THE OUTLOOK

There are approximately 6.1 million in-ground swimming pools in the United States. Each one needs chemicals, replacement parts, and regular maintenance throughout its lifetime, regardless of what interest rates are doing, regardless of whether anyone is building new pools, regardless of the broader economy. The market for maintaining these pools does not disappear in a downturn. It grows, steadily, as the installed base expands year after year.

Pool Corporation (POOL) sits at the wholesale distribution layer between the manufacturers who make pool products and the 125,000 contractors, service companies, and retailers who sell those products to end consumers. It is the largest business of its kind in the world, with 456 sales centres, approximately 40% of the US wholesale distribution market, and a flywheel of scale that has been compounding for over four decades.

The business is currently earning below its normalised level. New pool construction fell to just below 60,000 units in 2025, roughly half of pandemic-era peak volumes. Elevated financing costs have suppressed homeowner discretionary spending on pools, renovations, and upgrades for three consecutive years. The result is a peak-to-trough earnings-per-share (EPS) compression from $18.70 in 2022 to $10.85 in 2025.

The maintenance revenue base, which represents approximately 64% of Pool’s sales and comes from the inelastic, recurring demand of existing pool owners, held steady throughout this period and grew modestly as the installed base expanded. The competitive position did not weaken. Gross margins held within a few basis points of their historical range. Return on invested capital (ROIC) declined from its peak but remained above 16%, above the cost of capital, in a trough year for the industry. No meaningful market share was lost to competitors.

Pool Corporation is Approved in the Bearhold Universe. The business has demonstrated over four decades that its scale advantages compound, its competitive position strengthens through cycles, and its maintenance revenue base is structurally permanent.

My valuation puts Pool at approximately 15 years of embedded discounted cash flows at current prices. I hold a position initiated before publication of this report. This report explains the business, the thesis, and where the risks sit.

Disclosure: The author holds a position in POOL. This report reflects the author’s personal views and is not an investment advice. Investing carries the risk of permanent capital loss. Read the full disclaimer here

At a Glance

Company: Pool Corporation

Ticker: $POOL · NASDAQ

Sector: Industrials

Industry: Wholesale Distribution

Market Cap: $8.3 billion (at $225)

Status: Approved

First Coverage: April 2026

Valuation Zone: Attractive (last updated in April 2026)

1. The Business

The Origin

Pool Corporation traces its roots to 1980, when Frank St. Romain, a former warehouse manager, and his partner Richard Smith founded South Central Pool Supply in New Orleans. The founding insight was simple and durable: insurance companies and pool professionals needed a professional, geographically distributed source for pool supplies and equipment. The fragmented local dealers serving that need were doing it poorly.

For its first decade, South Central Pool Supply grew modestly, building a reputation for service and expanding steadily across the Southeast and into Oklahoma, Texas, and Tennessee. Growth was constrained by access to capital, as then-president Manuel Perez de la Mesa later described it, the company funded expansion almost entirely from retained earnings. By 1993, annual revenues had reached approximately $67 million.

The inflection came in 1993, when private equity firm Code Hennessey & Simmons acquired the company in a leveraged buyout (LBO), incorporated it in Delaware as SCP Holding Corp, and set out to use South Central Pool Supply as a platform for aggressive national consolidation. The strategy was straightforward: the US pool distribution market was deeply fragmented, and a well-capitalised consolidator with the right model could roll it up. Wilson B. ‘Rusty’ Sexton became Chairman and Chief Executive Officer (CEO).

In 1995, the company renamed itself SCP Pool Corporation and went public on the NASDAQ stock exchange under the ticker POOL, raising capital that retired its debt and funded further expansion. The initial public offering (IPO) proceeds unleashed the acquisition campaign: within five years, SCP had acquired dozens of regional distributors and opened scores of new sales centres. Annual revenues grew from $102 million in 1994 to $670 million by 2000.

A pivotal expansion came in August 2000, when SCP acquired Superior Pool Products from Arch Chemicals, at the time a 19-location distributor operating primarily in California, Arizona, and Nevada with revenues exceeding $80 million. Rather than absorbing Superior into the SCP network, management made the strategic decision to operate two distinct domestic distribution networks side by side, each with its own product selection, sales personnel, and supplier relationships. The logic was deliberate competition between internal networks to drive service quality, an unusual approach that has proven effective.

Manuel Perez de la Mesa, who joined as Chief Operating Officer (COO), became President and CEO in 2001 and led the company for nearly two decades. Under his leadership, Pool expanded internationally, launched the Horizon Distributors irrigation business, and broadened the product offering far beyond pool chemicals into building materials, outdoor living, and commercial pool products. At the end of 2018, Peter D. Arvan succeeded him as CEO, bringing over 20 years of distribution industry experience. On May 16, 2006, SCP Pool Corporation became Pool Corporation, a single name for what had become the world’s dominant force in pool wholesale distribution.

What Pool Corporation Does

Pool Corporation does not build pools. It does not install them, service them, or repair them. It is the wholesale distribution layer, the intermediary that sits between manufacturers and the fragmented network of pool professionals who serve end consumers.

When a pool builder needs concrete, pumps, filters, tiles, and chemicals for a new installation, they order from Pool. When a service technician needs a replacement part for a malfunctioning heater, they call Pool. When a specialty retailer needs to restock their shelves with chemicals and accessories, they order from Pool. Pool takes possession of product from over 2,200 manufacturers, holds inventory across 456 geographically distributed sales centres, and delivers to contractors and retailers, often the same day. It earns a margin on every product it passes through the supply chain.

This hub-and-spoke model is the foundation of the business. Pool does not compete nationally in any meaningful sense, it competes market by market. Each sales centre is positioned within a population centre, close to contractor customers, with well-stocked inventory and local delivery capability. A pool professional in suburban Houston does not care about national market share; they care whether their local Pool centre can get them what they need, today, at a competitive price. Pool’s network is built to answer yes to that question in virtually every significant pool market in North America.

The Five Networks

Pool Corporation operates through five wholly-owned subsidiary distribution networks, each serving a distinct segment of the market. Understanding these networks matters because they are not divisions of the same business in a superficial sense, they have different brand identities, different product selections, and in some cases, different types of customers. Pool maintains them separately by design.

SCP Distributors (SCP) is the original network, descended from South Central Pool Supply. It is the largest of the five, operating 224 sales centres domestically and 48 internationally as of year end 2025. SCP handles the full range of swimming pool supplies, equipment, and related leisure products for the domestic market and serves as Pool’s primary international distribution platform.

Superior Pool Products (Superior) is the second domestic swimming pool network, acquired in 2000 and deliberately maintained as a separate operation. Superior operates 75 centres, primarily serving pool builders, service companies, and retailers with a different product mix and distinct supplier relationships from SCP. The rationale for two separate domestic pool networks is intentional competitive tension: SCP and Superior compete in many markets, which management believes drives better service quality than a single monopoly distributor would provide. This may seem counterintuitive, why compete with yourself? but the track record suggests it works.

Horizon Distributors (Horizon) is Pool’s irrigation and landscape maintenance subsidiary, operating 88 sales centres focused on irrigation system components, professional turf care equipment, hardscapes, and landscape maintenance supplies. Horizon serves landscape contractors and commercial operators rather than pool professionals. The business shares many characteristics with the pool distribution model, fragmented customers, recurring demand, local delivery, and benefits from Pool’s purchasing scale and supplier relationships.

National Pool Trends (NPT), formerly known as National Pool Tile until a rebrand in November 2025, is a specialised network focused on swimming pool tile, composite pool finishes, decking materials, and interior pool surfacing products. NPT operates 19 standalone sales centres, but its reach is substantially broader: 124 SCP and Superior locations also feature NPT consumer showrooms where contractors and homeowners can view and select tile, decking, and surfacing options. NPT’s products are more discretionary in nature than chemicals or replacement parts, making this segment more sensitive to the construction and renovation cycle.

Sun Wholesale Supply is the most recently added network, acquired as part of Pool’s December 2021 purchase of Porpoise Pool & Patio. Sun Wholesale’s primary role is servicing the Pinch A Penny franchise network, a chain of independently owned specialty retail pool stores, the largest pool franchise system in the United States. Sun Wholesale supplies these franchisees with pool chemicals, equipment, and accessories, and owns a chemical re-packaging plant in Florida that produces proprietary branded products sold through the Pinch A Penny system. The Pinch A Penny network grew to over 300 franchise locations by year-end 2025, adding 10 new stores during the year and expanding into two new states.

The Revenue Mix

In 2025, approximately 64% of Pool’s sales came from recurring maintenance and minor repair of existing pool installations. These are the chemicals, replacement parts, and minor consumables that pool owners must purchase to keep their pools functional and safe, the non-discretionary foundation of the business. Approximately 22% came from remodeling, renovation, and upgrades, and the remaining 14% from new pool construction. This structure matters because the 64% non-discretionary base does not compress in downturns. It grows every year with the installed base, providing the revenue floor that sustains the business through every economic cycle.

Scale and Geography

Pool’s 456 sales centres at year-end 2025 are disproportionately concentrated in the markets with the greatest pool density. California, Florida, Texas, and Arizona together account for approximately 53% of annual net sales. Pool operates 77 locations in California, 67 in Florida, 55 in Texas. These are also the states receiving the largest net domestic migration inflows. The demographic tailwind of population movement toward warmer climates creates pool installation demand and maintenance volume growth that flows directly to Pool’s most dense network regions.

Internationally, Pool operates in Europe, primarily through the UK, France, Spain, Germany, Belgium, Croatia, Italy, and Portugal, and in Australia. International operations represented approximately 5% of net sales in 2025, growing 5% in local currency terms. Europe is structurally earlier in its pool distribution evolution than the US, representing a long-runway optionality rather than a near-term growth driver.

2. The Moat

Pool’s competitive advantage is scale economies shared, one of the most durable forms of moat in distribution because it is self-reinforcing and strengthens with each passing year.

How the Flywheel Works

Pool’s scale allows it to purchase from 2,200 suppliers at better prices than any regional distributor can negotiate. It passes a portion of those better economics to contractor customers as better pricing, better product availability, and faster delivery. Contractors consolidate purchasing with Pool. Pool’s volumes grow. Purchasing leverage increases. Smaller competitors are squeezed on margins and cannot match Pool’s service levels. They lose customers or exit. Pool absorbs their volume. Repeat.

This flywheel has been running for over 40 years. Revenue per sales centre has grown at nearly 6% per year since 2014, outpacing the approximately 3% annual growth in centre count over the same period. Each location is becoming more productive, not just the network expanding. That is the flywheel operating.

Four Pillars

The first pillar is purchasing leverage from scale. Pool’s volumes across 2,200 suppliers give it negotiating power that no regional competitor can match. Its preferred vendor programme concentrates purchasing to extract better terms, rebates, early-buy discounts, priority allocation, which flow into gross margin stability even as the revenue mix shifts through cycles.

The second pillar is product breadth. Pool offers over 200,000 products across more than 700 product lines and 40 product categories. A contractor purchasing from a smaller regional distributor faces a fundamentally inferior selection. In a service business where time is money, the convenience of a single trusted source with comprehensive inventory is worth more than marginal savings on individual items. No competitor can replicate this breadth without decades of supplier relationship building and the capital to finance the inventory position.

The third pillar is local delivery infrastructure. Pool’s sales centres sit within population centres near customer concentrations. The ability to deliver same-day or next-day to contractors in their local market is operationally critical. A pool professional who needs a pump motor today cannot wait for an online order. This geographic density is a physical infrastructure moat, expensive to build, slow to replicate, and self-reinforcing as Pool adds 8 to 12 new centres per year.

The fourth pillar is POOL360 digital integration. POOL360 is Pool’s proprietary business-to-business (B2B) digital platform, which includes e-commerce ordering, water testing software (POOL360 WaterTest), and a field service management application (POOL360 PoolService) that allows pool professionals to manage their scheduling, routing, billing, and customer relationships from a single mobile application. A contractor who has embedded POOL360 into their daily workflow is not switching distributors to save 2% on one product category. POOL360 reached an all-time high of 15% of total sales in FY2025, up from 10% in 2023. The trend has been consistently upward for four consecutive years, which is the most important forward-looking signal in the moat analysis.

Financial Evidence

Gross margins averaged approximately 29% to 30% across every year of the past decade, including the trough years of 2023 through 2025. A business losing its competitive position would see gross margins compress as it discounts to retain volume. Pool’s gross margins have not compressed.

ROIC averaged above 20% over the decade, and in 2025, with earnings per share (EPS) having fallen from a peak of $18.70 in 2022 to $10.85, ROIC still sits at approximately 16%. For a wholesale distribution business with no proprietary physical products, sustaining returns above the cost of capital through a decline of that magnitude is a meaningful signal that the structural economics of the business are intact. The capital invested in the network is generating returns. No meaningful market share has been lost to competitors in this period.

The supplier concentration itself confirms market power. Pool’s top three suppliers, Pentair, Zodiac, and Hayward, collectively account for approximately 43% of Pool’s cost of goods sold (COGS). These are large, sophisticated manufacturers who sell through Pool because Pool is the most efficient route to the fragmented contractor market. They depend on Pool. That dependency is evidence of distribution leverage, not exposure.

3. Financial Performance

The financial record of Pool Corporation over the past ten years tells two stories simultaneously. The first is structural: a business with genuine competitive advantages compounding revenue, earnings, and free cash flow at exceptional rates over the long arc. The second is cyclical: a business whose earnings expanded dramatically during an extraordinary pandemic-era demand surge and have since normalised back toward their historical range. Reading the numbers correctly requires keeping both stories in view.

Revenue

Revenue grew from $2.36 billion in 2015 to a peak of $6.18 billion in 2022, a compound annual growth rate (CAGR) of approximately 15% over seven years. Since 2022, revenue has declined to $5.29 billion in 2025, entirely reflecting the reduction in new pool construction and renovation activity. The maintenance revenue base did not contract. The compression is entirely in the discretionary segment.

The Pandemic Surge and the Normalisation

Understanding the operating margin trajectory requires understanding what happened during 2020 through 2022. For the decade prior to the pandemic, Pool’s operating margin averaged approximately 10%, a stable, predictable range that reflected the economics of a distribution business with consistent gross margins and a fixed cost structure.

