country. However, estimates suggest that the 2025 split was roughly as follows:

• ~$83 billion Southeast Asia

• ~$22 billion Taiwan

• ~$13 billion Brazil

• ~$10 billion other markets

According to the e-Conomy SEA 2025 report published by Google, Temasek and Bain & Company, the total non-grocery eCommerce GMV in Southeast Asia reached around $161 billion in 2025. Based on the estimated GMV split above, Shopee would account for roughly 52% of the market.

Its key competitors in the region are Lazada, Alibaba’s Southeast Asian eCommerce platform, TikTok Shop, and Tokopedia, the Indonesian platform acquired by ByteDance, the owner of TikTok.

One of the most important developments at Shopee has been the steady increase in its take rate. In the early years, the take rate was effectively close to zero. Today, it is around 12%. This was part of the strategy.

Shopee first wanted to get the marketplace flywheel spinning as quickly as possible. More sellers create more selection. More selection attracts more buyers. More buyers make the platform more valuable for sellers. And with every additional transaction, the platform becomes more relevant.

• On the seller side, Shopee attracted merchants with no or very low commissions and free shipping.

• On the buyer side, the company used vouchers, discounts, and marketing to get consumers to place their first orders.

This is also visible in the high marketing spend, which we will look at shortly. The goal was not short-term profitability. The goal was liquidity on the platform. Buyers and sellers had to get used to Shopee as their central marketplace.

The following video with Christiano Ronaldo (I’m wondering how much he was paid for that…) captures this consumer behavior in a much more entertaining way than any chart could. It may look funny at first glance, but it points to something important: In many markets, Shopee has become part of everyday consumer behavior. Discounts, vouchers, free shipping, entertainment, and convenience all helped train users to return to the platform again and again.

You could compare these subsidies to pouring fuel on a fire to make it burn faster. Once the fire is hot enough, it can sustain itself.

That is exactly what the take-rate chart above shows. Until 2017, Shopee was barely monetized. The platform first had to prove that it could generate enough demand, enough supply, and enough repeat purchases. Only once a marketplace becomes valuable enough for its users can it begin to raise fees without merchants and buyers immediately switching to competitors.

This is often described as a tipping point: the moment when an ecosystem starts to monetize its participants more aggressively, even though alternatives exist. If buyers and sellers still remain loyal to the platform, it signals confidence in the network effects and the stickiness of the merchant base.

That is why Shopee’s rising take rate matters so much. It shows that Shopee has not merely pushed GMV through the platform. It appears to have built real platform relevance. If Shopee had grown only because of subsidies, higher fees should have quickly led to market share losses. Instead, Shopee was able to increase monetization gradually over many years, from almost zero in the early phase to around 11–12% today.

Importantly, this take rate is not just a simple seller commission. Depending on the calculation, it also includes advertising revenue, transaction-related fees, payment and service revenue, and logistics-related value-added services. That is where the real leverage sits for high-margin revenue:

Merchants buy visibility, run ads, use fulfillment, process payments through the platform, and purchase additional services.

This is also where the profit improvement comes from. As a marketplace matures, two things tend to happen at the same time: take rates rise, while marketing spend per order falls. Most of the take-rate increase drops through to earnings because the variable costs attached to it are relatively low. That creates operating leverage.

The following chart illustrates this very well by showing both the take rate and the gross profit margin. As the take rate increased steadily, shown on the left axis, the gross profit margin, shown on the right axis, also improved. Shopee now earns a gross margin of around 30–35%, compared with only 5–10% revenue margins in a 1P model.

In the early years, Shopee was a deliberately subsidized marketplace. Revenue was still small, while losses increased as the company expanded. It is not visible in the chart bel, but in 2019, the EBIT margin was still -135%. Only in 2025 did Shopee reach positive full-year EBIT margin for the first time, at 3.5%.

This was part of the strategy. Sea invested aggressively in buyer acquisition, seller incentives, free shipping, discounts, marketing, and logistics to get the marketplace flywheel moving, as described above.

We will return to profitability from a broader perspective, as well as Shopee’s future profitability target, shortly.

8. Monee: the fintech lever on top of the Shopee ecosystem

Monee is Sea’s digital financial services segment and the third major building block alongside Shopee and Garena. It is one of the leading digital financial services providers in Southeast Asia, with a growing presence in Latin America. As a result, Monee is increasingly becoming a standalone growth and earnings driver within the group.

Strategically, Monee is particularly interesting because it builds on an advantage that many financial services companies do not have: Shopee already provides a massive user base, transaction history, merchant relationships, and payment data. While traditional banks or pure fintechs have to acquire customers at high cost, Sea can offer financial products directly where consumers and merchants are already active: inside the eCommerce ecosystem.

The business model consists mainly of digital payments, wallets, consumer credit, buy-now-pay-later products, merchant and SME financing, and digital banking offerings. The most important earnings driver today, however, is clearly the credit business.

Monee’s revenue and operating income primarily come from its consumer and SME credit businesses. At the end of 2025, Sea Ltd. reported outstanding loans of $8.0 billion, of which $7.4 billion were classified as short-term loans within current assets.

The numbers show how quickly Monee has developed. Until 2019, the segment was almost irrelevant, but since then it has gained significant importance. In 2025, revenue increased by 60.1% to $3.8 billion. In Q4 2025 alone, Monee generated revenue of $1.13 billion, implying a run-rate of more than $4.5 billion even before assuming any further growth.

Adjusted EBITDA increased by 43% in 2025 to $1.0 billion. Monee is therefore already highly profitable and no longer just a “long-term option.”

The more transactions run through Shopee, the more data is created. This data helps with credit decisions, risk assessment, fraud detection, and product selection. At the same time, buyers and merchants can use financial products directly where they are already active. This reduces friction and increases the likelihood that Monee becomes a natural part of the Shopee ecosystem.

The most interesting point is the flywheel between Shopee and Monee. Shopee brings buyers and sellers onto the platform. Monee makes payments and financing easier. Better financial products can then enable more transactions, because consumers gain additional purchasing power through installments or credit, while merchants can use working-capital solutions. More transactions strengthen Shopee. More Shopee activity gives Monee better data and more customer touchpoints. In the best case, both segments reinforce each other.

At the same time, Monee is also the segment where risk management matters most. Credit growth almost always looks attractive in good times. The key question is whether credit quality remains stable when growth is high or macro conditions become more difficult. So far, reported credit quality looks good: the 90+ days non-performing loan ratio stood at 1.1% at the end of 2025 and was stable compared with the previous quarter.

Subscribe now

8. Why the stock was so volatile

Sea’s share price volatility reflects the market’s constantly changing perception of the company. During the euphoric phase, Sea was seen as the dominant internet winner in Southeast Asia. Garena was highly profitable, Shopee was growing extremely fast, and the market was willing to accept heavy losses as necessary reinvestment to reach the tipping point, fuel the flywheel further, and secure the best market position in a large, underpenetrated, and fast-growing TAM. Once interest rates started to rise and growth-at-any-cost was no longer rewarded, the equation changed abruptly.

Over the past few years, the market has repeatedly put Sea into different narratives: first as a gaming-funded growth machine, then as a pandemic winner, then as a cash-burn risk, then as a profitability turnaround, and most recently as a profitable growth platform that is once again investing more aggressively.

Phase 1: IPO to 2019 — the still-unproven platform case

In the first years after the IPO, Sea was still a relatively small and difficult-to-categorize internet company. Garena was already profitable and generated cash flow, while Shopee was investing heavily and reporting large losses. At that point, the market still had to figure out whether Shopee was just an expensive eCommerce experiment or whether Sea could actually build a regional platform around it.

The narrative at the time was essentially this: Garena funds the option value of Shopee. The share price had already started to rise, but the major re-rating was still ahead. Shopee first had to prove that subsidies could translate into real marketplace liquidity: more buyers, more sellers, more selection, more orders, and eventually higher relevance for merchants.

Phase 2: 2020 to the end of 2021 — the perfect growth story

During the pandemic, all the pieces suddenly seemed to fit together. Free Fire was extremely strong, Garena generated large profits as a result, Shopee accelerated massively, and eCommerce in Southeast Asia received an additional push from lockdowns and changing consumer behavior. In 2021, Garena generated bookings of $4.6 billion and Adjusted EBITDA of $2.8 billion, while Shopee’s gross orders increased by 116.5% to 6.1 billion and revenue grew by 136.4% to $5.1 billion.

The narrative shifted to Garena as the cash cow, Shopee as the dominant eCommerce winner, and SeaMoney as an additional fintech option. In a zero-interest rate world, the market was willing to discount this growth far into the future. The stock rose to almost $370, which at the time implied a market capitalization of roughly $200 billion. Within less than two years, the stock had returned more than 800%.

In hindsight, this was also the most dangerous phase for investors, as the market priced in almost everything at once: permanently strong gaming, uninterrupted eCommerce growth, cheap capital, and a very high probability that Shopee would become highly profitable later on.

Phase 3: 2022 — the collapse and the forced path to self-funding

In 2022, the narrative turned brutally. Higher interest rates made unprofitable growth stocks less attractive, Garena normalized after the pandemic boom, Free Fire was banned in India, and Shopee continued to report large losses. Sea still reported $12.4 billion of revenue for 2022, but also a net loss of $1.7 billion. Shopee’s Adjusted EBITDA was negative $1.7 billion, after negative $2.6 billion in the prior year.

The market was no longer asking: how big can Sea become?

The new question was: can Sea become profitable at all if Garena can no longer fund everything?

Within roughly one year, the stock collapsed by almost 90% from its peak (and even after the recovery from the lows, Sea still trades around 75% below its all-time high today).

This pressure led to what was probably the most important management pivot in the company’s history. Forrest Li, founder, major shareholder, and CEO, made it clear that Sea could no longer rely on capital always being available. Management temporarily gave up cash compensation until the company reached “self-sufficiency.”

At the same time, costs were cut, expansions were scaled back, and the focus on profitability increased dramatically.