The COVID-19 pandemic created an extraordinary and temporary departure from that baseline. As families spent more time at home and sought to create or expand outdoor living spaces, demand for new pool construction surged dramatically. Combined with homeowners who were already invested in their properties, supported by rising home values, low mortgage rates, and a perception that their homes had become their primary entertainment and recreation venues, Pool’s revenue grew from $3.9 billion in 2020 to $6.2 billion in 2022. Operating leverage on a fixed cost base that could not scale as fast as revenue drove operating margins to 15.7% in 2021 and 16.6% in 2022, levels that had never been seen in the company’s history.

Since the second half of 2022, this surge has reversed. Elevated interest rates have suppressed new pool construction. Consumer discretionary spending has moderated from pandemic highs. Revenue has declined from its peak, and operating margins have returned to approximately 11% in 2025, essentially the pre-pandemic normal. This is not deterioration. The operating margin has returned to its historical range, not fallen below it. The 2021 and 2022 margins were the anomaly.

Earnings Per Share

Diluted EPS grew from $2.90 in 2015 to $18.70 at the 2022 peak, a 7-year CAGR of approximately 30%. From the peak, EPS has declined to $10.85 in 2025, reflecting the revenue normalisation described above. Neither number represents the mid-cycle earning power of the business. The $18.70 incorporated extraordinary pandemic demand; the $10.85 reflects a construction market at approximately half of normalised levels. The normalised earning power sits somewhere in between, and the Valuation section addresses this directly.

The share count has declined from approximately 44.3 million diluted shares in 2015 to 37.3 million in 2025, a reduction of approximately 16% through consistent buyback programmes. In 2025 alone, Pool repurchased $341 million of its own shares. That discipline compounds EPS growth on top of the underlying business performance.

Gross Margin

Gross margin has held in a narrow band of 28.6% to 31.3% across the full decade. The 2025 gross margin of 29.7% is essentially identical to 2024. Adjusting for the one-off $12.6 million import tax reversal that benefited 2024’s gross margin, the underlying 2025 gross margin actually improved 20 basis points year over year, reflecting Pool’s pricing discipline and supply chain management even in a softer demand environment.

Free Cash Flow and the OCF Story

Operating cash flow (OCF) was $888 million in 2023, $659 million in 2024, and $366 million in 2025. The decline from 2024 to 2025 appears dramatic but is substantially explained by working capital timing rather than a deterioration in the underlying business.

Three components of working capital combined to suppress OCF in 2025. First, inventory increased by $165 million during the year. Management explicitly attributed this to strategic purchasing ahead of anticipated vendor price increases, a deliberate decision to build stock before tariff-related and inflation-driven cost increases took effect. Second, accounts receivable increased by approximately $33 million, primarily driven by higher sales in December 2025, which fell outside the collection window for year-end cash. Third, accounts payable declined by approximately $44 million as Pool paid down supplier balances at a faster rate relative to the elevated payable levels of 2024, when the company had benefited from extended seasonal payment terms. These three working capital movements together subtracted approximately $150 million from operating cash flow. The fourth and largest factor was a $68.5 million federal income tax payment in 2025 that had been deferred from 2024, a one-time cash outflow with no impact on the underlying profitability of the business.

Adjusting for the deferred tax payment alone, 2025 OCF would have been approximately $434 million, or 107% of net income, consistent with Pool’s historical cash conversion rate. Free cash flow of $310 million therefore understates run-rate cash generation. The 2024 OCF of $659 million, with normal working capital dynamics, is the more reliable anchor for what the business can produce at stable conditions.

Inventory Days: A Decade-High That Deserves Scrutiny

Days inventory outstanding (DIO), the number of days Pool holds inventory before it is sold, has climbed steadily from approximately 96 days in 2020 to 135 days in 2025, the highest level in a decade. This is worth examining carefully because elevated inventory days affect multiple financial metrics simultaneously: they tie up working capital, reduce free cash flow, and, if the inventory cannot be sold at full margin, compress profitability.

The 2025 inventory build reflects a deliberate strategic decision by management. Pool purchased ahead of anticipated price increases, both from inflation running at approximately 2% to 3% in 2025 and from tariff impacts that management expected to flow through the supply chain. Pool explicitly disclosed this in the Q4 2025 earnings presentation, noting that inventory growth reflected inflation, strategic pre-buy purchasing, acquisitions, and new sales centre additions.

The bull case for this strategy is straightforward: if vendor prices rise as expected, Pool’s early-bought inventory sits on the balance sheet at below-market cost, and when sold, generates better margins than inventory purchased at higher post-increase prices. Pool has a long history of making similar strategic inventory decisions during inflationary periods, the company’s preferred vendor programme and early-buy arrangements with manufacturers are specifically designed for this kind of counter-cyclical purchasing.

The risk case is equally clear. If the price increases management anticipated do not materialise, if tariff impacts are reversed, if inflation moderates, or if demand remains soft and inventory sits longer than planned, then Pool is holding a larger stock of product than it needs, at cost prices that may be above what it can charge in a deflationary environment. The company acknowledges this explicitly in its risk factors: overestimating demand and purchasing too much of a particular product creates the risk that prices fall, leaving inventory that cannot be sold at optimal margins or, in a worst case, requires a write-down.

The chemical and maintenance portion of Pool’s inventory, approximately 14% of net sales comes from pool and hot tub chemicals, adds a specific wrinkle. Pool chemicals such as chlorine products, algaecides, and pH adjusters have finite shelf lives. While Pool manages this risk through its reserve for inventory obsolescence (currently $23.9 million), elevated holding periods increase exposure to degradation, regulatory change, and shifts in consumer preferences for specific chemical formulations. Equipment and building materials carry lower obsolescence risk but are more sensitive to technology cycles and construction activity levels.

To contextualise the scale: the 2023 DIO of approximately 139 days was the decade peak, coinciding with the period when Pool was drawing down the inventory it had aggressively built during the pandemic surge. The 2025 level of 135 days is approaching that level again from a different direction, a deliberate build rather than a demand-driven drawdown. The distinction matters, but the risk of being wrong is real and I carry it explicitly.

Balance Sheet and Equity Trajectory

Pool’s balance sheet has undergone a meaningful structural shift over the past several years that is worth examining carefully.

Total equity grew consistently from $256 million in 2015 through $1.31 billion in 2023, an expansion driven by profitable growth, modest leverage, and retained earnings accumulating on the balance sheet. From 2023 onward, equity has declined, from $1.31 billion to $1.27 billion in 2024 and $1.19 billion in 2025. This reversal is not a sign of financial distress, but it does reflect a clear and deliberate capital allocation choice.

The primary driver is the share buyback programme. Pool repurchased $306 million of shares in both 2023 and 2024, and $341 million in 2025, the largest annual repurchase in company history. These buybacks are funded partly from operating cash flow and partly from incremental borrowing, which is why total debt increased by $249 million in 2025. When a company borrows to buy back its own shares, equity shrinks, buyback proceeds reduce equity directly, while the debt that funded them sits on the liability side of the balance sheet.

Retained earnings tell the same story. Retained earnings peaked at approximately $700 million in 2023 and declined to $521 million in 2025, as the combination of buybacks and dividends paid out more than the net income being retained. Pool paid $185 million in dividends in 2025, marking the 21st consecutive annual dividend increase.

The debt-to-equity (D/E) ratio has risen from 1.00x at year-end 2024 to 1.30x at year-end 2025. This remains within management’s stated target leverage range of 1.5x to 2.0x on a debt/EBITDA basis (Pool’s preferred internal leverage metric). For context, the D/E ratio reached 2.98x in 2018 at a prior leverage peak, and the business remained healthy throughout. The current level is moderate and well within the boundaries of Pool’s financing capacity.

The question this raises for a long-term investor is whether aggressive buybacks at current prices represent good capital allocation. At $225 and approximately 15 years of embedded cash flows, buying back shares at attractive prices is a reasonable use of capital. Management has consistently demonstrated that they would rather deploy capital to buybacks than hold excess cash, and the per-share compounding effect of a declining share count is real regardless of entry price.

4. Competition

A Fragmented Market with One Dominant Player

The competitive structure of US pool wholesale distribution is unlike most industries. At national scale, Pool Corporation is effectively the only fully-formed competitor. The remaining approximately 60% of the US market is distributed among regional and local distributors who lack Pool’s purchasing leverage, product breadth, and logistics infrastructure. This is not a duopoly or a tight oligopoly, it is one dominant national platform and several hundred smaller regional players.

Pool holds approximately 40% of the US wholesale pool distribution market. Management’s 2025 annual report notes that the company believes its selection of pool products is the most comprehensive in the industry, with more than 700 product lines across approximately 40 product categories. That breadth, combined with geographic density that allows same-day delivery in virtually every major pool market, is a service proposition that smaller distributors simply cannot replicate at comparable cost.

The competitive gap has widened over time. While specific current data on competitor facility counts versus Pool’s is not available from the 2025 annual report, the long-term trajectory is clear from the financial record: Pool has been consistently gaining share through network expansion at a pace that smaller competitors cannot match, and the structural advantages of scale, purchasing leverage, product breadth, POOL360 digital integration, compound each year.

The Amazon Question

The e-commerce and direct distribution question is the one I hold with genuine uncertainty over a long horizon. Amazon already sells pool chemicals and basic equipment directly to consumers and some contractors. The question is whether direct-from-manufacturer or Amazon purchasing becomes cost-effective enough for contractor workflows to displace Pool as the primary source.

Pool’s defence has three components, and the evidence on each is currently favourable. Same-day local delivery is an advantage that Amazon cannot replicate without the same physical infrastructure investment Pool has made over four decades. Product breadth of 200,000 SKUs means a contractor relationship involves hundreds of different items across dozens of categories, the convenience premium of a single trusted source is real. And POOL360 integration at 15% of total sales and rising means Pool is building its own digital capability before the competitive challenge fully develops.

None of this permanently closes the question. It means the defence is well-constructed, the early evidence is encouraging, and the competitive outcome over the next decade looks meaningfully better than the disruption narrative implies. I monitor it rather than dismiss it.

The Leslie’s Contrast

Leslie’s (LESL), the publicly traded specialty pool supply retailer, is sometimes cited as a peer. It is not a direct competitor. Leslie’s is a consumer-facing retail operation, not a wholesale distributor, and its financial trajectory has been one of significant distress in recent years, with deep losses and balance sheet stress. The contrast with Pool’s wholesale distribution model is instructive: serving professional contractors with a more defensible service proposition, at higher volumes, with better economics, is structurally superior to consumer retail in this industry.

5. Management

Peter Arvan became President and CEO of Pool Corporation at the end of 2018, succeeding Manuel Perez de la Mesa, who had led the company for nearly two decades. Arvan brought over 20 years of distribution industry experience to the role, which matters more here than it would in most industries. Pool’s business model is operationally intensive, managing 456 sales centres, 2,200 supplier relationships, a 200,000-SKU inventory position, and a contractor customer base that values service consistency above all else. A CEO who understands distribution economics at the operational level is better equipped to manage this complexity than a generalist executive would be.

The tenure since 2019 coincides with a period of significant operational challenge: the pandemic demand surge, the subsequent normalisation, three years of construction suppression, and a deliberate technology platform build-out in POOL360. Through this cycle, the structural economics of the business have remained intact, gross margins held, ROIC stayed above the cost of capital at the trough, and market share was not lost. That is not a given outcome; it reflects operational execution.

Arvan holds approximately 86,400 shares, worth roughly $19 million at current prices. That represents approximately 0.23% of shares outstanding, a meaningful personal position for an individual, though not large enough to constitute a controlling interest or to align management and shareholder incentives in the way founder ownership does. It is the kind of holding that keeps a CEO focused on share price over a five-year horizon; it is not the kind that makes him a co-owner in spirit.

In December 2025, Arvan sold a portion of his holdings (approximately 11,000 shares) in a single transaction. I note this honestly rather than dismissing it. A single sale during a period when the stock was trading below its 52-week high, by a CEO who still holds $19 million in company shares, does not strike me as a meaningful negative signal. But it is a data point worth tracking. A pattern of consistent net selling would change my view; one transaction does not.

On the board, Manuel Perez de la Mesa, Arvan’s predecessor as CEO and the architect of Pool’s modern business model, remains a director and holds approximately 97,500 shares. His continued involvement at board level and his personal economic stake in the business is a meaningful signal: the person who built the flywheel believes in its continued value. Total insider ownership across all directors and officers stands at approximately 388,000 shares, or 1.04% of shares outstanding. This is not a founder-led business in terms of insider concentration, but neither is it a company where management has no economic skin in the outcome.

Compensation Structure

Arvan’s total compensation for FY2025 was approximately $5.3 million, below the median of the peer group Pool uses for benchmarking. Base salary is $900,000, roughly 17% of total compensation, meaning 83% of pay is variable and tied to performance outcomes. I consider this an appropriate structure. A CEO whose pay is predominantly fixed has limited downside exposure to poor decisions; a CEO whose pay is predominantly variable is aligned with shareholders in a meaningful way.

The long-term equity component, the largest slice of total pay, is split equally between time-based restricted stock (three-year cliff vesting) and performance-based restricted stock tied to diluted EPS targets. The performance awards pay out on a 0% to 150% scale depending on whether EPS targets are met, providing upside for outperformance and meaningful forfeiture risk if results disappoint. EPS as the performance metric is well-chosen for Pool: it captures the combined effect of earnings growth and the share buyback programme, both of which are within management’s control and directly relevant to shareholder value creation. The overall compensation structure is judged appropriate.

6. Growth Levers & Addressable Market

Four Structural Drivers

1. Construction Recovery

New pool installations fell from pandemic peaks of over 100,000 units annually to just below 60,000 units in 2025. The suppression is entirely financial: pools are primarily financed through home equity borrowing, and when mortgage rates rose sharply from 2022 onwards, homeowners deferred the decision. This is not a permanent structural decline in demand for pools, it is a rate-driven deferral.

The important nuance is that management’s 2026 guidance explicitly assumes new construction units consistent with 2025. Pool is not projecting a recovery; they are projecting another trough year. That conservatism is appropriate given current rate levels, but it also means the construction recovery is not priced into the business at current earnings levels. When rates normalise and construction recovers toward the historical average of 80,000 to 85,000 units annually, Pool captures that revenue incrementally, on top of a maintenance base that is growing regardless.

Each new pool is also not a one-time event for Pool. It joins the installed base permanently and begins generating recurring annual maintenance demand for 20 to 30 years. A recovery in construction from 60,000 to 80,000 units adds not just current-year construction revenue but a compounding permanent addition to the non-discretionary revenue base.

2. The Technology Upgrade Cycle

The installed base of US pools is ageing. Pools installed in the 1990s and early 2000s are now 25 to 35 years old and require equipment replacement. The technology gap between old and new equipment is substantial. Variable speed pumps use 75% to 80% less electricity than single-speed equivalents. LED pool lighting uses 70% less energy than incandescent. Automated pool controls, allowing smartphone management of pump schedules, heating, lighting, and chemical dosing, are now standard on new installations and desired retrofits on older ones.

California mandated variable speed pumps statewide from 2025. Pool operates 77 locations in California, making it the single largest state in Pool’s network. Other states historically follow California’s energy efficiency leadership with a multi-year lag. This mandate creates a legally-enforced replacement cycle across Pool’s largest market, with similar mandates likely spreading. The upgrade cycle is broad-based, multi-year, and structural, it does not depend on consumer confidence, interest rates, or new construction activity.

3. Sunbelt Demographics

Florida, Texas, California, and Arizona account for approximately 53% of Pool’s net sales. These are also the states receiving the largest net domestic migration inflows. Pool’s network concentration in these markets is not accidental, it is the result of decades of market-by-market expansion following population density. The ongoing migration toward warmer climates creates pool installation demand and maintenance volume growth that accrues directly to Pool’s most concentrated network regions without requiring any strategic repositioning.

4. Fragmented Market Consolidation

Pool adds 8 to 12 new sales centres annually. Each new centre is positioned to serve a local market more effectively than smaller competitors, absorbing volume through better service and pricing. In 2025, Pool added 8 new greenfield centres and acquired 3, for a net addition of 8 after 3 closures. At this pace, Pool’s structural lead over regional competitors widens year after year, and the flywheel described in the Moat section turns one more cycle.

Adjacent Opportunity

Pool’s expansion into irrigation, landscape maintenance, hardscapes, and outdoor living products through Horizon and NPT is early-stage. I treat it as optionality rather than a core driver. The Horizon 24/7 B2B platform and the NPT consumer showroom network are genuine value additions, but they do not yet move the needle on overall earnings in a material way. The potential is there; the execution record in these adjacencies is still developing.

7. valuation

What Today’s Price Assumes

My valuation framework expresses intrinsic value as years of embedded discounted cash flows. Rather than using a terminal value, which requires assumptions about perpetuity growth rates I find too speculative to be reliable. I model an explicit series of annual free cash flows per share, discount each one back to the present at an appropriate rate, and ask: how many years of future cash flows does today’s price already contain?

The framework has five zones. Exceptionally Attractive sits below 15 years. Attractive runs from 15 to 20 years. Hold covers 20 to 30 years. Expensive runs from 30 to 35 years. Exceptionally Expensive is anything above 35 years. I initiate new positions only in the Attractive or Exceptionally Attractive zones

The single most important decision in valuing Pool is choosing the right starting point for free cash flow (FCF) per share. Current FCF of approximately $8 to $9 per share is not the right number, it reflects a construction environment at roughly half of normalised levels, working capital timing distortions from the 2025 inventory build, and the deferred tax payment described in the financial section. Using current FCF as the base would be like pricing a hotel chain on occupancy rates during a recession.

The framework I apply here is the same used in prior Bearhold reports where current FCF understates normalised earning power: I use a normalised FCF per share estimate with an explicit explanation of why current figures are distorted. My base case is $15 per share, representing mid-cycle conditions, construction volumes recovering to approximately 80,000 to 85,000 units annually, normal renovation activity, maintenance base growth continuing at approximately 1% to 2% per year, and stock-based compensation (SBC) deducted from operating cash flow as a real economic cost to shareholders. The upper end of my range is approximately $17 per share, anchoring to 2024 OCF of $659 million as a more representative year for underlying cash generation capability.

I present the range rather than a single number because the honest answer is that normalised FCF has genuine uncertainty. If mid-cycle earning power settles at $13 rather than $15, because cost inflation proves stickier, or because the construction recovery is shallower than expected, the valuation picture is less attractive. I carry that uncertainty explicitly.

The Result

At approximately $225, my model puts Pool at around 15 years of embedded discounted cash flows, the lower boundary of the Attractive zone, approaching Exceptionally Attractive territory. I hold a position initiated before this report’s publication.

The 2026 guidance from Pool’s management, projecting EPS of $10.85 to $11.15 with no construction recovery assumed, confirms that the market is pricing the business at close to trough earnings. That is a different and more interesting situation than being priced at trough earnings with a trough valuation multiple. At 15 years in the Attractive zone, Pool’s current valuation is attractive, especially considering we are at a cyclical low point for earnings in such a high-quality business

Growth Engines

I evaluate the return potential of any business through two engines running simultaneously.

The first engine is FCF per share growth, the fundamental driver of intrinsic value over time. For Pool, this engine has multiple components: the construction recovery adds to the revenue base as rates normalise; operating leverage on a cost structure maintained through the downturn amplifies earnings as revenue recovers; and a consistent share buyback programme reduces the denominator. Pool repurchased $341 million of its own shares in 2025 at prices well below any reasonable mid-cycle intrinsic value estimate. That per-share compounding is real and operates independently of the business cycle.

The second engine is valuation re-rating. At 15 years at the lower boundary of Attractive, this engine is beginning to work in the investor’s favour. An investor adding at current prices captures the construction recovery through fundamental earnings growth and the subsequent re-rating from attractive territory simultaneously. That combination, fundamental compounding plus valuation correction, is what the framework is designed to identify.

8. Risks

Where the Thesis Can Be Wrong

1. Cyclical Earnings Compression

This is the most important risk, and the one most often understated. EPS declined from a peak of $18.70 in 2022 to $10.85 in 2025, a 42% compression over three years, driven by the normalisation of pandemic-era construction demand and elevated financing costs. This will happen again. In the next significant economic downturn, Pool’s discretionary revenue segment will compress. Construction will slow. Renovations will be deferred.

A long-term investor can look through cycles if the business is structurally intact, and the evidence strongly suggests it is. But the volatility is real and must be understood before entering a position. Pool is more cyclical than a pure-maintenance business would be, precisely because 36% of revenues are discretionary. The cyclicality is a feature, not a bug, of a business that participates fully in the upside of construction booms. But it means the entry price matters considerably, and owning Pool at the wrong price through a downturn is a genuinely uncomfortable experience.

2. Inventory Pile-Up

Inventory days of 135 in 2025 are approaching decade-high levels. Management’s rationale, purchasing ahead of anticipated price increases, is a standard and historically successful practice for Pool. But the downside scenario is real: if anticipated price increases do not materialise, or if demand remains soft, Pool is carrying more inventory than it needs at cost prices that may be above what the market will bear. This could result in either margin compression on the sale of elevated-cost inventory, or in a worst case, write-downs. Pool’s reserve for inventory obsolescence was $23.9 million at year-end 2025, not alarming in absolute terms, but worth monitoring against actual write-off trends in coming quarters.

3. E-Commerce Structural Challenge

Amazon and direct-from-manufacturer purchasing represent a long-term structural question for wholesale distribution economics. Pool’s defence, same-day local delivery, 200,000 SKU breadth, POOL360 digital integration, is well-constructed and already working. But the question is not closed over a 20-year horizon. I monitor POOL360 adoption rates, gross margin trends, and any signals of customer defection as the most important leading indicators of this risk.

4. Tariff Exposure

Pool sources products from manufacturers globally, including Chinese-manufactured pool equipment. The current tariff environment introduces cost uncertainty that Pool largely passes through to contractors, but the timing mismatch between cost increases and price realisation creates quarterly earnings volatility. Inflationary product cost increases included a 1% tariff impact in 2025. The tariff environment remains dynamic and will continue to create noise in quarterly results.

5. Housing Market Dependency

New pool construction correlates strongly with housing market activity, home equity availability, and consumer willingness to take on debt for home improvement. In a scenario where housing prices decline significantly, from a broader recession or from housing-specific stress, the construction recovery thesis is delayed or reversed. Consumers who have seen home equity erode are unlikely to finance pool installations. This is the scenario where my normalised FCF estimate of $15 proves too optimistic in the medium term.

6. Normalised Earnings Uncertainty

The honest acknowledgment is that $15 per share as normalised FCF is an estimate, not a number readable from the financial statements. If structural operating cost inflation proves stickier than expected, or if the construction recovery is shallower or slower than my model assumes, the normalised earning power may settle lower. That is a risk I hold explicitly rather than papering over with a point estimate.

What Would Change My Mind

Persistent gross margin compression below 28% for two or more consecutive years, not explained by a temporary mix shift, would signal something more structural than a cyclical trough. A sustained reversal in POOL360 adoption rates. A major supplier announcing direct distribution capability at scale. EPS on a normalised basis, after rates decline meaningfully, failing to recover toward $14 to $15. Any of these would warrant a fundamental reassessment of the thesis.

9. The Verdict

Pool Corporation is Approved in the Bearhold Universe. The business has demonstrated over four decades that its scale advantages compound, its competitive position strengthens through economic cycles, and its maintenance revenue base is structurally permanent. The financial track record, ROIC consistently above 16% through a 42% EPS compression from peak to trough, gross margins stable at 29% to 30% throughout, is empirical evidence of a genuine and durable moat. The cyclicality of the business does not disqualify it from the Approved designation; what matters is that the moat itself has not weakened during the downturn.

At approximately $225 and 15 years of embedded cash flows, Pool sits at the lower boundary of the Attractive zone. I hold a position initiated prior to this publication.

What makes the current moment analytically interesting is this: Pool’s management is guiding for 2026 EPS in line with 2025 and explicitly assumes no recovery in new pool construction. The market is therefore pricing the business at trough earnings, with no meaningful premium for a recovery that, at some point, is structurally inevitable. The construction suppression is rate-driven and financial in nature, not a permanent structural loss of demand for pools. When rates normalise and construction recovers, the earnings recovery is expected to be significant, and investors who entered in the Attractive zone will capture both the fundamental compounding and the re-rating from trough levels.

The risks are real, cyclical earnings volatility, the inventory build, the long-term e-commerce question. None of them, assessed honestly against the quality of what has been built over 40 years and the price at which it is currently available, changes my view that Pool Corporation belongs in the Bearhold Universe.

Disclosure: The author holds a position in POOL. This report reflects the author’s personal views and is not an investment advice. Investing carries the risk of permanent capital loss. Read the full disclaimer here

Sources

Pool Corporation FY2025 Annual Report on Form 10-K (filed February 2026).

Pool Corporation Q4 2025 Earnings Presentation (February 19, 2026).

Pool Corporation company history at poolcorp.com.

P.K. Data, Inc., US in-ground pool installation data.

The Outlook

There is a hormone produced in the human gut after you eat. It signals to your brain that you have had enough. It slows the stomach from emptying. It tells the pancreas to release insulin. For most of medical history, this hormone, glucagon-like peptide-1, or GLP-1, was just a footnote in endocrinology textbooks. Then a Danish pharmaceutical company spent thirty years studying it, synthesising molecules that mimic it, extending their half-life in the bloodstream from a few minutes to a full week, and iterating on the formulation until they could put it in a once-weekly injection. And then, remarkably, into a pill.

What Novo Nordisk has built is not just a drug franchise. It is a century-old institution that made the right scientific bet at the right moment in history, and found itself at the centre of what may be the most commercially significant pharmacological development of our generation. GLP-1 drugs are not a weight loss trend. They are a reclassification of obesity from a lifestyle problem to a treatable chronic disease, and Novo Nordisk is the company that did the reclassifying. My valuation framework puts the stock at 17 years of embedded discounted cash flows, firmly in the Attractive zone.

The business is on the Bearhold Watchlist, not yet Approved, and this report explains the distinction.

At a Glance

Company: Novo Nordisk A/S

Ticker: NVO · NYSE / NOVO B · Nasdaq Copenhagen

Sector: Healthcare

Industry: Pharmaceuticals

Market Cap: ~$168 billion

Dividend Yield: ~4.8% (price $37.5)

Current Stock Price: $37.50

First Coverage: April 2026

Bearhold Universe Status: Watchlist

Valuation Zone: Attractive (last updated April 2026)

This report reflects the author's personal views and does not constitute investment advice. Investing carries the risk of permanent capital loss. The author held a position in NVO during the research process and exited prior to publication. No position is held at the time of publishing. Read the full disclaimer here

1. The Business

From Insulin to the Drug That Changed Everything

The story of Novo Nordisk starts in 1923, the same year insulin was first commercially produced, just two years after its discovery in Canada. A group of Danish scientists recognised that this life-saving hormone would need to be manufactured reliably, at scale, for the millions of people with type 1 diabetes who would otherwise die without it. They built a small laboratory in Copenhagen and began producing insulin from pig and cow pancreas. That business became Novo Nordisk.

For most of the 20th century, the company’s identity was simple: it made insulin. Its early products were animal-derived insulins, extracted directly from livestock pancreas. These worked, but they were not identical to human insulin, which meant the body sometimes rejected them. In the 1980s, Novo Nordisk was among the first companies to produce human insulin through recombinant DNA technology, meaning scientists could engineer bacteria to produce an exact replica of the human molecule. This was a step-change in treatment quality and marked the company’s transition from a manufacturer to an innovator.

Through the 1990s and 2000s, the company developed long-acting insulins, versions that are released slowly into the bloodstream over 24 hours, eliminating the need for multiple daily injections. Tresiba (insulin degludec), one of Novo Nordisk’s flagship insulins, is a once-daily injection that lasts over 40 hours, offering far more flexibility than older formulations. These innovations matter because the management of type 2 diabetes, which affects roughly 500 million people globally, traditionally relied entirely on insulin injections. The more convenient and precise those injections became, the more patients could comply with treatment, and the better their outcomes.