This strategic shift was particularly visible in Shopee’s international expansion. During the pandemic, Sea had tried to roll out Shopee very aggressively beyond its core markets of Southeast Asia, Taiwan, and Brazil. Poland, France, Spain, India, Argentina, Mexico, Chile, and Colombia were all part of this phase. The logic was clear: if Shopee’s playbook of low merchant fees, free shipping, promotions, gamification, and aggressive marketing had worked in Southeast Asia, perhaps it could also work in other emerging markets and even in Europe.

But in 2022, this experiment was largely brought to an end.

• In France, Shopee had only launched in October 2021 and already shut down the platform again in March 2022. Shopee itself described France as an initial test market.

• India followed almost immediately afterwards. Shopee had launched there only a few months earlier and exited the retail market again on March 29, 2022. Reuters pointed at the weaker growth outlook and the fact that the withdrawal came shortly after the Free Fire ban in India.

• Spain was also shut down quickly, with the Spanish platform set to stop operations in June 2022.

• In September 2022, the next cut came in Latin America. Shopee fully exited Argentina and stopped local operations in Chile, Colombia, and Mexico. In those three markets, only a cross-border model remained, meaning products from sellers in other countries could still be sold there, but not from local sellers.

• At the beginning of 2023, Poland was also closed. Shopee had launched there in September 2021 as its first European market and ended operations in January 2023.

In effect, Shopee gave up the idea of rolling out the Southeast Asian playbook globally in the short term. The focus shifted back to the markets where Shopee already had strong relevance or a realistic path toward a strong market position: Southeast Asia, Taiwan, and Brazil.

This was the moment when the market forced Sea to make a new promise: we can do more than grow and burn cash: we can fund ourselves.

Phase 4: Early 2023 — Sea proves profitability

In 2023, with the release of Q4 2022 results, the first countermove arrived. Sea showed that the profitability lever was real. Costs had already been reduced, marketing had become more disciplined, unprofitable activities had been scaled back, and Shopee’s unit economics improved significantly.

Only one year earlier, in Q4 2021, Shopee was still fully in growth-at-all-costs mode. Revenue was around $1.6 billion, but the segment lost almost $941 million on an EBIT basis. The EBIT margin was -59%, and the EBITDA margin was -55%. Back then, Shopee was buying growth with high marketing spend, free shipping, subsidies, and aggressive expansion.

One year later, the picture looked completely different. In Q4 2022, revenue had already reached $2.1 billion. Cost of services fell from 75% of revenue to around 58%, while sales and marketing expenses declined from 53% of revenue to only 18%. At the same time, the gross margin jumped from 6% to 32%.

The result was a fundamental break in how the market perceived the business model. A segment that had generated almost $1 billion of EBIT losses in the prior-year quarter became a segment with positive EBIT of $109 million and positive EBITDA of $196 million in Q4 2022.

This was the first concrete proof that Shopee can be profitable.

Forrest Li therefore said:

The narrative shifted again. Sea was no longer just a cash-burning growth stock, but a company that could become profitable.

For the share price, that was initially positive. But the next question immediately appeared: if Sea is now profitable, is the company sacrificing too much growth?

Phase 5: Second half of 2023 — reinvestment and renewed skepticism

After Sea had proven that profitability was possible, management pivoted again. The logic was straightforward: now that the cost base was healthier and the company had become self-funding, Sea could start investing in growth again.

This shift was difficult for the market to digest. In Q2 2023, Sea announced that it would increase investments in the eCommerce business again. Forrest Li said that these investments could cause Shopee and the group to record losses again in certain periods, while also emphasizing that self-sufficiency and cost efficiency would remain central.

This was the second major management pivot:

• 2022 / early 2023: prove profitability.

• From mid-2023 onward: accelerate growth again.

The market reacted nervously because this looked, at least on the surface, like a return to old losses. On top of that came more intense competition, especially from TikTok Shop. As a result, the stock remained volatile despite operational improvements.

This created an almost paradoxical market reaction. When Shopee was still reporting large losses, investors demanded proof that the business model could become profitable. Management delivered that proof: costs were cut, subsidies were reduced, marginal markets were closed, and Shopee turned profitable surprisingly quickly.

But as soon as that proof was delivered, the concern shifted. The market was asking now whether that profitability had been achieved at the expense of growth. Sea had solved the first problem, and by doing so, it triggered the next narrative.

When management then started investing more heavily in growth again, that too was interpreted negatively. Instead of seeing reinvestment as a rational attempt to reaccelerate a stronger platform, the market feared a return to the old cash-burn patterns.

Phase 6: 2024 to the end of 2025 — profitable growth returns

In 2024 and 2025, the picture improved materially again. Sea showed that growth and profitability could potentially coexist.

• For 2024, Sea reported revenue of $16.8 billion, net income of $447.8 million, and Adjusted EBITDA of $2.0 billion.

• In 2025, growth accelerated further. Sea generated revenue of $22.9 billion, up 36.4%, net income of $1.6 billion, and Adjusted EBITDA of $3.4 billion. Shopee reached GMV of $127.4 billion, 13.9 billion orders, revenue of $16.6 billion, and Adjusted EBITDA of $880.6 million.

This created a new narrative. Sea was no longer the old hypergrowth cash-burn case with triple-digit or very high double-digit growth rates. It had become a platform with three functioning earnings engines. Shopee was growing strongly and becoming profitable. Monee was scaling as a fintech segment. And Garena was stabilizing, or even growing again.

The stock recovered significantly as a result, but remained far below its previous highs.

Phase 7: Late 2025 to today — the market wants proof of disciplined reinvestment

The latest share price movement shows that Sea still suffers from a trust problem. Whenever costs rise again or management invests more heavily, the market immediately remembers 2021 and 2022 and punishes the stock.

After the Q4 2025 results, the stock fell sharply because investors reacted negatively to higher operating expenses, rising marketing costs at Shopee, increased investment in Monee, and a slightly more cautious GMV growth outlook.

The 2026 outlook implies higher investment at Shopee, which once again scared investors:

Summary

The following chart summarizes the Adjusted EBITDA contributions across all segments. It captures the Sea story very well in a single image: Garena funded the group for a long time, Shopee burned massive amounts of capital, Monee was initially small, and since 2023/2024, all three segments have increasingly moved toward positive earnings contributions.

In the first phase through 2019, Sea was still a relatively small company. Garena was already generating positive Adjusted EBITDA, while Shopee deliberately reported losses to build the eCommerce flywheel. From 2020 onward, Garena then became an enormous cash flow engine. During the pandemic, the gaming business exploded, and those profits funded the aggressive investments in Shopee and later also Monee.

The critical point came in 2021/2022. Garena remained profitable, but normalized after the COVID boom, while Shopee’s losses became very large at the peak of its expansion. This was exactly when the central market question emerged: can Sea survive and become profitable if Garena can no longer cross-subsidize everything?

From Q4 2022, and especially in 2023, the management reset became visible. Shopee’s losses shrank dramatically and even turned positive for a while. At the same time, Monee increasingly became a real earnings contributor rather than just a fintech option. This changed the quality of the group: Sea was no longer dependent solely on Garena, but was developing multiple earnings sources.

Finally, in 2024 and 2025, the chart shows the new state of the business. Garena remains profitable, Shopee is much closer to sustainable profitability, or already positive depending on the period, and Monee is growing as an additional EBITDA lever.

9. Long-term potential and thoughts on valuation

The big question is simple: how much potential does Sea Ltd. still have, and what is the market currently pricing in?

Sea’s current market capitalization is around $51.9 billion. At the end of 2025, the company had $1.8 billion of financial debt, but also $4.2 billion of cash and $8.3 billion of investments. In other words, Sea had a net cash and investment position of roughly $10.7 billion, which brings the enterprise value to approximately $41.2 billion.

For 2026, I expect EBITDA of around $2.8 billion on a group level (taking into account HQ costs), implying an EV/EBITDA multiple of roughly 14.7x.

Let’s look at Sea through a simple sum-of-the-parts lens to see what the current valuation implies for Shopee.

• Garena: I value Garena at 5x Adjusted EBITDA. Here, I deliberately use Adjusted EBITDA rather than EBITDA, because Adjusted EBITDA better reflects the underlying cash generation of the gaming business. GAAP revenue is affected by revenue recognition, while Adjusted EBITDA is closer to the cash economics of bookings. Based on 2025 Adjusted EBITDA of $1.6 billion, this would imply a value of around $8 billion.

• Monee: I value Monee conservatively at the size of its outstanding loan book, or roughly $8 billion. This is approximately 8x 2025 EBITDA. Put simply, Monee could stop writing new loans today and would recover a large part of this amount in cash over the next twelve months.

• Shopee: The remaining value implied for Shopee would therefore be around $25 billion.

Shopee’s GMV is expected to grow by roughly 25% in 2026, while revenue should grow even faster as the take rate continues to increase.

The key question is therefore: what is Shopee worth once it reaches mature profitability?

To put this into context, let’s look at one of the most competitive eCommerce markets in the world: China. Four large players dominate the market:1

• Tmall is Alibaba’s flagship business-to-consumer platform, serving more than one billion buyers and over one million merchants. Launched in 2008 as Taobao Mall, it has grown into China’s largest B2C marketplace. In 2024, the number of new merchants joining Tmall increased by 83% year over year. Supported by powerful marketing tools, data analytics, and deep integration with Alibaba’s wider ecosystem, including Alipay and Cainiao logistics, Tmall’s estimated GMV reached almost $1.11 trillion in 2024.

• JD.com, founded in 1998 as a computer equipment vendor, has grown into one of China’s largest eCommerce platforms, with estimated GMV of $565 billion in 2024. JD is known for its supply-chain capabilities and self-operated model, which gives the company control over inventory, warehousing, and logistics. Its strategic partnership with Tencent, the Chinese technology company behind WeChat, has further expanded JD’s market reach across China.

• Douyin, developed by ByteDance, the company behind TikTok, has become a dominant force in China’s short-video and livestreaming ecosystem. Since introducing shopping features in 2018, Douyin has used its entertainment-driven platform to reach younger, mobile-first consumers. Its algorithm personalizes the shopping experience and blends content with commerce. In 2024, Douyin’s estimated GMV reached $477 billion.