The GLP-1 Discovery

In the 1980s, researchers discovered that the intestinal hormone GLP-1 had remarkable properties: it stimulated insulin release only when blood sugar was elevated, suppressed glucagon (the hormone that raises blood sugar), and slowed digestion. Unlike insulin, which works regardless of blood sugar levels and can cause dangerous hypoglycaemia (dangerously low blood sugar) if dosed incorrectly, GLP-1 had a built-in safety mechanism. It only worked when it was needed.

The problem was that natural GLP-1 is destroyed in the bloodstream within a few minutes. To be useful as a drug, scientists needed to create a version that lasted long enough to have a therapeutic effect. This is where Novo Nordisk’s decades of molecular engineering experience paid off. Their scientists modified the GLP-1 molecule to make it resistant to the enzyme that breaks it down, and bound it to a carrier protein to extend its life in the body. The result was semaglutide, a once-weekly injectable that mimics GLP-1 with dramatically extended duration.

The Same Drug, Different Names

This is one of the most frequently misunderstood aspects of Novo Nordisk’s business, so I want to explain it clearly.

Ozempic and Wegovy contain the exact same active ingredient: semaglutide. The difference is the dose. Ozempic, approved for type 2 diabetes, is available at doses of 0.5mg and 1mg per week. Wegovy, approved for obesity, is dosed at 2.4mg per week. Rybelsus is the oral version of semaglutide, available in daily pill form at doses of 7mg and 14mg, approved for diabetes. The Wegovy pill, approved by the FDA (US Food and Drug Administration) in January 2026, is an oral version for obesity at 25mg per day.

The reason these identical molecules have different names is both regulatory and commercial. Regulatory agencies approve drugs for specific indications at specific doses, a company cannot simply sell its diabetes drug off-label for obesity without a separate approval process. More importantly, it is commercial. Insurance companies in the US have historically refused to cover obesity as a disease, meaning they will reimburse Ozempic for a diabetic patient but not Wegovy for an obese patient who does not have diabetes, even though the molecule is the same. By separating the brands, Novo Nordisk can price and position each product for its specific reimbursement channel. Ozempic’s lower dose and diabetes indication gives it a clearer path to insurance coverage. Wegovy’s obesity approval opens a different, harder, but rapidly expanding reimbursement channel.

Why GLP-1 Is Replacing Traditional Insulin Therapy

The conventional treatment pathway for type 2 diabetes used to look like this: start with oral medication (metformin), add more oral medications as the disease progresses, and eventually transition to daily insulin injections as the pancreas loses its ability to produce enough insulin on its own. Patients often spent years on multiple medications and ended up injecting insulin multiple times daily, managing complex carbohydrate intake, and still experiencing high rates of cardiovascular disease, kidney disease, and nerve damage.

GLP-1 drugs changed this trajectory. Clinical trial after clinical trial showed that semaglutide not only controlled blood sugar as effectively as insulin but also reduced body weight, lowered blood pressure, reduced inflammation, and, critically, reduced cardiovascular mortality. The SELECT trial, completed in 2023, showed that Wegovy reduced major adverse cardiac events such as heart attack and stroke by 20% in people with obesity and established cardiovascular disease. The FLOW trial showed Ozempic reduced the progression of chronic kidney disease by 24%. These are outcomes that insulin, despite decades of use, has never convincingly demonstrated.

The result is that GLP-1 drugs are no longer just an add-on to diabetes therapy. They are increasingly the first-line treatment, with insulin reserved for patients who cannot tolerate GLP-1 drugs or whose disease has progressed beyond what GLP-1 can manage. This is why Novo Nordisk’s older insulin products, Victoza, Levemir, are declining. Victoza sales fell 43% in constant currency terms in FY2025. These are not accidents. They are the natural consequence of a superior drug replacing its predecessor.

How a Diabetes Drug Became the Defining Weight Loss Treatment

The weight loss effect of GLP-1 drugs was noticed early in clinical trials but treated initially as a side effect. Patients taking these drugs for diabetes were losing significant body weight, not through stimulant effects, but because the brain was receiving genuine satiety signals, reducing appetite naturally. Novo Nordisk’s scientists understood what this meant: there was a separate, enormous market here.

The STEP trials, completed in 2021, tested semaglutide 2.4mg specifically for obesity in people who did not have diabetes. Participants lost an average of 14.9% of their body weight. That may sound modest, but in the context of obesity pharmacology, where most approved drugs had achieved 3–5% weight loss at best, it was extraordinary. The FDA approved Wegovy in June 2021, and the obesity treatment market was never the same.

Today the business operates across three segments. Obesity care generated DKK 82.3 billion in FY2025, up from essentially zero in 2019, when it barely existed as a category, growing 31% at CER (constant exchange rates). Diabetes care generated DKK 207.1 billion, growing 4% at CER. Rare disease, treatments for growth hormone disorders and haemophilia, generated DKK 19.6 billion, growing 9% at CER. Total revenue reached DKK 309 billion in FY2025, up from DKK 157.5 billion five years earlier.

2. The Moat

What Makes This Business Hard to Compete with

Novo Nordisk’s competitive position rests on four reinforcing pillars, and understanding each one is important before looking at the financial results.

The first is molecular depth. Semaglutide has been in development since the early 2000s and has accumulated a clinical evidence base that no competitor can replicate quickly. Cardiovascular outcomes trials take five to seven years to complete and cost hundreds of millions of dollars. Novo Nordisk has completed the SELECT trial for obesity, the SUSTAIN-6 trial for diabetes, and the FLOW trial for kidney disease. These are not marketing claims, they are data sets generated from tens of thousands of patients over years. A new entrant today, even with a technically equivalent molecule, would need to spend a decade and billions of dollars to generate comparable evidence. That is structural protection.

The second is manufacturing expertise. GLP-1 drugs are biologics, they are not simple chemical compounds. They are produced through complex biological processes that require precision at every step, from fermentation to purification to formulation for injection or oral delivery. The development of a stable oral semaglutide formulation, which required solving the problem of getting a large peptide molecule through the stomach without being destroyed, is a technical achievement that took years of R&D investment. The current DKK 60 billion annual capital expenditure is explicitly aimed at expanding this infrastructure. It takes time and expertise that cannot be purchased overnight.

The third is commercial infrastructure. Novo Nordisk has one of the deepest specialist sales forces in pharmaceutical history, built over a century of selling to endocrinologists and diabetologists. These relationships are not purely transactional, prescribers who have spent fifteen years trusting Novo Nordisk’s insulin products were natural early adopters of Ozempic. That trust transfers, and it creates a distribution advantage that is invisible in the financial statements but profoundly real.

The fourth is patient stickiness. GLP-1 drugs require continued use to maintain their effect, body weight returns on discontinuation, and blood sugar control deteriorates. This means a patient who starts on Ozempic or Wegovy is, in practice, a recurring revenue stream. It is not a perfect annuity, discontinuation rates are real and I will discuss them in the Risks section, but the baseline adherence creates a revenue durability that few pharmaceutical franchises can match.

The empirical evidence of the moat is in the margins. The operating margin has ranged from 41.3% to 45.8% across every year from 2015 to 2025, averaging 43.1% over the decade. A business that is losing its competitive position does not sustain margins at that level across ten years of significant disruption, including pricing pressure from biosimilars on older products, a global pandemic, supply constraints, and a complete transformation of the revenue mix.

3. The Paradox at the Heart of the Business

One Company Treating the Same Disease it is Helping Prevent

I want to pause on something that I have not seen discussed clearly enough in most analyses of Novo Nordisk. It is a genuine strategic tension at the heart of the business model, and it deserves honest examination.

Novo Nordisk sells two categories of products. In the diabetes segment, it sells drugs that treat and manage diabetes, products that patients typically use for life, generating stable, recurring revenue. In the obesity segment, it sells drugs that cause significant weight loss, and we now have substantial evidence that sustained weight loss reverses or prevents type 2 diabetes in a meaningful share of patients.

I do not think this resolves neatly. Here is how I think about it.

In the near term, the two businesses are growing simultaneously and serve largely different patient populations. Most Ozempic patients are type 2 diabetics with established disease. Most Wegovy patients are obese without a diabetes diagnosis. The overlap exists but it is not yet the dominant dynamic.

In the medium term, say five to ten years, if GLP-1 drug adoption reaches a meaningful share of the obese population, and if adherence rates are sustained, we should expect a measurable reduction in the incidence of new type 2 diabetes cases. This is good for society and genuinely good for patients. For Novo Nordisk’s diabetes franchise, it means the pool of newly diagnosed type 2 diabetics will eventually shrink. At the same time, the surviving diabetes franchise will treat a patient population with more advanced and complex disease, the patients for whom lifestyle intervention and GLP-1 therapy alone were not enough.

In the long term, the company’s own strategic positioning tells you what management believes: obesity care is the growth engine of the next decade, and diabetes care is the stable cash-generating foundation. The restructuring announced in FY2025, redirecting DKK 8 billion in annualised savings toward obesity and diabetes innovation, confirms this. They are not treating this as a zero-sum game within the portfolio. They are betting that the obesity market is large enough to more than offset any erosion in the diabetes base.

I think this is probably correct, but I hold it with appropriate uncertainty. The honest risk is that if obesity drugs penetrate much faster and at much higher adherence rates than current models project, the diabetes franchise could decline faster than the obesity franchise grows. I do not think this is the base case, the sheer scale of the undiagnosed and untreated obesity population is simply enormous, but it is the kind of second-order risk that deserves a place in any serious analysis of this business.

This unresolved dynamic is one of the reasons the obesity franchise has not yet earned the Approved designation, the long-term interaction between these two segments is genuinely uncertain in ways that a decade of insulin history was not.

4. Financial Performance

A Decade in Numbers

The ten-year financial record of Novo Nordisk is, on almost every metric except one, exceptional.

Revenue grew from $15.7 billion in FY2015 to $48.5 billion in FY2025, a compound annual growth rate of approximately 11.9% over ten years. The growth was not linear. Through 2021, the diabetes franchise expanded steadily. Then the GLP-1 obesity inflection arrived: from 2022 onwards, the company added roughly seven to nine billion dollars of revenue per year. The 2025 figure is more than three times the 2015 base.

Diluted EPS grew from $0.99 in FY2015 to $3.61 in FY2025, a compound annual growth rate of 13.8%. This is the number I trust most as a proxy for earnings power, because it is clean, consistent, and not distorted by the current capex cycle. The share count declined from 5,155 million to 4,448 million over the decade, a consistent buyback programme that reduces the per-share denominator. Dividends per share grew from $0.37 in 2015 to $1.82 in 2025.

The gross margin averaged approximately 83.9% from 2015 through 2025, near the pharmaceutical ceiling, reflecting proprietary manufacturing and durable pricing power. In FY2025, gross margin fell to 81.0%. This is a real and material decline, driven primarily by the enormous surge in cost of goods sold as the company scaled manufacturing rapidly and absorbed restructuring costs related to facility consolidation.

The operating margin has ranged from 41.3% to 45.8% across every year of the decade, averaging 43.1%. FY2025 came in at 41.3%, adjusted for the DKK 8 billion restructuring charge (In September 2025, Novo Nordisk announced a “company-wide transformation” involving approximately 9,000 job cuts, this resulted in a one-off DKK 8.0 billion restructuring cost), underlying operating profit grew 13% at constant exchange rates, and the reported operating margin would have been meaningfully higher. The consistency of margins at this level across ten years of significant business model change is one of the most powerful signals of competitive quality I have seen in this type of analysis.

ROIC (return on invested capital) was 73.4% in FY2015 and declined to 26.8% in FY2025 as the capital base expanded with the manufacturing buildout. The directional decline is expected, the denominator grew faster than the numerator as capital was deployed into assets not yet generating returns. The absolute level of 26.8% is still exceptional for a business of this scale. Eli Lilly’s ROIC reached 33% in FY2025 after a decade of expansion from 10.9% in 2015. The two companies are converging at the high end of the global pharmaceutical industry.

The FCF Story, Why the Headline Number is Not the Full Story

Free cash flow (FCF: cash left after all operating expenses and capital investment) was remarkably stable from 2015 through 2023, growing from $4.7 billion to $10.2 billion. OCF (operating cash flow: cash generated from the business before investment) margins throughout this period ranged from 35% to 47%, and FCF margins ranged from 24% to 36%.

In FY2025, FCF collapsed to $4.5 billion, a margin of 9.4%, compared to a ten-year average of approximately 28%.

The cause is entirely capital expenditure. Capex (spending on factories and equipment) was $935 million in FY2015, representing 17% of OCF. By FY2025, it had reached $14.1 billion, 76% of OCF. The company spent DKK 60 billion on property, plant, and equipment in FY2025, principally to build manufacturing capacity for GLP-1 drugs.

The important distinction is that OCF margins remained within historical norms: 38.5% in FY2025. The business is not generating less cash from its operations, it is reinvesting aggressively in infrastructure. Because FCF per share is temporarily distorted, I use EPS as my primary earnings proxy throughout this report. The operating health of the business is intact.

The Balance Sheet Has Changed

Debt-to-equity was 0.02x in FY2015, essentially no debt. By FY2025, it had risen to 0.68x. The company now carries meaningful net debt, having taken on borrowings to fund both the manufacturing buildout and several acquisitions, including the Catalent manufacturing sites that Novo Holdings purchased in 2024 for $11.7 billion and subsequently deployed to Novo Nordisk’s production. This is manageable at current earnings levels, but it represents a structural shift from the near-pristine balance sheet that characterised this company through most of the last decade.

5. Regional Breakdown

The Numbers by Region (FY2025)

The detailed revenue data from the FY2025 annual report tells a story that the headline figures do not.

US Operations generated DKK 173.2 billion in total sales in FY2025, growing 3.4% as reported (8% at CER). Within that, Wegovy in the US generated DKK 51 billion, up from DKK 45.8 billion in 2024. Ozempic US sales were DKK 88.5 billion. These are extraordinary numbers for two products that barely existed five years ago.

International Operations generated DKK 135.9 billion, growing 10.5% as reported (14% at CER). The breakdown within International Operations reveals where the real opportunity sits:

EUCAN (Europe and Canada) generated DKK 66.1 billion, growing 14.9% as reported (16% at CER). Wegovy in EUCAN generated DKK 15.4 billion, up from DKK 7.7 billion in 2024, a doubling in a single year. This is what the early phase of a proper launch looks like. Europe is roughly two years behind the US in GLP-1 obesity adoption, constrained by more conservative payer systems and slower reimbursement approvals. The trajectory is clear.

Emerging Markets (mainly Latin America, Middle East, and Africa) generated DKK 30.4 billion, growing 3.1% as reported (8% at CER). Wegovy in Emerging Markets generated DKK 6.1 billion in FY2025, up from DKK 2.7 billion in 2024. This is one of the most interesting long-term opportunities and one of the most underdiscussed. Obesity prevalence in Latin America and the Middle East is extremely high, Brazil, Mexico, and Saudi Arabia are among the most obese nations on earth. The barrier is affordability and reimbursement coverage. As Novo Nordisk develops lower-cost formulations and access programmes, this market has a long runway.

APAC (Japan, Korea, Oceania, and Southeast Asia) generated DKK 20.7 billion, growing 18.8% as reported (25% at CER). Wegovy in APAC generated DKK 5.8 billion, up from DKK 1.9 billion in 2024, a tripling in a single year. Japan and South Korea have both launched Wegovy relatively recently, and the cultural and clinical context is different from the West. Japanese patients tend to be obese at lower BMI (body mass index) thresholds than Western patients, and the regulatory framework for obesity treatment has historically been restrictive. As awareness grows and reimbursement expands, APAC represents a meaningful expansion opportunity.

Region China generated DKK 18.7 billion, growing 0.8% as reported (5% at CER). This is the most complicated region, and I will address it directly in the Risks section. Wegovy in China generated DKK 796 million, still a very small number relative to the scale of the opportunity. China has a massive obesity and diabetes burden, but the semaglutide active ingredient patent expires there in 2026, which means low-cost domestic biosimilar competition is coming. The next two to three years in China will be a test of brand loyalty versus price.

The Global Penetration Opportunity

Here is the number I keep coming back to: the World Health Organization estimates approximately 890 million adults globally have obesity. Current GLP-1 treatment penetration is in the low single digits. The United States has an obesity prevalence of roughly 42% in adults, and fewer than 6% of eligible patients are currently on a branded GLP-1 drug. Even after the explosive growth of the last three years, the penetration story is genuinely in its early chapters.

6. Competition: Novo Nordisk vs. Eli Lilly

The Duopoly That Defines the GLP-1 Market

Novo Nordisk and Eli Lilly together represent the vast majority of commercial GLP-1 (glucagon-like peptide-1) prescriptions globally. No other company is remotely close to their combined scale in either diabetes or obesity. Understanding how these two businesses compare is essential.

Novo Nordisk pioneered this market. It was the first to achieve large-scale commercial success with semaglutide, the first to build a manufacturing infrastructure capable of supplying tens of millions of patients globally, and the first to bring an oral GLP-1 pill for obesity to market. That first-mover advantage is real and it matters, both commercially and in terms of the clinical evidence base accumulated across cardiovascular disease, kidney disease, and liver disease.

But Eli Lilly has overtaken Novo Nordisk in prescription market share, and that shift deserves honest acknowledgment. By the third quarter of 2025, Lilly held more than 57% of US monthly GLP-1 prescriptions across diabetes and obesity, with Novo Nordisk at approximately 43%, down from a position of clear leadership just two years earlier. The primary driver is tirzepatide, which activates two hormone receptors simultaneously, GLP-1 and GIP (glucose-dependent insulinotropic polypeptide), compared to semaglutide’s one. In Lilly’s SURMOUNT-1 trial, tirzepatide produced average weight loss of approximately 21–22% at the highest dose, compared to semaglutide’s 15% in the STEP trials. When a drug delivers meaningfully better efficacy and is available in adequate supply, prescribers and patients notice. Lilly’s 2026 revenue guidance projects approximately 27% growth. Novo Nordisk’s 2026 guidance is for negative adjusted sales growth at constant exchange rates. The divergence in near-term momentum is real and not easily dismissed.

Where It Gets More Complicated — The Pill Battle

The injectable competition is clear: Lilly leads on efficacy. But the oral market, which is where the next major wave of GLP-1 adoption is likely to come from, driven by patients who have consistently refused injections, is a more nuanced picture, and one that has shifted meaningfully in Novo Nordisk’s favour in just the past two weeks.

Novo Nordisk’s Wegovy pill was approved by the FDA in December 2025 and reached over 600,000 US prescriptions in its first few weeks. On April 1, 2026, 12 days before this report, the FDA approved Eli Lilly’s oral GLP-1, orforglipron, now branded as Foundayo. The pill competition is now live and direct.

The initial market framing was that Novo had first-mover advantage but Lilly had a convenience edge: Foundayo is a small-molecule drug that can be taken at any time with or without food, while the Wegovy pill is a peptide that requires a 30-minute fast each morning. For patients who already struggle with daily medication adherence, that restriction was seen as a meaningful disadvantage for Novo.

Then, on April 2, 2026, one day after Foundayo’s approval, Novo Nordisk presented the ORION study at the Obesity Medicine Association’s annual conference in San Diego. The study used a population-adjusted indirect comparison of data from the OASIS 4 trial (Wegovy pill) and the ATTAIN-1 trial (Foundayo), and the results were notable. Oral semaglutide showed 3.2 percentage points greater weight loss than orforglipron on a real-world adherence basis. On tolerability, patients on orforglipron had approximately four times higher odds of discontinuing due to any adverse event, and nearly 14 times higher odds of discontinuing specifically due to gastrointestinal side effects. A separate patient preference survey of 800 adults found that 84% favoured the oral semaglutide profile over orforglipron, and 65% of those respondents said the morning fasting requirement would not significantly disrupt their daily routine.

I want to be clear about what this data is and what it is not. The ORION study is an indirect comparison across two separate trials, not a head-to-head study with identical protocols. The researchers themselves flagged substantial uncertainty in the tolerability findings, the confidence interval on the gastrointestinal discontinuation figure runs from 2.0 to 96.0, which is wide enough to counsel humility. The study was also funded and presented by Novo Nordisk, which means it should be read with appropriate critical distance even if the methodology is standard. No direct head-to-head trial between these two pills exists, and neither company is likely to fund one voluntarily.

With those caveats stated, the direction of the finding is meaningful. In a chronic disease drug that patients take daily for the rest of their lives, tolerability and real-world adherence matter more than any single efficacy number. A drug that patients stay on compounds its benefit over years and generates recurring revenue. A drug that patients are significantly more likely to discontinue due to side effects loses both the clinical benefit and the commercial durability. The convenience narrative that Lilly was relying on to offset Novo’s first-mover advantage in the oral market has been meaningfully complicated by this data.

The Margin and Balance Sheet Picture

On financial quality, Novo Nordisk maintains the structural advantage. Its gross margin averaged approximately 83.9% from 2015 through 2024, falling to 81.0% in FY2025, compared to Eli Lilly’s expansion from 74.8% in 2015 to 83.0% in FY2025. Lilly has only just reached the margin level that Novo Nordisk has sustained for a decade. Operating margins tell the same story: Novo Nordisk’s range of 41.3% to 45.8% throughout the decade versus Lilly’s expansion from 18% in 2015 to 45.6% in FY2025. Both companies are now experiencing the same capex-driven FCF (free cash flow) compression, Lilly’s FCF margin was 9.2% in FY2025 versus Novo Nordisk’s 9.4%, as both build manufacturing infrastructure for the same market opportunity. On leverage, Novo Nordisk is more conservatively capitalised at 0.68x debt-to-equity versus Lilly’s 1.60x.

One metric I flag for both companies is inventory. Eli Lilly’s Days Inventory Outstanding, a measure of how long products sit in inventory before being sold, increased from 196 days in FY2021 to 352 days in FY2025, nearly a doubling in four years. This could reflect pre-launch positioning or manufacturing buffer-building, but it is a number worth watching in future quarterly reports for signs of demand forecasting error.

The Summary

The injectable market: Lilly leads on efficacy with tirzepatide and has taken prescription share from Novo Nordisk. That is a fact.

The oral market: Novo Nordisk leads on efficacy and, based on the most current available data, leads on tolerability as well, though the evidence is indirect and requires a head-to-head trial to confirm definitively.

The financial quality: Novo Nordisk has the stronger historical margin profile and more conservative balance sheet.

The pipeline: Novo Nordisk’s CagriSema, filed for FDA approval in December 2025 with a decision expected around October 2026, produced 22.7% weight loss in the REDEFINE 1 trial, essentially matching tirzepatide’s injectable efficacy. If approved, it closes the injectable efficacy gap before Lilly’s next-generation retatrutide (a triple receptor agonist) reaches market.

I would describe the current state as a genuine competition between two exceptional businesses, with Lilly holding the commercial momentum in injectables and Novo Nordisk mounting a more credible defence in the oral market than the market appears to currently price in. The coming 18 months, CagriSema approval, head-to-head oral data if it emerges, and the real-world prescription trends between Wegovy pill and Foundayo, will determine whether Novo Nordisk’s defensive position stabilises or continues to erode. At $37.50, I believe the current price more than compensates for that uncertainty.

7. Growth Levers & Addressable Market

The Pipeline: Novo Nordisk’s Response to the Lilly Challenge

CagriSema is the most important near-term pipeline catalyst. It is a fixed-dose combination of cagrilintide (a long-acting amylin analogue, amylin is another satiety hormone produced by the pancreas) and semaglutide 2.4mg. In the REDEFINE 1 Phase 3 trial, participants lost an average of 22.7% of body weight assuming full adherence to treatment, and 20.4% on a real-world basis regardless of adherence. Both figures substantially exceed Wegovy’s current results. Novo Nordisk filed the NDA (new drug application) with the FDA in December 2025, and a decision is expected approximately October 2026.

Zenagamtide (amycretin), a single molecule that activates both GLP-1 and amylin receptors, is entering Phase 3 trials in 2026 in both injectable and oral forms, with early-phase data showing weight loss in the 20% range. The semaglutide 7.2mg dose achieved 20.7% weight loss and has received a positive opinion from the EMA (European Medicines Agency), with an FDA submission also filed. Wegovy was approved for MASH in the US in FY2025, MASH affects approximately 6% of the global population and has very limited approved treatment options, making this a meaningful new revenue channel.

The oral Wegovy pill, approved by the FDA in December 2025, is worth emphasising specifically. Injection hesitancy is a real and documented barrier to GLP-1 adoption. A meaningful share of patients who would benefit from these drugs decline or discontinue them because of the injection requirement. An oral formulation that achieved 16.6% average weight loss in trials, better than any previously approved oral obesity drug, removes that barrier entirely. This is a market expansion story, not just a market share story.

8. Management

Lars Fruergaard Jørgensen, Mike Doustdar, and a CEO Transition at a Critical Moment

Lars Fruergaard Jørgensen served as CEO from 2017 through August 2025. His tenure included the most transformative period in Novo Nordisk’s modern history, the pivot to obesity care, the launch of Wegovy, the extraordinary growth from 2021 to 2023, and the beginning of the current manufacturing buildout. On May 16, 2025, the company announced he would be stepping down following a period of market challenges and declining share price. On July 29, 2025, Mike Doustdar, then head of International Operations, was named as successor, with the formal handover taking place on August 7, 2025.

I read the transition thoughtfully. Jørgensen built the strategic architecture of the current business. Doustdar is, in many ways, the commercial architect of its execution: as head of International Operations, he oversaw the global Wegovy launch and was responsible for the market access strategy that determined how quickly the drug reached patients outside the US. His appointment is not a reversal of strategy. It is a shift in emphasis, from the scientific and strategic decisions of the build-out phase to the commercial and operational execution required to turn that investment into revenue.

The capital allocation record under the previous leadership is solid. Share count declined by approximately 14% over the decade through consistent buybacks. R&D (research and development) spending reached DKK 52 billion in FY2025, 16.8% of revenue, reflecting a genuine commitment to the pipeline rather than cost-cutting to protect near-term margins. The DKK 8 billion restructuring announced in FY2025, reducing the global workforce by approximately 9,000 positions, signals that management is willing to make uncomfortable structural decisions. I read that positively.

The governance structure deserves a note. Novo Holdings, controlled by the Novo Nordisk Foundation, holds approximately 77% of the voting rights through a dual-class share structure. This insulates management from short-term market pressure. It is a structural positive for a business with a multi-decade strategic horizon, though minority shareholders have limited influence over capital allocation decisions.

9. The Compounding Pharmacy Story

The GLP-1 compounding story is important context for understanding the FY2025 performance, and its resolution is one of the reasons I believe the near-term earnings trajectory is better than the FY2026 guidance implies.

When demand for Wegovy and Ozempic surged in 2022 and Novo Nordisk’s manufacturing could not keep pace, the FDA placed semaglutide on its official drug shortage list. Under US law, compounding pharmacies can produce copies of drugs on the shortage list. What followed was a large parallel market: an estimated 3.7 million Americans accessing compounded semaglutide through telehealth platforms at $150–$400 per month versus the $1,349 list price of Wegovy.

The FDA declared the shortage resolved on February 21, 2025. The legal basis for most large-scale compounding of semaglutide no longer exists. Courts sided with Novo Nordisk and the FDA in the key preliminary injunction hearings. As of September 2025, Novo Nordisk had filed 140 lawsuits and issued over 1,000 cease-and-desist letters against compounders. The management team acknowledged that the persistence of compounded semaglutide was a meaningful drag on branded volumes in 2025. As compounding recedes, that volume either transitions to branded Wegovy or is lost, but the channel overhang is clearing.