• Pinduoduo (PDD), the owner of Temu, has become one of China’s fastest-growing eCommerce platforms since its launch in 2015. It built its success around group buying, social commerce, and a consumer-to-manufacturer model that allows consumers to influence production through real-time demand. Today, Pinduoduo is also China’s largest direct-to-consumer agriculture platform. In 2024, estimated GMV reached $723 billion.

Alibaba’s Taobao/Tmall ecosystem and PDD are roughly in the range of 2–2.5% EBITDA/GMV, depending on the exact earnings metric and GMV estimate used. JD is lower, but less comparable because of its heavier 1P and logistics model.

Even though Southeast Asia (and Brazil) are highly competitive markets, the underlying market is still growing. Southeast Asian eCommerce is expected to roughly double by 2030.

And as China shows, large eCommerce markets do not necessarily end with only one profitable winner. More than one scaled player can coexist profitably. This is especially relevant because MercadoLibre and TikTok Shop address different users and customer behaviors, which means they are not always competing head-to-head with Shopee in the same way.

Now assume that Shopee doubles its GMV again by 2030 versus 2025, reaching around $250 billion. Given the expected 25% GMV growth in 2026 alone, this does not seem like an overly aggressive assumption.

If Shopee still earned only 0.7% EBITDA/GMV, the 2025 level, EBITDA would reach roughly $1.8 billion by 2030. But if Shopee moved into management’s long-term target range of 2–3% EBITDA/GMV, EBITDA would land somewhere between $5 billion and $7.5 billion.

10. Why Sea might still be misunderstood today

Sea is harder to categorize than a classic quality company. It is earlier in its corporate life cycle, which means the range of possible outcomes is much wider.

The thesis ultimately depends on whether Shopee can defend its market leadership while continuing to improve monetization. If it can, Sea could offer a very rare combination: high structural growth, a leading platform position, multiple monetization levers, and a stock price that is still weighed down by the memory of the difficult years.

That does not automatically make Sea a safe investment. Quite the opposite. The risks are higher than for many classic compounders. Competition, customer behaviour, regulation, credit risk, currency movements, capital allocation, and technological change are all real. But the potential reward is also higher. If Sea truly becomes the leading platform compounder in Southeast Asia, today’s phase may one day look like the early re-rating of a more mature company.

And who knows, by 2030 or 2035, there may even be a fourth major segment. Big Tech companies have repeatedly reinvented themselves by making relatively small investments into new business models with potentially huge TAMs. Sea may or may not do the same, but the possibility of future optionality is part of what makes the company interesting.

11. Risks that should not be ignored

The biggest risk is competition. Shopee is the market leader, but the market is not static. TikTok Shop, Lazada, Temu, local players, and social commerce models could keep profitability under pressure for longer than expected. The most dangerous outcome would be a market where GMV continues to grow, but every platform is forced to buy a large part of that growth through subsidies.

The second risk is Monee. Credit businesses often scale impressively in good times, but the true quality only becomes visible over a full credit cycle. Strong growth rates only create value if underwriting, pricing, and risk management remain robust. NPLs, provisions, and the composition of the loan book should therefore be watched closely.

The third risk is capital allocation. Sea has shown that it can make large bets. In Shopee’s case, that was the right decision. But it could become a problem in the future if new markets, new product categories, or fintech expansion are pursued too aggressively. For Sea, capital discipline is therefore a central part of the thesis.

Fourth, Garena remains a cyclical and hit-driven profit pool. Free Fire has been an extraordinary success, but no game is immortal. Garena either has to preserve the value of its existing franchise or develop new content, without allowing the segment to lose too much relevance for group profitability.

Conclusion

Sea is not the typical “boring” Slow Compounding company I usually look for.

It is more complex, more volatile, and clearly riskier than many of the businesses I normally like to study. Competition is intense, the company operates across very different markets, Monee adds credit risk, and management’s capital allocation discipline will remain important. This is not a simple buy-and-forget compounder.

But that is also why the opportunity exists.

Sea still seems to be viewed through the scars of the past: the pandemic boom, the collapse in Garena, Shopee’s heavy losses, aggressive expansion, abrupt cost cuts, and then renewed reinvestment. Yet underneath that messy history, the company has changed meaningfully. Shopee is now much larger, more relevant, and increasingly monetized. Monee has become a real earnings contributor. Garena is no longer the whole story, but still provides a profitable pillar.

The long-term question is whether Shopee can defend its leading position while moving closer to mature profitability. If it can, the numbers become very interesting. Thinking out to 2030 or even 2035, Sea could still have many years of growth ahead, driven by rising eCommerce penetration, higher take rates, advertising, logistics density, digital payments, and consumer finance.

The valuation does not look demanding for that kind of upside. Based on a simple sum-of-the-parts view, the market does not seem to assign an ambitious value to Shopee. That does not eliminate the risks, but it makes the setup attractive: the downside appears more manageable than the stock’s volatility suggests, while the upside could be substantial if Sea continues to execute.

Never forget, that over the long-term, the stock price will reflect the underlying profitability (and its embedded expectations for future earnings growth).

Never forget: over the long term, the stock price will follow the underlying earnings power of the business — and the market’s expectations for how that earnings power can grow.

I still need to do more research to get deeper into the details, especially around competition, Shopee’s country-level economics, Monee’s loan book, and the sustainability of long-term margins. But the first impression is clear: Sea is too interesting to ignore.

Does Sea Ltd. deserve a full Deep Dive?

This article was only a Company Snapshot.

For many Substackers, this might already count as a Deep Dive. But for Slow Compounding, a real Deep Dive would mean going several layers deeper: country-level competition, Shopee’s unit economics, Monee’s loan book, credit quality, management’s capital allocation, and the long-term earnings power of the whole ecosystem.

Sea is not the typical “boring” compounder I usually write about. It is more complex, more competitive.

But that is also what makes it interesting, as the upside seems quite compelling.

A potential structure could look like this:

• Part 1: Southeast Asia and Brazil - the markets behind the opportunity

• Part 2: Garena

• Part 3: Shopee

• Part 4: Monee

• Part 5: Management, capital allocation

• Part 6: The full investment story & valuation

So I would like to hear from you:

That, in one sentence, is the Röko case.

Röko was founded by the former Lifco CEO, one of the key architects behind a business that delivered extraordinary shareholder returns over a very long period. And those returns stood out not only within the Nordic serial acquirer universe, but even relative to Berkshire Hathaway.

That is what makes Röko so interesting. This is not just another Swedish serial acquirer with a decentralized structure and a long list of acquisitions. It is the second act of someone who already helped build one of the best compounding machines in Europe.

The market, of course, is not willing to simply assume that this will work again. Röko is still young, its listed history is short, and the stock has already lost much of the initial excitement that often surrounds a newly public serial acquirer. As so often happens, once the first wave of enthusiasm fades, attention fades with it.

But that is usually where things get interesting.

Because Röko is not trying to copy Lifco in a superficial way. In many ways, it looks like an attempt to build on the most important lesson Lifco taught Fredrik Karlsson: the real strength of the model was never being tied to one specific vertical. It was having the freedom to allocate capital across a much broader opportunity set. That is exactly why Röko appears to have been built as a pure sector-agnostic serial acquirer for from day one.

My view is that this is what makes the setup so compelling. Röko combines a proven playbook, a highly experienced founder, a decentralized model, aligned incentives, and a long runway for acquisitions. The ingredients of a very strong long-term compounder are clearly there.

That is why Röko’s story is so compelling.

In our Quality Growth framework, Röko checks several boxes:

• 🧠 A founder with a rare track record: Fredrik Karlsson already helped turn Lifco into one of Europe’s great compounders

• 🌍 A broader and more sector-agnostic acquisition model: built around the idea that the best opportunities are not confined to one niche

• 🏢 A decentralized structure with permanent-owner DNA: local managers keep responsibility, while HQ stays lean

• 🤝 Aligned incentives after acquisitions: sellers often remain invested, which matters in a people-driven model

• 💰 A capital allocation story first and foremost: the real value driver is disciplined reinvestment, not financial engineering

• 📈 A long runway: Röko is still early in its journey, with plenty of room to scale if quality remains high

• ⚖️ A more interesting setup after the derating: the stock no longer trades as if success were already guaranteed

This Company Snapshot is the quick guided tour. It explains what Röko is, why Fredrik Karlsson’s background matters so much, how the compounding machine is supposed to work, and why the company may be more interesting than its still-short public track record suggests.

And if that resonates, the full three-part Deep Dive (74 pages) picks up from here — moving from the story to the evidence: Lifco’s influence, the acquisition playbook, the economics of the model, the financial profile, the risks, and ultimately the long-term risk/reward.

👔 Company Name: Röko AB (“Röko”)

🔎 ISIN: SE0023950795

🔧 Business model: sector-agnostic serial acquirer

🌍 Geographic exposure: mainly Sweden, UK (sales in Europe)

📈 Stock Price: SEK 1,534

💰 Market Capitalization: SEK 22.4 billion (~$2.4 billion)

👨‍💼 Number of CEOs since foundation: 1

👨‍👩‍👦 Founder-/Owner-operator: Yes

📅 CEO tenure: Since 2019

🥇 Insider ownership: ~21.0% of shares

📊 10Y EPS CAGR: N/A

🔁 Reinvestment profile: High

💸 Capital intensity: Capital-light

🏰 Moat: Seller proposition, Simplicity, experienced management team around Fredrik Karlsson

🧨 Main risks: Execution, macro pressure, key-person risk

🌳 Slow Compounding fit: Yes – high-quality serial acquirer

Business Model in a Nutshell: Röko is a Sweden-based serial acquirer that has completed 33 platform acquisitions and in a variety of industries across Europe. The Company’s ambition is to be the perpetual owner of small and medium-sized businesses. Röko’s organization has more than 100 years of combined experience working with founder- and entrepreneur-led companies in a wide range of industries and the individuals in executive management positions have worked at the Company since its first year of operation (2019).

If you are new to the world of serial acquirers, please check out my fundamental piece before you continue with Röko:

Subscribe now

Refining the Lifco Playbook into a World-Class Compounder

Fredrik Karlsson already built one of Europe’s great compounders once. The natural question now is whether he can do it again.

That, in one sentence, is the Röko case.

Röko was founded by the former Lifco CEO – one of the key architects behind a business that delivered extraordinary shareholder returns over a very long period. And those returns were remarkable not only within the Nordic serial acquirer universe, but even relative to Berkshire Hathaway.

Most company descriptions start with the usual labels: Swedish serial acquirer, decentralized model, niche businesses, long-term owner. All of that is true, but none of it gets to the heart of why Röko is interesting. Investors should not be looking at Röko because it is just another Nordic acquisition vehicle. They should be looking at Röko because it is the second act of the man who helped turn Lifco into one of the most successful serial acquirers in Europe.

And that matters.

Because if Lifco’s exceptional compounding was not just unique to Lifco, but the result of a repeatable system of capital allocation, decentralization, and a disciplined acquisition engine, then Röko could become one of the most interesting younger compounders in Europe today.

A good place to start is with Lifco. Over the last two decades, Lifco has not merely been a successful acquirer. It has been the standout performer among the major Sweden-based serial acquirers. From 2001 to 2025, Lifco compounded EBITA at 22.5% annually, ahead of Addtech (including AddLife) at 16.7%, Lagercrantz at 13.9%, and Indutrade at 13.1%. That is an extraordinary gap over such a long period.

This is important because the Nordic serial acquirer space is already full of impressive businesses. Beating that group is not the result of financial engineering or a good year or two. It usually reflects something deeper: a better system for sourcing, judging, buying, and holding companies over time.

Lifco built exactly that kind of system. The model was simple, but powerful: acquire profitable niche businesses, keep the group highly decentralized, maintain a very lean headquarters, demand strong cash generation, and avoid the value destruction that often comes from over-integration. The subsidiaries kept their entrepreneurial culture, while the parent company provided capital allocation discipline and a permanent home. In other words, Lifco behaved less like a conventional industrial group and more like a compounding machine.

That machine did not emerge fully formed from day one.

Fredrik Karlsson did not inherit the Lifco investors know today. When he took over in 1998, Lifco was still a much narrower business and first needed restructuring rather than expansion. Karlsson initially focused on improving profitability, simplifying the group, and pushing responsibility down into the organization. After Carl Bennet took Lifco private in 2000, Karlsson sold large parts of the Health & Self-Care operations, refocused the company on dental distribution, and lifted margins from low single digits to high single digits through a decentralized model built on simplicity, local accountability, and a relentless focus on profit and cash flow. But once that turnaround work was done, he recognized that dental alone offered limited growth. The real inflection point came in 2006: with the merger of Lifco and Sorb Industri and a more deliberate push into acquisitions, Karlsson evolved from turnaround manager into capital allocator. That shift laid the foundation for modern Lifco: a diversified, decentralized serial acquirer that compounded EBITA at an exceptional rate for years.

By the time Fredrik Karlsson stepped down as Lifco CEO due to a compensation dispute with majority owner Carl Bennet, Fredrik Karlsson had delivered striking results. Over his full tenure beginning in 1998,

• revenue rose from roughly SEK 600 million to roughly SEK 12 billion in 2018, a CAGR of 16%,

• EBITA increased from roughly SEK 30 million to roughly SEK 2.2 billion, a CAGR of 24%,

The above chart helps illustrate that. Lifco’s current shape is very different from its earlier form. In the beginning, Dental was the dominant segment. Over time, however, the real engine of growth shifted toward a broader and more diversified structure, especially through Demolition & Tools and System Solutions. What makes this so interesting is not just the growth itself, but what it says about Karlsson’s underlying lesson from Lifco: the real source of value creation was never one narrow industry vertical. It was the model.

That distinction matters enormously for Röko.

Fredrik Karlsson’s time at Lifco seems to have taught him one especially important lesson: the biggest limitation in an acquisition model is often self-imposed. Lifco started out as a much narrower dental-focused business, but the strongest profit growth did not come from staying within that original vertical. It came from expanding beyond it. Between 2006 and 2025, Dental grew EBITA by roughly 13.0x, while Systems Solutions – Lifco’s more sector-agnostic segment – increased profits by about 32.9x, compounding at roughly 20.2% annually.

The core of Röko is not simply to replicate Lifco, but to build from day one around the lesson Lifco taught Karlsson: that the real power of the model lies not in being tied to one industry, but in having the freedom to allocate capital wherever the combination of quality, valuation, and management is most attractive. In that sense, Röko is the more deliberate, broader, and more sector-agnostic expression of the same underlying playbook.

And that is where the story shifts from Lifco to Röko.

After Fredrik Karlsson left Lifco in early 2019, he founded Röko 2019 together with Tomas Billing. So this is not an outsider trying to copy the Nordic serial acquirer playbook after admiring it from afar. This is one of the people who helped build the playbook trying to start again from scratch — arguably with a cleaner structure, a wider mandate, and the benefit of hindsight.

The ambition is straightforward: Röko wants to be the permanent home for small and medium-sized niche businesses with strong market positions, healthy margins, attractive cash generation, and capable management teams that want to remain in place. That last point is central. Röko is not offering an exit into a bureaucratic corporate machine. It is offering entrepreneurs continuity, autonomy, and long-term ownership. In that sense, it is much closer to the best serial acquirers than to private equity.

Since inception, Röko has already built a meaningful portfolio across a range of niche businesses. The company is sector-agnostic, but that does not mean it is indiscriminate. Röko is not buying “anything.” It is looking for specific economic characteristics: recurring or resilient demand, strong margins, low capital intensity, solid market positions, and management teams worth backing. The group’s financial targets also reflect that discipline, with a focus on annual earnings growth, EBITA margins above 15%, and leverage that remains manageable.

This is where the investment case gets interesting.

The easiest way to misread Röko is to see it as “Lifco 2.0” and stop there. The more accurate view is that Röko seems designed around the lessons learned from Lifco. It appears broader in scope, more explicitly sector-agnostic, and highly intentional in its organizational philosophy. Röko does not want to centrally optimize its subsidiaries into sameness. It wants to preserve what already works at the local level and simply provide the capital allocation framework, governance, and financial discipline needed to compound over time.

In other words, Röko is not trying to create value through integration. It is trying to create value through selection.

That is the key. In serial acquiring, the real edge rarely comes from heroic post-deal transformation. It usually comes from buying the right businesses, at the right price, with the right people, and then not ruining them afterward. Röko’s model appears built around exactly that principle.

There are a few features that make the setup especially appealing. First, Röko often keeps sellers and managers invested through minority stakes and staged ownership structures. That helps align incentives and reduces the risk of paying everything upfront before long-term performance is proven. Second, the company remains operationally decentralized, which should help preserve entrepreneurial energy inside the subsidiaries. Third, the headquarter is intentionally lean, which matters because bloated headquarters are often the silent killer of decentralized acquisition models.

Just as importantly, Röko does not seem to have pursued growth at any price so far. One of the biggest risks in young serial acquirers is that management becomes overly eager to prove momentum and starts compromising on acquisition quality. That can damage the model before it really scales. So far, however, Röko appears to have added businesses with solid profitability rather than diluting the overall portfolio. If that continues, it would be one of the strongest early indicators that the company is serious about protecting quality.

Of course, none of this means Röko is risk-free.

The company is still young. Its listed history is short. Its public track record is nowhere near as established as Lifco’s, and the model remains heavily tied to key individuals, especially Fredrik Karlsson, Tomas Billing, and Johan Bladh. A decentralized structure also requires an unusual degree of judgment: because the center stays lean, the initial filtering and underwriting of acquisitions matter even more. If the judgment slips, the model can look fine on the surface for quite a while before problems show up. On top of that, acquisition cadence is never perfectly smooth, and embedded obligations tied to minority buyouts are worth watching carefully over time.

Those are real risks. But they are also the kinds of risks one often has to accept when looking at an earlier-stage compounder rather than an already fully proven one.

That brings us to valuation.

The market’s initial enthusiasm has already cooled. Röko no longer trades like a newly listed premium compounder with endless optimism priced in. Instead, the stock now sits at the low end of its peer group on forward earnings. Röko trades around 22.4x forward earnings (2026), compared with roughly 25.9x for Indutrade, 28.8x for Lifco, 31.5x for Lagercrantz, and 35.2x for Addtech.

Röko is younger, less proven, and still needs to demonstrate that its acquisition engine can perform consistently through time and across market conditions. But that is also exactly what makes the current setup more compelling than it would have been at a much richer multiple.

At around the low 20s on forward earnings, the investment case does not require an aggressive rerating to work. If Röko can execute well and continue compounding earnings at a high rate (~15% p.a.), most of the return can come from the business itself rather than from multiple expansion. A rerating toward more mature Nordic peers would be helpful, but it should not be the core of the thesis. The real thesis is simpler: if Röko can reinvest a large share of its cash flows at high incremental returns for a long time, shareholder value should follow.

And that takes us back to the original question.

Was Lifco’s success mainly about Lifco itself? Or was it, to a meaningful degree, about Fredrik Karlsson’s system?

Röko is the live experiment. It is the attempt to prove that the key ingredients behind one of Europe’s great serial acquirers can be rebuilt in a new structure, with a broader opportunity set and the benefit of everything learned the first time around. That does not guarantee success. But it does make Röko far more interesting than the average newly listed acquisition platform.

My view is that Röko already deserves a place on the watchlist of anyone serious about high-quality serial acquirers. It is not yet as proven as Lifco, Indutrade, Addtech, or Lagercrantz. But that is precisely the appeal. Investors are not looking at a finished masterpiece. They are looking at an earlier-stage compounding story that is still being built in public – by someone who has already shown he knows what a great one looks like.