The TrumpRx Pricing Deal, Volume for Price

In November 2025, Novo Nordisk reached an agreement with the Trump administration under which Wegovy and Ozempic prices will be reduced to $350 per month through the TrumpRx government portal, down from the $1,349 Wegovy list price. The deal also extends Medicare coverage of Wegovy for obesity for the first time. In exchange, Novo Nordisk received a three-year exemption from the pharmaceutical tariffs the administration announced, tariffs that would otherwise apply at 100% to patented drugs imported without a Most Favored Nation (MFN) pricing agreement. Given that Novo Nordisk manufactures substantially in Denmark and Europe, the tariff exemption through approximately 2028 is valuable insurance during the period when its US manufacturing buildout is still coming online.

Management expects a negative low-single-digit impact on global sales growth in 2026 from the pricing agreement. That is a real near-term headwind. The medium-term logic is that lower prices plus expanded Medicare access could ultimately drive higher volumes that more than offset the per-unit reduction. I believe that logic is sound, but it will take time to play out.

10. Valuation

What the Market is Pricing in, and What I Think it is Missing

My valuation framework for Bearhold Research expresses intrinsic value as a number of years of embedded discounted cash flows. Rather than using a terminal value, which requires assumptions about perpetuity growth rates that I find too speculative to be reliable, I model an explicit series of annual free cash flows, discount each one back to the present, and ask a simple question: how many years of future cash flows does today’s price already contain? The answer tells me whether I am paying a fair price, a cheap price, or an expensive one.

The framework has five zones. Exceptionally Attractive sits below 15 years. Attractive runs from 16 to 20 years. Hold covers 21 to 30 years. Expensive runs from 31 to 35 years. Exceptionally Expensive is anything above 36 years. I initiate new positions only in the Attractive or Exceptionally Attractive zones and add most aggressively when a business I understand well enters Exceptionally Attractive territory. I sell when a position reaches Expensive or Exceptionally Expensive and reallocate to Attractive names in the Bearhold Universe.

17 Years — Attractive Zone

At $37.50 per share, Novo Nordisk sits at around 17 years of embedded cash flows, firmly in the Attractive zone of the Bearhold framework.

The most important thing to understand about this number is what it does and does not assume. It does not assume a flawless recovery. It explicitly accounts for the near-term FCF (free cash flow) compression driven by the capex cycle, modelling the gradual normalisation as the manufacturing buildout matures rather than assuming an immediate return to historical cash generation levels. It applies a growth rate that is a meaningful haircut to the company’s historical FCF per share CAGR, acknowledging competitive pressure from Lilly in the injectable market, gross margin headwinds from the TrumpRx pricing deal, and patent expiries in China. And it uses a discount rate at the conservative end of the reasonable range for a business of this quality, reflecting the genuine operational uncertainty of this particular moment.

In other words, 17 years is not an optimistic number dressed up as a conservative one. It is what the arithmetic produces when you take the near-term headwinds seriously.

At 17 years, the thesis requires the capex cycle to normalise broadly on schedule, revenue to recover from the 2026 transition year, and FCF per share to compound at a rate that reflects the underlying quality of the franchise rather than the distortions of the current investment cycle. If those things happen, and I believe they will, for reasons I have set out throughout this report, then 17 years represents a genuine margin of safety in a business whose quality justifies a much higher price.

Growth Engines

I evaluate the return potential of any business through two engines running simultaneously.

The first engine is FCF per share growth, the fundamental driver of intrinsic value over time. For Novo Nordisk, the near-term FCF figure is distorted by the capex cycle, and as that distortion clears over the next two to three years, FCF per share growth should re-converge with the underlying earnings power of the franchise. The combination of revenue growth, operating leverage, and a consistent share buyback programme, which has reduced the diluted share count from 5,155 million in FY2015 to 4,448 million in FY2025, provides a quiet compounding mechanism that operates in the background of every other assumption in the model. Even modest share count reduction adds to per share growth without requiring any improvement in the absolute cash generation of the business.

The second engine is valuation re-rating. At 17 years in the Attractive zone, this engine is working in the investor’s favour. The real optionality lies in the scenario where execution delivers, CagriSema approved, capex cycle matures, gross margins recover, oral market share consolidates in Novo Nordisk’s favour, and the market re-rates the stock from Attractive back toward the upper end of Hold or beyond. In that scenario, the investor earns the fundamental compounding of FCF per share growth and a valuation multiple expansion simultaneously. That combination is what makes the Attractive zone entry compelling for a long-term investor

11. Risks

Every investment thesis has a version of events where it is wrong. I want to walk through the scenarios that would genuinely change my view on Novo Nordisk, not the boilerplate risks that appear in every pharmaceutical analysis, but the specific dynamics that keep me thinking carefully about this position.

The Gross Margin is the Number I Watch Most Closely

This is not the risk that gets the most attention, but it is the one I consider most important to the long-term thesis. Novo Nordisk’s gross margin fell from a decade average of approximately 83.9% to 81.0% in FY2025. The official explanation is a combination of rapid manufacturing scale-up costs, one-off restructuring charges, and the initial inefficiency of newly commissioned facilities. If that explanation is correct, gross margins normalise as the new capacity reaches full utilisation over the next two to three years, and the underlying earnings power of the business is largely intact.

But there is an alternative explanation that I cannot dismiss. The TrumpRx pricing deal reduced Wegovy and Ozempic prices to $350 per month on government channels. Competitive pressure in the oral market is pushing both Novo Nordisk and Lilly toward $149 starting prices for their pills. Biosimilar competition is coming in China and eventually in Western markets as patents expire. If these pricing pressures are structural rather than temporary, the long-term gross margin of this business may settle permanently lower than its historical range. A business that earns 79% gross margins rather than 84% is still exceptional, but the difference compounds significantly over a decade of growth at this scale. I do not think this is the most likely outcome, but it is a risk I watch carefully.

The Capex Cycle

Novo Nordisk is spending DKK 60 billion per year, approximately $9 billion, on capital expenditure, and has committed approximately USD 5.6 billion in additional US manufacturing investment through 2028. The total committed capital across the current buildout cycle runs into the tens of billions of dollars. This infrastructure is being built on the assumption that GLP-1 demand will continue to grow rapidly for years and that Novo Nordisk will capture a meaningful share of that growth.

The risk is not that demand for GLP-1 drugs disappoints in aggregate. I believe the structural demand case is overwhelming, as I discussed in the market view section. The risk is more specific: that Novo Nordisk’s share of that demand disappoints relative to the assumptions embedded in the capex decisions. If Lilly continues to gain injectable market share at Novo Nordisk’s expense, and if the oral market develops more slowly than projected, then the company will have built manufacturing capacity for volumes it cannot fill. Capital expenditure is largely irreversible. Factories cannot be unbuilt. The financial consequence would be years of elevated depreciation charges on underutilised assets, compressing returns on invested capital precisely when the business needs to demonstrate that the investment cycle is paying off.

I think this risk is manageable but it is not theoretical. Management made their capacity decisions when Novo Nordisk’s growth trajectory looked dramatically more positive than it does today. The FY2025 full-year sales growth of 10.3% at constant exchange rates versus the FY2026 guidance of negative adjusted sales growth represents a sharp deceleration. How much of that deceleration is transitional, compounding headwinds, pricing adjustments, China patent expiry, and how much of it represents a more durable slowdown in the underlying franchise, will determine whether the capex cycle was visionary or premature.

China Patent Expiry

All of Novo Nordisk’s core semaglutide products have their active ingredient patents expiring in China in 2026. Ozempic, Wegovy, Rybelsus, the entire franchise. China generated DKK 18.7 billion in FY2025 sales and was growing at 5% at constant exchange rates. Novo Nordisk’s Wegovy launch in China has barely begun, with only DKK 796 million in FY2025 revenue, meaning the obesity opportunity there is largely untapped at the moment the moat is about to be breached.

The practical impact will not be immediate. Biosimilar manufacturers need time to build commercial scale, achieve regulatory approvals, and establish distribution networks. Novo Nordisk’s brand recognition, safety data, and clinical relationships provide a buffer. But Chinese pharmaceutical companies are sophisticated, well-capitalised, and experienced in bringing biosimilars to market quickly. Several domestic companies were already preparing semaglutide biosimilars well before the patent expiry. The pricing pressure in China over the next two to three years will be significant, and the obesity market, which was supposed to be a major long-term growth driver in the world’s most populous country, will develop in a far more competitive and lower-margin environment than the one that drove the Western growth story.

The Injectable Efficacy Gap

Tirzepatide’s approximately 21% average weight loss versus semaglutide’s 15% is a real clinical difference that is influencing prescribing behaviour. Novo Nordisk held approximately 59.6% of global branded GLP-1 volume market share in FY2025, but the US prescription trend, Lilly at 57% of monthly prescriptions and rising versus Novo at 43% and falling, is the more relevant near-term signal. The direction matters as much as the level.

The honest risk here is timing. CagriSema, which matches tirzepatide’s efficacy at 22.7% weight loss, is filed for FDA approval with a decision expected around October 2026. If the approval is delayed, through an unexpected complete response letter, additional data requests, or manufacturing inspection issues, Novo Nordisk remains in the gap year for longer than projected. Every additional month without CagriSema is another month of injectable market share drifting toward Lilly, another month of compounders and prescribers defaulting to tirzepatide for new obesity patients, and another month of the narrative calcifying around Lilly as the dominant player. The business does not collapse in this scenario, but the re-rating catalyst is deferred and the share count of prescribers who have built tirzepatide habits grows.

Real-world GLP-1 Adherence, The Recurring Revenue Moat May be More Fragile Than it Appears

I described the recurring revenue nature of GLP-1 drugs as a feature of the moat, patients who stay on therapy for life generate predictable, growing cash flows. The critical assumption is that patients actually stay on therapy. The real-world data on this is sobering. First-year discontinuation rates for injectable GLP-1 drugs in real-world settings have been estimated at 40 to 50% in multiple analyses, significantly higher than the dropout rates observed in tightly controlled clinical trials.

The reasons are well-documented: gastrointestinal side effects concentrated in the dose-escalation phase, cost and coverage barriers, weight loss plateaus that disappoint patients expecting linear progress, and the practical difficulty of managing a weekly injection in daily life over years. If a meaningful share of the patients who started on Wegovy or Ozempic in the 2022 to 2024 wave have already discontinued, the installed base of recurring prescriptions is smaller than the volume data implies, and the forward revenue from that cohort is lower than a pure adherence model would project.

This dynamic also has a second-order implication for the capex cycle. If real-world adherence is significantly worse than clinical trial data suggests, the demand projections management used when authorising DKK 60 billion annual capital expenditure may have been built on an assumption that the treated population compounds reliably over time. If the population turns over faster, patients starting, stopping, and restarting, the demand profile becomes more volatile and harder to forecast accurately.

The Diabetes-obesity Paradox, long-term Structural Uncertainty

I discussed this in section 3 as a paradox rather than a risk, but at a multi-decade horizon it becomes one. If GLP-1 obesity drugs achieve the penetration rates the most optimistic projections envision, treating hundreds of millions of people globally over the next twenty years, the downstream effect on type 2 diabetes incidence will be measurable. The patients who avoid diabetes because of sustained GLP-1 treatment are patients who do not eventually need insulin, metformin, and diabetes-specific medications. Novo Nordisk’s diabetes franchise, still generating DKK 207 billion in FY2025, by far the larger segment, will face a structurally shrinking addressable market over a long enough time horizon.

I do not think this plays out as a crisis. The transition will be gradual, the diabetes franchise will continue generating strong cash flows for many years, and the obesity revenue replacing it operates at similarly high margins. But any honest long-range model for this business needs to account for the possibility that the company’s most important product is, over a long enough horizon, cannibalising the market for its second most important product line. The net effect is probably positive, obesity revenue grows faster than diabetes revenue declines, but the uncertainty is real and deserves acknowledgment.

The Verdict

Bearhold Universe Status: Watchlist

The Approved designation in the Bearhold Universe is purely qualitative. It has nothing to do with valuation, price, or near-term earnings visibility. It asks one question: has this business demonstrated, through a sufficient track record, that it can sustain excellence in its competitive arena the way the best businesses in history have? The answer determines the designation. The price determines when I act.

Novo Nordisk’s insulin franchise answers that question without hesitation, and I want to be clear about why. This is not a business that stumbled into a good decade. It is a business that built a durable competitive position over a century, through two world wars, through the transition from animal insulin to recombinant DNA technology, through the commoditisation of older molecules, through the arrival of new drug classes that threatened its core. In every one of those transitions, Novo Nordisk did not just survive. It adapted, invested, and emerged with a stronger position than it entered with. The operating margin holding between 41% and 46% across every single year from 2015 to 2025, through a global pandemic, a complete revenue mix transformation, and a manufacturing buildout of historic proportions, is the financial expression of a century of that institutional resilience. The insulin franchise is Approved, unambiguously and permanently.

The obesity franchise is where I have to be honest about what I know and what I do not yet know. And the distinction matters enormously to me, because the Approved designation is a statement about demonstrated quality, not about the quality I believe is coming.

What I know is this. Semaglutide’s clinical evidence base is exceptional. The SELECT trial reducing major adverse cardiac events by 20% in people with established cardiovascular disease. The FLOW trial reducing chronic kidney disease progression by 24%. The STEP trials producing average weight loss of 14.9%, extraordinary by any historical standard in obesity pharmacology. These are not marginal results. They are practice-changing outcomes that have permanently altered clinical guidelines across cardiology, nephrology, and obesity medicine simultaneously. The commercial execution behind these results has been equally impressive. DKK 82 billion in obesity care revenue in FY2025 from essentially nothing six years ago, with a global manufacturing buildout that represents the largest capital commitment in the company’s history. Management made the right strategic bet, made it early, and executed it at scale.

What I do not yet know is whether Novo Nordisk will dominate the obesity market the way it dominated the insulin market. And that distinction is exactly where the Watchlist designation lives.

Eli Lilly’s tirzepatide currently demonstrates superior injectable efficacy, approximately 21% average weight loss versus semaglutide’s 15% in their respective pivotal trials. That gap is real and it is influencing prescribing behaviour in measurable ways. By the third quarter of 2025, Lilly held more than 57% of US monthly GLP-1 prescriptions, having overtaken Novo Nordisk from a position of no market presence just three years earlier. That is a competitive trajectory that deserves honest acknowledgment. Novo Nordisk pioneered this market and is currently being challenged within it by a competitor with a better efficacy number in the most commercially important indication.