Stay tuned for Part 1 of the upcoming deep-dive!

The last two years weren’t pretty for Teqnion. Execution got harder, the numbers got choppy, and it became easy to write the story off. With cracked fundamentals and a declining stock price, also attention from investors faded.

But a lot has been shifting under the surface: a new structure, new governance and a much higher bar in terms of acquisitions.

My bet is that once Teqnion shows sustainable revenue and EPS growth again, investor interest will return and the stock price will likely reflect that shift as well.

Teqnion’s current phase resembles Lagercrantz roughly fifteen years ago. Lagercrantz, one of Teqnion’s role models, has been through something similar. After a couple of tough cycles, progress looked slow for years and profitability stayed muted. The turnaround didn’t come as a sudden leap—it came gradually, as the business mix improved and the organization got sharper. Results changed cohort by cohort, almost quietly, until the margin profile looked like a different company. Over the years, EBITA margin increased from 5% to 17%.

Lagercrantz’ stock is up 33x since then. Teqnion is in a similar position today.

That’s why Teqnion’s “story” is so compelling.

In our Quality Growth framework, Teqnion checks several boxes:

• 🧩 A decentralized portfolio of niche operators with sustainable cash flows

• 🔧 A reset of the M&A playbook toward higher-quality, higher-margin targets

• 💰 Capital allocation discipline with a clear and reshaped focus on cash returns

• 🧑‍💼 Incentives that matter—alignment at both HQ and subsidiary level

• 📈 A long runway if the group can keep upgrading quality while scaling responsibly

• 🧱 Balance sheet pragmatism—built to keep doing deals through cycles, not just in good years

This Company Snapshot is your quick guided tour: what Teqnion actually is, how the compounding machine is supposed to work, where it broke under scale, and why the recent steps look promising (while the stock has not reflected any of the improvements in my opinion).

If that resonates, the upcoming three-part Deep Dive (77 pages) picks up from here—moving from the story to the evidence: the upgraded playbook, cohort math, margin mix, deal economics, and ultimately the full risk/reward and valuation.

Subscribe now

Download

👔 Company Name: Teqnion AB (“Teqnion”)

🔎 ISIN: SE0012308088

🔧 Business model: niche industrial serial acquirer

🌍 Geographic exposure: Sweden, UK (sales in Europe)

📈 Stock Price: SEK 169

💰 Market Capitalization: SEK 2.9 billion (~$321 million)

👨‍💼 Number of CEOs since foundation: 2

👨‍👩‍👦 Founder-/Owner-operator: Yes

📅 CEO tenure: Since 2009

🥇 Insider ownership: 5.0% of shares

📊 10Y EPS CAGR: ~23%

🔁 Reinvestment profile: High

💸 Capital intensity: Capital-light

🏰 Moat: M&A edge due to size, type of acquisition targets and deal-making approach

🧨 Main risks: Execution, quality drift, M&A engine, key-person risk

🌳 Slow Compounding fit: Yes – high-quality serial acquirer

Business Model in a Nutshell: Teqnion is a small Swedish sector-agnostic serial acquirer that describes itself as a builder of a conglomerate of industrial companies. The company is run as a decentralized group of subsidiaries, with a strong emphasis on culture, cash generation, and M&A.

If you are new to the world of serial acquirers, please check out my fundamental piece before you continue with Teqnion:

The “new Teqnion”: Gaining Momentum

As a small Swedish serial acquirer, Teqnion’s fundamental improvements are outpacing the rate at which the market is re-rating its valuation.

The management team has spent the last few months reshaping its internal processes and refining its acquisition strategy. This marks the beginning of a new multi-year era characterized by enhanced portfolio quality and greater resilience. As these improvements translate into expanding EBITA margins, the company is well-positioned for a significant re-rating of its valuation multiple, however, the key driver of shareholder returns is earnings growth.

Teqnion buys niche industrial businesses in Sweden and the UK, across a wide range of technical niches. The model has already proven it can compound at a high rate. Since inception, revenue has grown at ~26% p.a. and EBT at ~35% p.a. over roughly 15 years.

Teqnion appears to be going through a phase that resembles Lagercrantz (the smallest of the successful “Big 4” serial acquirers from Sweden) roughly fifteen years ago. After the dotcom bust and the GFC, Lagercrantz experienced a long period where profit growth was sluggish despite acquisitions, and its margin profile was modest (EBITA around ~5%). Over time, Lagercrantz steadily acquired higher-margin businesses, increased the share of more proprietary products with structurally higher gross margins, and improved its “operating system”. Lagercrantz didn’t jump from ~5% EBITA to >17% overnight; it got there slowly, cohort by cohort, as higher-quality businesses accumulated.

History rhymes: over the past two years, Teqnion went through a rough period, tightened its “operating system”, i.e. its governance and organizational system, and sharpened its capital allocation toward higher-quality companies without giving up on valuation discipline. Management aims to earn back its investment within five years on a free cash flow basis. Teqnion is now at a similar revenue scale to where Lagercrantz was about 15 years ago, has clearly laid out the shift toward higher-quality acquisitions, and is deliberately building the governance infrastructure required to scale a decentralized model.

Over the past two years, the company has been pushed into a “reset” of both its acquisition playbook and its internal operating system: raising the bar on deal quality, tightening governance, and rebuilding the portfolio follow-up discipline that a decentralized model needs once it reaches meaningful scale. The leading indicators are pointing in the right direction: higher-quality acquisitions, improving portfolio control and early signs that the struggling bucket is shrinking, underpinned by a jump in gross margin and EBITA margin and improved cash flow during 2025. I expect this positive development to turn into a trend of sustainable shareholder value creation over many years. The market’s current valuation does not reflect these pivotal changes at all (see below).

There are near-term catalysts that could move the stock, mainly the as profitability steps up over the next few quarters. The real upside, though, is multi-year: Teqnion may be entering a new era: acquiring more profitable businesses that are structurally more resilient and likely to grow faster organically than the legacy portfolio while maintaining valuation discipline and letting the consolidated margin gradually move higher. This kind of change rarely shows up from one quarter to the next. It tends to play out step by step, cohort by cohort, year after year (as will be demonstrated mathematically). But with serial acquirers, those small improvements in acquisition quality and portfolio governance can compound into a very different earnings profile over time, especially when the stock is still priced with a healthy dose of skepticism.

The first hard evidence is the 2025 acquisition cohort. Teqnion completed nine acquisitions in 2025, and the profitability of what it bought is striking: roughly ~24% EBITA margin versus a historical group margin closer to ~10%. It reflects a deliberate shift in what Teqnion is buying: fewer “fixing-heavy” categories, more focus on niche leadership, pricing power, and cash conversion, and a move away from structurally margin-capped business models like typical contract manufacturing or highly cyclical areas such as housing-related businesses. Overall, the 2025 cohort looks like the kind of mix that can lift group profitability over time, as long as underperformers don’t continue to offset the progress.

What surprised me most is that the better mix didn’t come at the expense of valuation discipline. Based on the pro forma earnings of the acquired subsidiaries, my estimate implies a cash-free purchase multiple of roughly ~4.2x EBITA (earn-outs included) and an immediate cash outflow multiple of about ~2.5x (earn-outs excluded). For businesses earning EBITA margins in the 20s, those are extremely attractive numbers.

These changes are now starting to show up in the numbers. Q3 2025 EBITA was up ~63% year over year. That’s an important signal, especially since 2024 was the year when the “process maturity gap” became painfully visible in reported numbers. Management ties the improvements to operational changes made during the year and highlights that the biggest impact has come from improving the Swedish businesses, while the UK operations have continued to progress steadily and the acquired companies were of higher quality. That’s a strong signal in the early stages of a serial acquirer’s “repair cycle”: the struggling bucket stops deteriorating, the portfolio stabilizes, and new cohorts come in with structurally higher profitability.

Despite the improving trajectory, the market is still pricing Teqnion as if it will remain stuck. At roughly ~13.6x P/E (excluding any assumed acquisitions for 2026), the valuation suggests investors have limited confidence in durable margin expansion and in the repeatability of the upgraded acquisition playbook. By contrast, larger Swedish peers such as Lifco, Indutrade, Addtech, and Lagercrantz currently trade in the ~25–35x range.

The valuation gap is so wide that it creates meaningful optionality. If Teqnion keeps improving acquisition quality and the upgraded operating system reduces portfolio volatility, a rerating tailwind becomes plausible, even without heroic growth assumptions. A doubling of the multiple over ten years adds about 7% p.a. to returns; over five years, it adds about 15% p.a., excluding any underlying earnings growth.

Teqnion is still very small relative to its “role models.” By revenue, it is only about ~17% of Lagercrantz and ~8% of Addtech and ~6% of Lifco. That small base matters in two ways. First, Teqnion is still earlier in its compounding journey, with a long runway if it continues to raise the quality of what it buys. Second, it has historically been more exposed to “small portfolio math”: just a handful of underperforming subsidiaries can meaningfully drag down consolidated margins (as seen during 2024).

To explain why Teqnion appears to be at an inflection point, this Deep Dive breaks the company’s evolution into four phases.

Phase 1 (2006–2009) was the capital-constrained founder era. Teqnion started without external capital and could only fund acquisitions with internally generated cash. That constraint shaped the early portfolio: cheaper businesses, often with more operational issues. As a result, the founders had to be deeply hands-on: stepping in as interim CEO, pushing sales, improving efficiency, and turning businesses around. The model worked—until it didn’t. The 2008 crisis exposed how fragile the setup was. High leverage, limited cash buffers, and aggressive reinvestment left the group vulnerable when the cycle turned. Teqnion recorded its first and only loss in that period, and one early subsidiary (UpTech Norden AB) went bankrupt in September 2009.

Phase 2 (2009–2018) was the rebuild and slow compounding phase. Teqnion was essentially restarted at a smaller scale, with Johan Steene staying on as CEO and focusing on rebuilding credibility and stability. Deal flow was constrained by limited capital and by the simple fact that most sellers want cash, which meant Teqnion often had to buy lower-quality companies that it could actually afford. As a result, Johan was frequently pulled too deep into day-to-day operations. For years, the group ran with an extremely lean HQ. That made the model operationally restricted, but it also helped shape the culture: pragmatism, cash flows, and long-term ownership. Importantly, this period also set up the next phase. By around 2017, Johan recognized that a listing and access to external capital could accelerate the acquisition pace and move Teqnion into a new scale regime.

Phase 3 (2019–2024) was the post-IPO acceleration and the period where scaling complexity started to outpace process maturity. The pre-IPO private placement and the IPO brought in meaningful capital and allowed Teqnion to move from one acquisition every year or two to multiple acquisitions per year, which accelerated growth materially. Over time, the M&A capability became more professional and the quality mix improved, including higher-margin and lower-sensitivity acquisitions (notably in the UK).

But this is also where the process maturity gap quietly opened. The portfolio grew into a more complex organism. More subsidiaries, more geographies, more variance in business quality, while the operating system (measurement, trend detection, escalation, and standardized follow-up) did not mature fast enough.

In a decentralized model, the holding company’s job is not to run subsidiaries day-to-day. Yet in 2024, management effectively acknowledged that only about two-thirds of subsidiaries performed as desired. That’s far from a steady-state ambition where perhaps 90% are fine and only 10% need special attention. The problems had accumulated over years, but a benign macro environment and maybe a bit of luck kept them from becoming visible. When interest rates rose and conditions tightened, it was like the tide going out: weaknesses that had been masked by good times became obvious.

Phase 4 (2025–?) is the current era: closing the process maturity gap. The agenda is explicit and has two parts.

• First, Teqnion is tightening governance and follow-up: standardized KPI sheets, a clearer reporting cadence, concrete action plans, faster escalation, more accountability, and a willingness to replace subsidiary CEOs if performance doesn’t improve. In that sense, Teqnion is “more hands-on” again, but in a different way than in Phase 1. It’s not about personally running subsidiaries. It’s about being hands-on in measurement, intervention, and organizational discipline.

• Second, Teqnion is upgrading its acquisition strategy: fewer startups and “turnaround” cases, higher-quality and higher-margin targets, lower macro sensitivity, and more businesses with proprietary products or defensible niche positions. In fact, this shift started a bit earlier, but 2025 is when it was being implemented at full speed and scale.

The combination of fixing the operating system while continuing to acquire at a higher pace and with better quality is what makes this moment feel different. Many serial acquirers either pause M&A to repair the portfolio (and lose compounding momentum), or they keep doing deals while problems build in the background (and eventually it blows up). Teqnion is trying to combine both: tighten the governance, reduce the negative tail, and keep compounding. The outcome is unlikely to be a sudden sustainable margin jump. More realistically, you get a slow but persistent uplift in group profitability as higher-margin cohorts build weight and the struggling bucket stops offsetting the progress.

The shareholder and management setup supports the long-term thesis. Teqnion is founder-led, with meaningful insider ownership and a relatively concentrated shareholder base that includes long-term oriented investors (the largest shareholder is the family office of Spotify founder Daniel Ek).

Serial acquirers need discipline through cycles. The models tend to fail when incentives and time horizons push management toward short-term behavior. Teqnion’s governance and ownership structure increases the odds that management can make the boring but correct decisions: walk away from bad deals, build a resilient “operating system”, intervene earlier, and accept some temporary friction to build a more durable compounding machine.

Putting it all together, the key conclusion is not that every issue has been solved. It’s that the direction of travel is now consistent across the system. Growth is once again driven by M&A, but the signals suggest the acquired cohort is of materially higher quality. Earnings power is stabilizing and improving compared with 2024. Gross margin improvements throughout 2025 support the thesis that the portfolio is gradually becoming structurally better. And while leverage has increased, the path to de-risking is clear: fix the negative tail and let higher-margin acquisitions flow through the P&L.

For an investor, the attractive part is the asymmetry. Teqnion is a small platform that has already proven ~26% p.a. revenue growth and ~35% p.a. earnings growth since inception, and it may now be entering a multi-year quality upgrade, while still trading at a multiple that implies limited confidence in that upgrade. If the “new Teqnion” becomes visible in the consolidated numbers over the next few years, any rerating would simply be the natural by-product of a better business becoming obvious, not the thesis itself. Add attractive earnings growth on top, and you could end up with very solid returns going forward.

“Momentum is back” (Johan Steene, CEO, 2025 Q3 report)

Stay tuned for Part 1 of the upcoming deep-dive!

Kinsale is one of those businesses.

It doesn’t sell code or subscriptions. It underwrites weird, non-standard risks that other insurers don’t want on their books – and still manages to post underwriting margins most peers can only dream of. It’s founder-led, focused on a single niche, and has been executing the same playbook with discipline for more than a decade.

In our Quality Growth framework, Kinsale checks a lot of boxes:

• 🏗 A focused, specialist business in a deregulated niche (E&S) with real pricing power.

• 📈 A long runway for profitable growth, driven by market share gains and E&S migration.

• 💸 Strong, repeatable unit economics with sub-80% combined ratios and attractive float.

• 🧑‍💼 An owner-operator management team with skin in the game and a conservative culture.

• 🧱 A robust balance sheet and conservative reserving, built to survive bad years, not just good ones.

This Company Snapshot is your quick guided tour: what Kinsale actually does, why its economics are different from a typical insurer, and which risks you should care about before you size the position.

If that resonates, the upcoming five-part Deep Dive will take you from this overview into the details of regulation, reinsurance, reserves, and loss triangles—the parts most investors never read, but where the real edge lives.

👔 Company Name: Kinsale Capital Group (“Kinsale”)

🔎 ISIN: US49714P1084

🔧 Business model: Specialty insurer (Excess & Surplus lines)

🌍 Geographic exposure: United States

📈 Stock Price: $393

💰 Market Capitalization: $9.1 billion

👨‍💼 Number of CEOs since foundation: 1

👨‍👩‍👦 Founder-/Owner-operator: Yes

📅 CEO tenure: Since 2009

🥇 Insider ownership: 5.5% of shares

📊 10Y EPS CAGR: ~41%

🔁 Reinvestment profile: High (float-funded growth)

💸 Capital intensity: Capital-light

🏰 Moat: Specialization, Focus, Technology & Cost Advantage

🧨 Main risks: Cyclicality, Broker concentration, Investment risk, Reserve adequacy, Regulatory risk

🌳 Slow Compounding fit: Yes – high-quality underwriting, long growth runway

Business Model in a Nutshell: Kinsale is a U.S. specialty insurer focused 100% on commercial Excess & Surplus Lines, insuring hard-to-place risks that standard carriers decline. The company runs a high-volume, low-hit-rate underwriting model, supported by a proprietary tech platform that lets underwriters quote small and mid-sized accounts quickly and selectively. Economics are driven by disciplined underwriting (combined ratio mid-70s) and float-funded investment income.

Is Kinsale the best insurance business in the world?

An insurance company can be a wonderful business.

An insurance company sells a “Collect Now. Pay later” type of product. This generates the so-called “float”, which the insurance company can use to generate investment income.

An insurance company sells coverage for risks. Customers pay a premium today. The insurance company collects the premiums, while they don’t have to pay a claim right away. The time gap between collecting the policy premium and paying out claims can be significant, especially in certain types of insurances.

During this time gap, the insurance company has a large pool of money—the “float”—sitting in its bank account. The company doesn’t own this money (it’s a liability on their books because they might have to pay it back), but they get to invest it in stocks, bonds, and other assets.

That’s why an insurance company has a double-profit-engine.

• Profit from underwriting: That’s the income from the underlying insurance business.

• Profit from the Investments: The money earned from these investments is a major source of profit for the insurance company.

Think of “float” as an interest-free loan from policyholders that the insurer uses to make more money. As investor Warren Buffett famously explained, this allows an insurer to use “free money” for its own benefit.

Kinsale looks like the kind of stock most investors scroll past: a mid-size U.S. insurer, writing commercial policies in something called the “Excess & Surplus Lines” market. On the surface that sounds like the dusty back room of the insurance world. In reality, Kinsale sits in one of the most attractive corners of U.S. P&C – and has quietly turned it into a compounding machine. Since the IPO in 2016, the share price is up more than 20-fold, powered not by financial engineering but by old-fashioned underwriting profits and disciplined growth.

The niche Kinsale serves is simple to describe: it insures risks that the standard market does not want. Kinsale insures “hard-to-place”-risks, as these insurance coverages are anything but standardized products.

When a retail broker has tried the usual admitted carriers and gets only declines, the file often lands on the desk of an E&S wholesale broker – and from there on Kinsale’s screen. The company wants to see as many of these “hard-to-place” risks as possible, then cherry-pick the ones where the odds are clearly in its favour. It operates exclusively in U.S. commercial lines, with an average policy size of around $15,000, across a broad range of property and casualty classes – from construction liability and small business casualty to catastrophe-exposed property and various niche segments.

Economically, Kinsale’s engine is built on volume, selectivity and cost discipline. Each year the company receives close to 900,000 submissions but binds only a small single-digit percentage of them. That low hit rate is a feature, not a bug: it reflects a model where underwriters are encouraged to focus on the most attractive opportunities. At the same time, retention on existing policies is healthy and average premiums have been rising. The result has been a decade of 20–30%+ premium growth, far ahead of the broader industry.

What makes this more than “just” growth is the quality of the underwriting. In an industry that, on average, barely breaks even on underwriting over a cycle, Kinsale has consistently posted combined ratios in the mid-70s, with an expense ratio in the low 20s.

The combined ratio is a crucial metric in the insurance industry:

In simple terms, these ratios are just ways for an insurance company to measure if it’s making money or losing money on the policies it sells:

• The loss ratio is the “claims” ratio. It answers the question: “For every dollar we collected in premiums, how much did we have to pay out in claims and the paperwork to handle those claims?”

• The expense ratio is the “running the business” ratio. It answers the question: “For every dollar we collected in premiums, how much did we spend on salaries, marketing, office space, and other overhead costs?”