CagriSema is Novo Nordisk’s answer. Filed with the FDA in December 2025 with a decision expected around October 2026, it produced 22.7% weight loss in the REDEFINE 1 trial, essentially matching tirzepatide’s headline figure and doing so with a novel dual-mechanism approach combining semaglutide with a long-acting amylin analogue. If CagriSema is approved and demonstrates competitive real-world efficacy, it closes the injectable gap before Lilly’s next-generation retatrutide reaches market. The pipeline response is credible. But it is a promise, not yet a result. And the Approved designation requires results.

The oral market is equally unresolved. On April 1, 2026, twelve days before this report was published, the FDA approved Eli Lilly’s orforglipron, now branded Foundayo, as the second oral GLP-1 pill for obesity. The ORION indirect comparison data, presented the following day, suggests the Wegovy pill has meaningful advantages in both efficacy and tolerability. Patients on orforglipron showed approximately four times higher odds of discontinuing due to adverse events and nearly fourteen times higher odds of discontinuing specifically due to gastrointestinal side effects. These are striking numbers and they support a compelling narrative for the Wegovy pill’s commercial position. But this is an indirect comparison across separate trials, funded and presented by Novo Nordisk, with wide confidence intervals and no head-to-head trial to settle the question definitively. The oral market battle between these two drugs will play out in real prescriptions over the next twelve to eighteen months, and those real-world results are what the quality assessment requires, not an indirect comparison published the day after a competitor’s approval.

This is what the Watchlist is designed to capture. The insulin franchise is proven. The obesity franchise is promising, credibly positioned, and backed by a pipeline that could decisively establish Novo Nordisk’s leadership in the new competitive arena. But the competitive outcome has not yet been determined. The prescription share trajectory, the CagriSema approval and launch, and the oral market real-world data will together tell the story that the insulin franchise told over decades, except compressed into the next two to three years because the competitive intensity demands it.

The specific triggers that would move Novo Nordisk from Watchlist to Approved are these. CagriSema receiving FDA approval and demonstrating competitive or superior real-world efficacy versus tirzepatide in its first year of commercial prescription data. And the oral market prescription trends confirming over at least two to three quarters that the Wegovy pill’s tolerability advantage holds in actual patient behaviour, that patients are staying on it longer and discontinuing less than Foundayo in the real world, not just in an indirect trial comparison. When those two conditions are met, the obesity franchise will have earned the track record the designation requires. It will have demonstrated that Novo Nordisk can do in obesity what it did in insulin, build a leadership position and defend it against serious competition through product quality, not just first-mover advantage.

At $37.50 and 17 years of embedded cash flows, the price sits in the Attractive zone of the Bearhold valuation framework. The business I have described in this report is exceptional. The near-term headwinds, negative 2026 guidance, China patent expiry, the capex cycle compressing free cash flow, the CEO transition, are all real and all visible in the numbers. None of them concern me as a long-term investor. What keeps Novo Nordisk on the Watchlist rather than in the Approved column is not weakness. It is the honest acknowledgment that the most important competitive chapter of this company’s modern history is being written right now, in real time, with the outcome still genuinely uncertain. I will be watching that chapter closely. And when the evidence confirms what the insulin franchise already demonstrated about this company’s ability to build and defend category leadership, the designation will change.

This report reflects the author’s personal views and does not constitute investment advice. Investing carries the risk of permanent capital loss. The author held a position in NVO during the research process and exited prior to publication. No position is held at the time of publishing. Read the full disclaimer here

Sources:

Novo Nordisk Annual Report FY2025 (DKK)

Novo Nordisk Annual Report FY2024;

FDA Declaratory Order on semaglutide shortage, February 21, 2025;

Novo Nordisk NDA filing for CagriSema, December 18, 2025; REDEFINE 1 and REDEFINE 2 Phase 3 trial data;

SELECT cardiovascular outcomes trial;

FLOW kidney disease trial;

ORION indirect treatment comparison, Obesity Medicine Association 2026, April 10–12, San Diego;

OPTIC patient preference study, Novo Nordisk, October–November 2025;

FDA approval of Foundayo (orforglipron), April 1, 2026;

White House TrumpRx / MFN pricing announcement, November 2025;

Trump Administration pharmaceutical tariff Executive Order, April 2026;

Morningstar GLP-1 market analysis, December 2025;

J.P. Morgan GLP-1 market projections, 2026.

There is a line in Accenture’s FY2025 annual report that I keep returning to. It appears in the section about AI adoption among enterprise clients, and it reads with the kind of candour that large companies rarely commit to print. The gap between AI mindshare and actual adoption, the company writes, exists because “the enterprise reinvention required to truly unlock the value of advanced AI is hard and has significant costs.” They go on to note that data preparedness is nascent, organisations are siloed, cloud and ERP modernisation is still incomplete, and workforces lack the skills to operate in an AI-enabled environment.

This is Accenture explaining, in its own annual report, exactly why its clients need Accenture.

And here is the paradox at the centre of the most important question in enterprise technology right now: is AI the thing that makes Accenture indispensable, or is it the thing that eventually makes Accenture unnecessary? I do not think the answer is simple. I think it is one of the most genuinely complex questions in business strategy today, and I want to work through it honestly rather than offer the comfortable narrative that Accenture’s investor relations team would prefer.

What Accenture Actually is?

Before discussing AI’s impact, it is worth being precise about what Accenture actually sells. It is not a technology company. It is not a software company. It is the world’s largest professional services firm, a business that sells human expertise, at scale, to the world’s largest organisations. Its 779,000 employees generated $69.7 billion in revenue in FY2025. That revenue splits almost exactly in half between consulting, project-based work advising clients on strategy, technology implementation, and transformation, and managed services, longer-term contracts where Accenture runs operations, maintains systems, and manages processes on behalf of clients.

The consulting half is what most people think of when they think of Accenture: teams of analysts and consultants deployed to client sites to deliver projects. The managed services half is less visible but more financially durable, multi-year contracts with meaningful termination costs that convert to revenue slowly and predictably. The consulting business grew 5% in local currency in FY2025. The managed services business grew 9%. That divergence is not an accident, and it is central to understanding how AI will affect this company.

The Surface Narrative, and Why it is Wrong in Both Directions

There are two simple narratives about AI and Accenture, and I think both are wrong.

The first is the bull narrative: AI creates enormous demand for implementation, change management, and enterprise transformation work. Clients need help deploying AI safely and at scale. Accenture is the partner they turn to. GenAI (generative AI) bookings reached $5.9 billion in FY2025, nearly doubled from the prior year. Revenue from generative AI and agentic AI reached $2.7 billion, tripling year-over-year. The company has 77,000 AI and data professionals, up from 40,000 in FY2023. It has trained over 550,000 of its employees in generative AI fundamentals. This is a company that positioned itself early, invested $3 billion in AI capability beginning in FY2023, and is now capturing the implementation wave. The AI opportunity is additive, not destructive.

The second is the bear narrative: AI automates exactly what junior consultants do. Writing code, analysing data, producing presentations, drafting documents, summarising research, building financial models, all of these tasks are being compressed by AI tools that any client can buy for $20 per user per month. The pyramid model that underlies Accenture’s economics, many juniors supporting fewer seniors, with juniors doing the volume work and seniors doing the judgment work, collapses when AI does the junior work. Revenue per engagement compresses. Headcount requirements fall. The business model is structurally impaired.

Both narratives capture something real. Neither captures the full picture.

The Pyramid Problem, This is the Real Risk

Let me start with the bear case because I think it is more structurally important than the bull case, even though the bull case is more visible in the near-term numbers.

Accenture’s operating model is built on a leverage pyramid. A small number of senior partners and managing directors sell and oversee client relationships. A larger number of managers and senior analysts do the intellectual work, designing solutions, leading workstreams, managing client relationships day-to-day. And a very large base of junior analysts and associates does the volume work, building models, writing code, conducting research, producing deliverables. This pyramid works economically because the juniors are relatively cheap, they generate billable hours that the senior layer monetises at a premium, and the pyramid widens at the base as the firm grows.

AI directly compresses the base of this pyramid. A junior analyst who previously spent three days building a financial model can now produce the same output in three hours with AI assistance. A developer who previously wrote 200 lines of code per day can write 800 lines with an AI coding assistant. A research team that previously spent two weeks analysing industry data can complete the same analysis in two days with AI-powered synthesis tools. These are not hypothetical capabilities — they are tools that Accenture’s own clients are deploying right now, and that Accenture’s own employees are using internally.

The implications for the business model are significant. If junior labour is three to four times more productive with AI, you need three to four times fewer juniors to deliver the same volume of work. That is not a problem if revenue grows proportionally, if the addressable market expands fast enough to absorb the productivity gain. But it is a fundamental structural problem if clients start asking why they should pay for 20 junior consultants when 5 can now deliver the same output. The answer, “because we have 779,000 people and can deploy them globally”, becomes less compelling when the leverage comes from AI rather than headcount.

The financial evidence of this tension is already visible, though subtle. Accenture’s gross margin fell to 31.9% in FY2025 from 32.6% in FY2024. The company attributed this to higher payroll costs. But the more revealing data point is the headcount reduction the company initiated in FY2025, $344 million in severance charges for “headcount reductions we are making in a compressed timeline.” A company that is simultaneously tripling its AI revenue and cutting headcount in a compressed timeline is not just managing capacity. It is restructuring its delivery pyramid in real time.

The Managed Services Moat

Here is where I think the bear narrative goes too far. It focuses almost entirely on the consulting business and largely ignores the managed services business, which is both larger and structurally different in ways that make it far more resistant to AI disruption.

Managed services, the $34.6 billion half of Accenture’s business, are long-term contracts where Accenture runs operations on behalf of clients. Application maintenance, infrastructure management, business process outsourcing, security operations. These contracts typically run three to five years with significant termination costs. They are based on outcome commitments, Accenture guarantees certain service levels, response times, and cost savings, rather than on the hourly billing of consultant time.

This is critically important. A client who has outsourced their SAP environment, their finance operations, or their cybersecurity monitoring to Accenture on a five-year contract does not reduce their payments because AI makes Accenture’s delivery team more efficient. Accenture captures the productivity gain from AI as margin expansion rather than passing it through as price reductions. The economics of managed services actually improve with AI, the same service level can be delivered with fewer people at lower internal cost, while the contractual revenue remains fixed.

This is the opposite dynamic from the consulting business, where clients can and will renegotiate based on observed productivity improvements. In managed services, the productivity gain is Accenture’s to keep. And the managed services business grew 9% in FY2025, faster than consulting at 5%, suggesting that clients are moving more work into this format precisely because they want to lock in AI-enabled efficiency gains without managing the complexity themselves.

The Agentic AI Moment

There is a dimension of the AI story that I think deserves specific attention because it is moving faster than most investors appreciate. Agentic AI, AI systems that can take autonomous actions, chain multiple tasks together, and operate continuously without human intervention, is beginning to change what enterprise AI deployment looks like.

Accenture describes deploying agentic AI systems that can “reinvent core business operations, streamline workflows and boost agility.” A client referenced in the annual report is deploying a system with 90 agents and 3,000-plus employees working alongside them. This is not a productivity tool layered on top of an existing workflow. It is a fundamental redesign of how work gets done, with AI agents operating in parallel with humans rather than simply assisting them.

For Accenture, agentic AI is both an opportunity and an existential question. The opportunity is that designing, deploying, and managing multi-agent systems at enterprise scale is genuinely complex work that requires deep expertise in AI architecture, change management, and process redesign, exactly the kind of work Accenture sells. The existential question is whether the agents themselves eventually replace the consultants who deployed them. An agentic system that automates a business process does not need to be maintained by a team of consultants indefinitely, it runs. The deployment engagement generates one-time revenue. The ongoing advisory relationship it displaces was recurring revenue.

This is the deepest tension in Accenture’s AI story. It is selling the tools that, if fully successful, reduce the long-term demand for its core product. Every enterprise AI transformation it helps a client achieve makes that client slightly less dependent on Accenture. The most successful consulting relationship is one that eventually makes itself unnecessary, and AI is accelerating that timeline.

The $5.9 Billion Number

Accenture reported $5.9 billion in generative AI bookings in FY2025 and $2.7 billion in generative AI revenue. These numbers are cited prominently in the annual report and in every investor communication. They are real, they are growing fast, and they tell you something important: clients are paying Accenture to help them deploy AI.

But the annual report contains a parenthetical that most analysts gloss over. These numbers, Accenture notes, “reflect only revenue and bookings specifically related to advanced AI and do not include data, classical AI or AI used in delivery of our services.” In other words, the $2.7 billion is a narrow slice of AI-related activity, specifically defined to exclude AI that Accenture uses internally to deliver its services more efficiently.

This matters because the most transformative AI happening inside Accenture right now is not the $2.7 billion, it is the AI that its own consultants and engineers are using daily to do their jobs faster. That AI does not show up in the headline AI revenue figure. It shows up in the gross margin compression, the headcount restructuring, and the quiet redesign of delivery pyramids that is happening across every large professional services firm simultaneously.

The $5.9 billion in generative AI bookings is the revenue opportunity. The pyramid restructuring is the cost reality. The net effect of both determines whether AI is a net positive or net negative for Accenture’s long-term economics.

My Honest Assessment

I think Accenture navigates the near-term AI transition better than most investors expect and worse than the company’s own narrative implies.

Better than expected because the managed services business, which now represents 50% of revenue and is growing faster than consulting, is structurally insulated from AI-driven price compression in ways that the consulting business is not. The long-term contractual nature of managed services means Accenture captures AI productivity gains as margin rather than passing them through as price cuts. This is a meaningful and durable economic advantage that the bear case on Accenture’s business model largely ignores.

Worse than the company’s narrative implies because the consulting business is facing a genuine structural challenge that cannot be resolved by rebranding it as AI-enabled transformation work. Clients who are becoming more sophisticated about AI, who are building internal capabilities, hiring their own AI teams, and deploying their own tools, will increasingly ask whether they need a 20-person Accenture consulting team or whether three of their own people with AI tools can deliver comparable output. That question gets harder to answer in Accenture’s favour with each passing year as AI tools improve.