If the combined ratio is below 100 percent, the insurer is generating an underwriting profit from its core business. However, if the ratio exceeds the 100 percent mark, it signifies that the insurance operations are running at a loss, which must be offset by returns from investments.

This key figure thus provides a quick overview of the profitability and financial health of an insurance company.

And guess what:

Kinsale is the most profitable insurance company that I have ever seen in my investment career. Over the last 3 years, Kinsale posted an average combined ratio of ~77%, while the average for the E&S competitors is closer to ~92 %. That’s a huuuuuge difference. (It represents an EBIT-margin of ~23% for Kinsale vs. ~8 % for the competition. Kinsale makes a 3x higher EBIT contribution from every dollar earned in premiums.)

That isn’t a small efficiency tweak, it is a structural gap – driven by a focused portfolio, tight control of underwriting and claims, and a proprietary technology platform that automates the routine work and lets underwriters concentrate on risk selection. The company writes business across all 50 states, uses reinsurance intelligently to shape its risk profile, and converts a large portion of accounting earnings into cash.

Behind the machine is a founder-led management team with real skin in the game: Kinsale has had one CEO since 2009, insider ownership is meaningful, and capital allocation has been refreshingly straightforward – reinvest into the core underwriting engine as long as attractive opportunities are available, keep leverage conservative, and avoid empire building. The business is capital-light, benefits from float, and has grown EPS at ~40% per year over the last decade.

The best is yet to come: Despite its phenomenal success over the last one and a half decades, Kinsale today has a market share of ~1.6 %. The growth runway is huge.

Currently, management views the stock as too cheap, as they announced a $250 million share-buyback program and reiterated the quality of the underlying business model.

This is not a risk-free story (no investment is risk-free): insurance is cyclical, Kinsale is exposed to catastrophe losses and to a concentrated broker base, and the regulatory framework for E&S and reinsurance matters a lot. Valuation also reflects much of the company’s success. But if Kinsale can preserve its underwriting edge and continue to take share in a still-fragmented E&S market, it has the ingredients of a long-term compounder: a focused niche, structural cost and speed advantages, a long runway for reinvestment and owner-operator governance.

The following Deep Dive will unpack each piece of that puzzle – from the plumbing of E&S and reinsurance to Kinsale’s loss triangles, technology stack and runway for growth.

It will be structured in five Parts plus one additional fundamental piece:

• Part 1: Insurance 101 and E&S playground

• Part 2: Regulation: Why E&S looks the way it does

• Part 3: Kinsale’s Business Model

• Part 4: The Kinsale machines: Understanding the Financials

• Part 5: Compounder?

• “The Fundamentals”: Reinsurance 101

This Company Snapshot is the 10,000-foot view. In the Deep Dives, we’ll zoom in on three layers:

1. The basics of the U.S. E&S market and surplus-lines regulation that makes Kinsale’s niche possible;

2. The inner workings of its underwriting machine – from submission funnel and pricing discipline to claims handling, technology stack and reinsurance program;

3. The hard numbers behind the story, including combined ratio drivers, reserve development, capital allocation and valuation.

Along the way we’ll dig into material you’ll rarely see discussed elsewhere: Kinsale’s statutory loss triangles and what they reveal about reserve quality, how reinsurance flows through the P&L and impacts the balance sheet, and much more.

Very few actually build a portfolio that forces them to think in decades.

In this article, I’m opening up my own portfolio – 20 businesses I’m comfortable owning for a very long time. From Nordic serial acquirers like Lifco, Teqnion and Röko to compounders like Novo Nordisk, Copart and Topicus, this is the real-world expression of my investment philosophy: slow compounding, high returns on capital, and management teams that know how to allocate cash.

I’m sharing it because I believe the best way to understand an investor is to look at where their own money lives. This portfolio is my checklist in action:

High ROIIC, reinvestment runways, sensible leverage, and a culture that cares about creating long-term shareholder value rather than pleasing the next quarter.

Think of this article as a trailer for what’s coming next on this Substack.

Every single company you see below will get its own:

• Company Snapshot – a concise, reader-friendly overview of the business model, economics and key metrics.

• Deep-Dive – a more detailed breakdown of the thesis, capital allocation, risks, valuation and how it fits into my “Slow Compounding” and 3-Engines framework.

If you’re curious about why Kinsale dominates a niche part of insurance, how InPost’s parcel lockers can scale, or what makes Lifco’s acquisition machine so powerful, those answers are coming – company by company.

This is the roadmap for the next months of content: I’ll go down the list, explain the logic behind each position, and update you when the facts or valuations change. If you want to follow that journey in real time – and discover the companies that might be worth holding for years, not weeks – now is the perfect moment to subscribe.

Subscribe now

Before we zoom into individual companies, it’s worth taking a step back and looking at the portfolio as a whole.

The Portfolio at 10,000 Feet

(1) Position sizing & concentration:

The first chart shows the weight of each position. Position sizes are deliberately kept in a relatively tight band: the largest holdings – AQ Group, Mader Group, Novo Nordisk, Kinsalte and Medpace – currently sit around 6–7%, most others cluster between 3–5%.

This reflects the way I think about risk: I want every position to be large enough to matter, but I don’t want any single stock to be able to derail the portfolio on its own.

(2) YTD performance - price is noisy

The second chart shows year-to-date stock performance. It’s a good picture of how noisy markets can be in the short term.

TerraVest, Mader and Medpace have had a very strong year so far, while several other holdings – from Novo Nordisk and Copart to Judges Scientific and InPost – are meaningfully down. The underlying businesses did not change that dramatically ytd; sentiment and multiples did. Note that I didn’t own Novo Nordisk and Judges Scientific at the beginning of the year and I only have built the position during summer.

(3) Drawdowns

The third chart shows drawdowns from the respective share price highs. Many holdings are sitting 30–60% below previous peaks; a few have even seen drawdowns of ~70%.

For me, this is the real‐world “cost of admission” for long-term equity returns. If you want to participate in multi-year compounding, you have to accept that even great businesses will periodically look terrible on a price chart.

Of course, it’s painful to see others making a lot of money with Nvidia, Palantir and other AI-related companies right now. The S&P 500 is up more than 17 % YTD. It doesn’t show up in the performance of my holdings this year.

It literally feels like this:

(4) Earnings power - EBIT(A) growing at a blended ~13.7 % this year (taking into account current portfolio weight)

The fourth chart shows EBITA growth over the last twelve months. Here, the picture looks very different from the price action.

Most companies in the portfolio are growing EBITA at healthy single- to double-digit rates, with stand-outs like Medpace, TerraVest, Dino and Topicus posting very strong numbers. A few names are temporarily flat or down – that’s part of the game – but at the portfolio level, earnings are moving up for the portfolio, not down.

This gap between fundamentals (EBITA growth) and market perception (drawdowns) is where I see opportunity. Over time, I expect returns to track earnings power far more closely than they track quarterly mood swings. Aside from Brown & Brown and TerraVest Industries, all growth came without any dilution!

(5) Geographic mix - a Nordic & North American tilt

The fifth chart breaks the portfolio down by country of listing. Roughly:

• ~a third of the portfolio is in Sweden,

• ~a quarter in the US,

• with additional exposure to Canada, Germany, Denmark and Poland,

• plus smaller positions in Australia, the Netherlands and the UK.

This is not a top-down macro call. It’s simply where I currently find the kind of businesses I’m looking for: capital-light compounders, often founder-influenced, with long reinvestment runways.

(6) Geographic mix - a Nordic & North American tilt

The final chart shows the portfolio split by business model:

• roughly half in serial acquirers – decentralised groups like Lifco, Röko, Addtech, Teqnion, Judges, Momentum, TerraVest and Topicus,

• roughly half in other compounders – insurance specialists, healthcare and pharma, logistics and software,

• plus a small cash buffer.

Serial acquirers are attractive to me because they can reinvest large amounts of cash into small, repeatable deals at high incremental returns (soon a separate post about this will follow). The other compounders often have different engines – pricing power, volume growth, operating leverage, buybacks – but the goal is the same: turn today’s cash flows into a much larger stream of cash flows tomorrow.

Taken together, these charts show what I’m trying to build: not a portfolio that looks good every quarter, but a collection of businesses that can quietly compound value over many years – even if the share prices take a very noisy path to get there.

Subscribe now

The Holdings - 1-Minute Business & Investment Case

AQ Group AB

Business in one sentence
AQ Group manufactures components and systems such as electrical cabinets, inductive components, wiring systems and sheet metal parts for demanding industrial customers around the world.

Why I own it

• Deep integration into customers’ products and processes – switching costs can be meaningful.

• Exposure to long-term trends like electrification, automation and industrial outsourcing.

• Decentralized, engineering-driven culture with a long track record of profitable growth.

• Very well managed serial acquirer flying under the radar of many investors despite their long-term success and high cash-generation.

Mader Group

Business in one sentence
Mader Group is an Australian technical services provider that maintains heavy mobile and fixed equipment for customers in mining, energy, transport and other industrial sectors worldwide.

Why I own it

• Asset-light service model: selling skilled labour and uptime, not building big factories.

• Mission-critical: broken equipment costs clients far more than Mader’s service fees.

• Attractive economics when utilization is high and the installed base of heavy equipment keeps growing.

• They have a >50 % market share in Australia with ~12% EBITDA margin, still growing >15% currently. They entered Canada & U.S. a few years ago. This market is several times bigger than Australia, and penetration is much lower, while EBITDA-margin comes in at ~20% in North America.

• Founder-influenced culture, with the founder Luke Mader owning ~51 % and Non-Executive Board member Craig Burton owning ~18 %.

Novo Nordisk

Business in one sentence
Novo Nordisk is a global leader in diabetes and obesity treatments, driven by its GLP-1 drug franchise (e.g. Ozempic, Wegovy) and a broader cardiometabolic pipeline.

Why I own it

• Enormous secular tailwind: rising global rates of obesity and diabetes.

• Strong R&D engine and manufacturing scale in biologics, difficult for new entrants to replicate.

• Optionality from expanding the GLP-1 platform into cardiovascular and other indications.

• Risks (competition, pricing, political pressure) are real, but the economic engine is very powerful.

• The global market is huge. Despite losing market share to Eli Lilly lately, there should be enough growth for both current duopolists.

Kinsale Capital Group

Business in one sentence
Kinsale is a U.S. specialty insurer focused exclusively on the excess & surplus (E&S) lines market, insuring risks that standard insurers often avoid.

Why I own it

• Underwrites non-standard risks where pricing is more rational and less commoditized.

• Strong underwriting culture with a long history of extremely attractive combined ratios.

• Asset-light business model: most capital sits in the investment portfolio, not in fixed assets.

• Long growth runway as Kinsale’s market share is still tiny, while they are taking ~0.2% in market share every year

Medpace Holdings

Business in one sentence
Medpace is a global contract research organization (CRO) that runs clinical trials for biotech, pharma and medical device companies from Phase I through Phase IV.

Why I own it

• “Picks and shovels” to the biotech and pharma R&D cycle, without betting on individual drugs.

• Focused, mid-sized player – more nimble and aligned than some of the mega-CROs.

• Asset-light, people-and-process business with high margins and strong cash generation.

• Owner operator, as the founder is still involved as CEO.

• Long runway as more development work is outsourced and trial complexity increases.

TerraVest Industries

Business in one sentence
TerraVest is a Canadian serial-acquirer industrial group focused on manufacturing niche equipment like propane and NGL transport and storage vessels, home heating products and energy-processing equipment.

Why I own it

• Strong positioning in small but critical segments of energy and infrastructure.

• Cash-generative, often family-owned businesses rolled into a disciplined capital allocation framework.

• Management with a track record of value-accretive M&A and operational improvement.

• Exposure to replacement demand and safety/regulatory spend, which tends to be less cyclical.

Teqnion AB

Business in one sentence
Teqnion is a Swedish industrial group and serial acquirer that buys and operates small niche technology and industrial companies, run in a decentralized way.

Why I own it

• Typical Nordic serial acquirer playbook: small deals, disciplined pricing, decentralized governance.

• Portfolio of niche businesses with sticky customer relationships and specialized know-how.

• Focus on profitability and cash generation rather than chasing revenue at any price.

• Long runway: the universe of small industrials in Scandinavia and beyond is huge.

• Despite some operational headwinds over the last 24 months, they have refoucsed their M&A activity on higher-quality companies (many in UK lately), with the first impact showing off in Q3/2025. In my opinion, the market does not yet recognize the tremendous improvement in ROIIC profile going forward.

Medios AG

Business in one sentence
Medios is building a leading European Specialty Pharma platform, focusing on the supply and compounding of medicines for complex, chronic and rare diseases.

Why I own it

• Positioned in a structurally growing niche: specialty drugs and personalized medicine.

• Integrated model across wholesale, compounding and related services creates network effects with pharmacies and clinics.

• Regulatory know-how and quality systems are a real competitive moat in this space.

• Scalable platform economics if management continues to execute M&A and integration sensibly.

• The company is transitioning into higher margin-businesses and better cash conversion.

• The Market Capitalization sits at €350 million and the Enterprise Value (EV) comes in at ~€455 million. Over the last 4 quarters, Medios has generated ~€63 million in Free-Cashflow.

Evolution AB

Business in one sentence
Evolution is a B2B provider of live casino and online gaming content, licensing live-dealer tables and game shows to online casinos worldwide.

Why I own it

• Powerful platform economics: studios and game IP can be monetized globally at very high incremental margins.

• Revenue share model aligns Evolution with operators’ success and creates “recurring” income.

• Strong product innovation and brand with a long runway in both live casino and game shows.

• Highly cash-generative business with opportunities for both reinvestment and shareholder returns.

• Evolution has been facing regulatory and operational headwinds over the last 24 months.

Judges Scientific PLC

Business in one sentence
Judges Scientific is an AIM-listed buy-and-build group that acquires and develops niche scientific instrument businesses with strong global reputations.

Why I own it

• High-margin, high-moat products used in research, industry and academia.

• Extremely disciplined acquisition approach with a focus on returns and cash payback.

• Conservative balance sheet and culture – “boring” in the best possible way.

• Many of the acquired companies are founder-led, niche leaders with sticky customer bases.

Dino Polska S.A.

Business in one sentence
Dino Polska operates one of the fastest-growing chains of mid-sized proximity supermarkets across Poland, focused on everyday grocery needs close to where people live.

Why I own it

• Attractive “local discounter” format with strong customer frequency and basket stability.

• Integrated supply chain and own logistics support good margins and execution.

• Inflation and private-label trends can be a tailwind for efficient mid-market grocers.

• High store rollout potential with still-low saturation in many regions. With just 7.8 stores per 100k inhabitants on average country-wide and the most dense voivodship constantly increasing the bar while not showing any signs of a leveling-out with 19.2 stores per 100k inhabitants now, there should remain a long growth runway going forward.

Lagercrantz Group AB

Business in one sentence
Lagercrantz is a Swedish technology group and serial acquirer, owning >100 niche companies that provide value-creating technical solutions in electronics, communications, mechatronics and other segments.

Why I own it

• Classic decentralized serial acquirer in the Nordics with a long history of smart acquisitions and capital allocation.

• Focus on niche B2B markets where small companies can hold very strong positions.

• Strong culture around ROCE, pricing power and cash generation.

• Structural tailwind as industrial customers outsource more specialized technology solutions.

Brown & Brown, Inc.

Business in one sentence
Brown & Brown is one of the world’s largest insurance brokers, earning commissions and fees for arranging coverage and providing risk management services.

Why I own it

• Asset-light, fee-based model: Brown & Brown doesn’t carry the underwriting risk on its own balance sheet.

• Recurring, diversified revenue across industries, geographies and product lines.

• Long acquisition history of buying local agencies and integrating them into a performance-driven culture.

• Rising complexity of risks (cyber, climate, etc.) makes advice more valuable, not less.

Addtech AB

Business in one sentence
Addtech is a Swedish technical solutions group made up of ~150 independent companies selling high-tech products and solutions to industrial and infrastructure customers.

Why I own it

• Proven serial acquirer model: small, disciplined deals in niche technical areas.

• Decentralized structure that keeps entrepreneurship alive at the subsidiary level.

• Exposure to long-term themes like electrification, automation and infrastructure upgrades.

• Strong track record of compounding through a mix of organic growth and acquisitions.

Copart, Inc.

Business in one sentence
Copart runs an online auction platform for salvage and used vehicles, connecting insurers, dealers and buyers across a global network of yards and marketplaces.

Why I own it

• Two-sided marketplace with powerful network effects and data advantages.

• Structural growth in total loss vehicles, used car demand and emerging markets.

• Highly scalable digital platform with attractive incremental margins as volumes grow.

• Substantial real-asset footprint (land and yards) that would be hard to replicate from scratch.

InPost SA

Business in one sentence
InPost is a leading out-of-home e-commerce logistics platform in Europe, best known for its network of automated parcel lockers and pickup points across multiple countries.

Why I own it

• Locker-based last-mile delivery is cheaper, more convenient and provides much superior unit-economics compared to home-delivery.

• Strong first-mover advantage and network density in key markets like Poland and the UK.

• Significant operating leverage as locker utilization rises over time.

• Huge untapped market potential outside of Poland.

(Note: APM stands for Automated Parcel Machine, i.e. self-service parcel lockers)

Momentum Group AB

Business in one sentence
Momentum Group is a Nordic serial acquirer that acquired companies that supply industrial components, maintenance services and related solutions to industrial customers and infrastructure.

Why I own it

• Picks-and-shovels exposure to the health of Nordic industry and infrastructure.

• Recurring demand from MRO (maintenance, repair and operations) activities.

• Decentralized structure with specialist local brands close to the customer.

• Opportunity to roll up more niche distributors and service providers over time.

Lifco AB

Business in one sentence
Lifco is a Swedish serial acquirer that owns a large portfolio of niche businesses across three segments: Dental, Demolition & Tools, and Systems Solutions.

Why I own it

• One of the “blueprint” Nordic serial acquirers with a phenomenal long-term track record.

• Very decentralized, profit-focused culture with minimal bureaucracy.

• Disciplined acquisition criteria and a long runway in European SMEs.

• Strong capital allocation history and a management team that thinks in decades, not quarters.

Topicus.com

Business in one sentence
Topicus.com is a pan-European vertical market software (VMS) group, spun out of Constellation Software, focused on mission-critical software platforms for public and private sector niches.

Why I own it

• Classic VMS economics: high switching costs, sticky customers, recurring SaaS-like revenues.

• Backed by the Constellation playbook and culture of disciplined, decentralized M&A.

• Vast runway in under-penetrated European software niches.

• Strong alignment with long-term compounding through reinvesting free cash flow into acquisitions.

Röko AB

Business in one sentence
Röko is a Swedish serial acquirer and perpetual owner of small, profitable, asset-light businesses across Europe, often buying majority stakes while leaving founders invested.

Why I own it

• Co-Founded and led by former Lifco-CEO Fredrik Karlsson with deep experience in the Lifco model of compounding through acquisitions.

• Focus on high-margin, asset-light niche leaders with strong market positions.

• Minority-retention model keeps founders and managers aligned for the long run.

• Still early in its listed life, with a long runway to compound if discipline is maintained.

• Scaled from 0 to SEK 1,323 million (~$140 million) in EBITA within 6 years.

What to Expect If You Subscribe

On the Substack, my plan is to:

• Publish company snapshots for portfolio companies and for companies on the watchlist (as I seek to further increase the quality of the portfolio)

• Publish deep dives

• Write framework pieces on topics like ROIIC, capital allocation, and the “engines” of long-term shareholder returns (and more).

If you’re interested in discovering the companies worth holding for years, not chasing the story of the week, then this is what you can expect here.

👉 If that resonates with you, consider subscribing to follow the journey and the upcoming deep dives on these businesses.

Subscribe now