The most honest summary is this: Accenture is one of the most capable organisations on earth at helping large companies navigate technology transitions. It has done this successfully through the internet era, the cloud era, and the mobile era. Each transition generated significant consulting revenue as clients needed help adapting. Each transition also eventually reduced the long-term demand for certain types of advisory work as the new technology became standard.

AI is the same pattern, but faster and more fundamental. Will the business emerge from the AI transition as large, as profitable, and as structurally advantaged as the one that entered it. On that question, I am genuinely uncertain. The managed services business argues yes. The consulting pyramid economics argue no. And the agentic AI dynamic, where Accenture’s most successful work makes its clients less dependent on it, introduces a long-term structural headwind that has no easy resolution.

For investors, the honest framing is not whether Accenture is a good business, it clearly is. It is whether the current price adequately reflects the structural uncertainty of a 779,000-person consulting firm navigating a technology transition that is, by its own admission, compressing the economics of the very work that built it.

That question deserves its own valuation analysis. But the starting point for that analysis has to be honest about what AI does to the consulting pyramid, not just what it does for the AI bookings number.

This report reflects the author’s personal views and is not an investment advice. Investing carries the risk of permanent capital loss. Read the full disclaimer here

There is a number that appears in almost every corporate earnings release, every leveraged buyout pitch, every bank credit memo, and almost every discussion of whether a company is financially healthy.

That number is EBITDA, Earnings Before Interest, Taxes, Depreciation, and Amortisation.

And in my view, it is one of the most overused and most misleading metrics in corporate finance.

This is not a fringe opinion. Charlie Munger once called EBITDA “Bullsh*t earnings.” Warren Buffett has spent decades explaining why depreciation is a very real cost that management teams conveniently prefer to ignore. But despite those warnings, EBITDA remains the dominant lens through which banks assess a company’s ability to repay debt, and the dominant shorthand through which analysts compare businesses.

The better alternative, Operating Cash Flow, sits right there in every set of financial statements, far more revealing, far harder to manipulate, and almost universally ignored in favour of the number that makes everything look bigger.

Here is why that matters, and why as a long-term investor, you should train yourself to reach past EBITDA every time.

What EBITDA Actually Is

EBITDA starts with net income and adds back four things: interest expense, tax expense, depreciation, and amortisation. The resulting number is supposed to represent a company’s core operating earnings, what the business generates before the effects of how it is financed, how it is taxed, and how its assets wear out over time.

The logic behind the addbacks seems reasonable on the surface. Interest expense varies depending on how much debt a company carries, strip it out so you can compare businesses with different capital structures. Tax rates vary by jurisdiction, strip those out too. Depreciation and amortisation are non-cash charges, add them back because no cash actually left the building when the accountant recorded them.

The problem is that the moment you reconstruct EBITDA in your mind, you realise it lives entirely in the income statement. It is built from revenue and expenses as recognised by the company’s accountants, not from cash that actually moved. And that distinction, which seems technical, turns out to be enormous in practice.

What Operating Cash Flow Actually Is

Operating Cash Flow (OCF) starts in a completely different place. It begins with net income and then adjusts for everything that happened between the income statement and the company’s actual bank account.

The most important adjustments are the working capital movements: things like changes in accounts receivable (money owed to the company by customers), inventory (goods sitting in a warehouse waiting to be sold), and accounts payable (money the company owes to its suppliers).

These three items represent the business cycle in brief, the lag between when revenue is recognised on the income statement and when cash actually arrives, and the lag between when expenses are recognised and when they are actually paid.

A company that books $100M in revenue but has not yet collected any of it has a wonderful income statement and an empty bank account. OCF captures that reality. EBITDA does not.

Why EBITDA Is Almost Always Higher

Here is a mechanical truth that most financial commentary glosses over: in most businesses, in most years, EBITDA will be higher than Operating Cash Flow. Sometimes significantly higher.

The reasons are structural. Working capital typically consumes cash as a business grows, receivables and inventory expand as sales increase, and this cash outflow never touches the income statement. Meanwhile, depreciation and amortisation, which were added back to create EBITDA, represent real economic consumption of the asset base that will eventually require real cash to replace. A piece of machinery that depreciates over ten years does not magically regenerate itself at the end of year ten. The cash to replace it has to come from somewhere.

When you add back depreciation to create EBITDA and then use that number to assess a company’s financial health, you are implicitly claiming that the machinery doesn’t need replacing. Every manufacturing company knows that is not true.

The Manipulation Problem

EBITDA is not just theoretically imprecise. It is also practically easy to inflate.

The most straightforward manipulation is recognising revenue aggressively. A company can book a sale the moment goods leave the warehouse, even if the customer hasn’t paid or won’t pay for six months. That revenue flows directly into EBITDA. It does not flow into Operating Cash Flow, where the receivables balance would swell visibly and alert any attentive analyst that something is off.

This is not a hypothetical risk. It is one of the most common patterns in corporate fraud.

Sunbeam, the American appliance manufacturer, provides a textbook case. In the late 1990s, CEO Al Dunlap, known as “Chainsaw Al”, drove reported earnings higher by selling products to retailers at heavily discounted prices with generous return rights, recognising the revenue immediately. EBITDA looked healthy. Operating Cash Flow told a very different story, the company was consuming cash at an alarming rate as receivables ballooned. Sunbeam filed for bankruptcy in 2001.

Closer to the present, many highly leveraged companies that appeared financially manageable through the EBITDA lens, revealed their true fragility when interest rates rose and their working capital cycles deteriorated. The income statement said they were profitable. The cash flow statement said they were running dry.

Why Banks Use EBITDA Anyway

If EBITDA is so flawed, why does almost every bank in the world use it as the primary measure of debt capacity?

The answer is partly historical, partly structural, and partly, I will be direct, because it serves the bank’s commercial interests.

Historically, EBITDA became the standard in leveraged finance during the 1980s leveraged buyout boom. Dealmakers needed a metric that could justify higher debt levels on acquisitions, and EBITDA, which ignores working capital movements, strips out the cost of replacing assets, and excludes the interest expense on the very debt being assessed, produced the highest possible number. It became institutionalised.

Structurally, the banking system adopted it so broadly that any single bank switching to OCF-based underwriting would be at a competitive disadvantage, they would approve smaller loans than their peers, lose deals, and see revenue decline. The incentive to use the more conservative metric is weak when competitors are not.

And then there is the commercial reality: a higher EBITDA justifies a larger loan. A larger loan generates more fee income, more interest income, and a bigger balance sheet. There is a direct financial incentive for the banking system to use the metric that produces the highest debt capacity, and EBITDA is that metric.

I spent fifteen years in institutional finance. I have sat in credit committees where the EBITDA multiple was the headline figure and the cash flow statement was barely discussed. This is not a theoretical observation, it is a standard practice.

What to Use Instead

Operating Cash Flow is not perfect either. It can be managed through timing of receivables collections, stretching of payables, and opportunistic working capital draws before year-end. A skilled CFO can compress the working capital cycle in the fourth quarter to produce a better-looking OCF number than the underlying trend warrants.

But the manipulation is harder, the signals are clearer, and the number is fundamentally more honest because it reflects cash that actually moved.

For assessing a company’s debt capacity, OCF minus maintenance capex, what some call Owner Earnings or Free Cash Flow, is the most relevant figure. It represents what the business actually generated after keeping the existing asset base functional. That is what is available to service debt, pay dividends, fund growth, or return to shareholders. EBITDA does not answer that question cleanly. FCF does.

For comparing profitability across businesses, operating margin, operating income as a percentage of revenue, is a more reliable starting point than EBITDA margin, because operating income includes depreciation and therefore acknowledges the cost of the assets generating the revenue.

The Investor’s Takeaway

When I look at a business, I use EBITDA as a starting point at most, a rough orientation before I do the real work. The questions that matter to me are:

How does Operating Cash Flow compare to EBITDA? If the gap is consistently large, I want to understand why. A large and growing gap is often the first sign that something is wrong in the working capital cycle.

Is FCF growing alongside revenue, or is the business consuming more cash as it scales? A business that grows revenue but consistently burns cash is not compounding, it is borrowing against the future.

What is the capex-to-OCF ratio, and how much of capex is maintenance versus growth? Maintenance capex is not optional. It should never be ignored when assessing a business’s true earning power.

EBITDA will tell you what a company wants you to think about its earnings. Operating Cash Flow will tell you what is actually happening.

The Outlook

There is a condition that affects nearly one billion people on earth. Most of them don’t know they have it. Their doctors haven’t diagnosed it. Their partners have complained about the snoring, the gasping, the restless sleep, but the link to a treatable medical condition has never been made. Obstructive sleep apnea, or OSA, is one of the most prevalent and undertreated chronic conditions in modern medicine, and ResMed has spent thirty years quietly building the most comprehensive platform for diagnosing, treating, and managing it.

This is not a story about a niche medical device company. It is a story about a platform business with thirty million cloud-connected patients, a proprietary data advantage that compounds with every device sold, and a software layer that makes it operationally difficult for the healthcare providers who use it to switch to anyone else. The hardware is just the door into the ecosystem.

Revenue has grown at 11.9% annually for a decade. Operating margins have expanded from 24% to 33%. Free cash flow has increased more than fivefold. And yet fewer than 20% of the people who need this product have ever received it.

The stock is not cheap. At 23 years of embedded cash flows, ResMed sits in the Hold zone of the Bearhold valuation framework, a fair price for a business of exceptional quality. I am not buying here. But I am watching, and this report explains exactly what I am waiting for.

At a Glance

Company: ResMed Inc.

Ticker: $RMD · NYSE

Sector: Healthcare

Industry: Medical Devices & Instruments

Market Cap: $32.65 billion (at $224)

Dividend Yield: ~0.95% ($2.12 per share, FY2025)

Status: Approved, Bearhold Universe

First Coverage: April 2026

Valuation Zone: Hold (last updated April 2026)

Disclaimer: This report reflects the author’s personal views and is not an investment advice. Investing carries the risk of permanent capital loss. Read the full disclaimer here

The author does not currently hold a position in $RMD at the time of publication.

1. The Business

What the Company Does

ResMed makes the devices and software that treat sleep-disordered breathing. The flagship products are CPAP (Continuous Positive Airway Pressure) and APAP (Automatic Positive Airway Pressure) machines, compact bedside devices that deliver pressurised air through a mask while the patient sleeps, keeping the airway open and eliminating the breathing disruptions that cause OSA. The condition, if left untreated, is associated with hypertension, stroke, type 2 diabetes, and coronary artery disease, which is why the clinical community increasingly treats it as a cardiovascular risk factor, not just a sleep nuisance.

Revenue breaks down into three streams: devices at approximately 52% of the total, masks and accessories at 36%, and Residential Care Software at 12%. The structure matters. Devices are event-driven, a patient buys one at diagnosis and replaces it every few years. Masks are recurring, replaced every three to six months for as long as the patient stays on therapy. Software is subscription-based. The combination produces a revenue profile far more stable than a pure device company, with the recurring mask and software streams anchoring the base even in softer device years.

How the Business Was Built

ResMed traces its origins to June 1989, when Dr. Peter Farrell founded the company, originally called ResCare, in Sydney, Australia, to commercialise nasal CPAP technology invented by Professor Colin Sullivan at the University of Sydney in 1981. Baxter Healthcare had licensed the technology and then decided not to pursue it. Farrell recognised what Baxter had missed.

ResMed Inc. was incorporated in Delaware in March 1994 and went public on June 1, 1995, trading on NASDAQ. The company moved its primary listing to the New York Stock Exchange in September 1999.

For its first two decades, ResMed was a device company that grew by making better CPAP machines and expanding geographically. The inflection came around 2014, when the company began embedding cellular connectivity into devices as standard, a decision that looks obvious in hindsight but required genuine conviction at the time. Connected devices enabled remote monitoring at scale. Clinicians could adjust therapy settings without requiring patients to return to a clinic. And every connected patient became a source of real-world clinical data. That data now feeds the machine learning algorithms that make each new device generation smarter than the last. Today, ResMed manages over 30 million cloud-connected patients through its AirView platform, with more than 10 million active on myAir, the patient-facing therapy management app.

The flywheel is real: more patients generate more data, which improves therapy outcomes, which attracts more patients and more providers into the ecosystem.

The Philips Recall, and What Actually Happened

In June 2021, Philips issued a recall of millions of CPAP, bilevel, and ventilator devices after discovering that the polyester-based sound abatement foam inside the machines was degrading and potentially releasing particles and gases into the patient’s airway. It was a significant product safety failure that triggered multi-billion euro legal settlements and years of operational disruption for Philips.

ResMed was the only large-scale alternative with the manufacturing capacity to absorb displaced patients quickly. The impact on ResMed’s numbers was real but more targeted than widely reported. Overall revenue growth remained broadly consistent with historical trends. The visible surge was concentrated in the US, Canada, and Latin America device segment specifically, where device revenue grew approximately 24% in FY2022 and approximately 35% in FY2023, compared with around 9% in the year before the recall. By FY2024, device growth in that segment had normalised to approximately 11%, and FY2025 came in at 9.8% for total revenue, a return to the underlying demand trajectory. The recall pulled forward some volume and concentrated it geographically, but the business was not propped up by it.

Scale

ResMed employs more than 10,600 people across more than 140 countries. The US, Canada, and Latin America generate approximately 58% of Sleep and Breathing Health revenue. Combined Europe, Asia, and other markets contribute approximately 29%. Residential Care Software, sold only in the US and Germany, rounds out the remaining 12%. Manufacturing is split across Australia, Singapore, and the US, with the company running an active foreign currency hedging programme to manage the exposure this creates.

2. The Moat

Four Walls, One Ecosystem

ResMed’s competitive position is not a single advantage, it is four advantages that reinforce each other.

The first is clinical data. ResMed has more real-world sleep therapy data than any organisation on earth. Thirty million cloud-connected patients generate continuous information, therapy compliance, apnea-hypopnea index, mask leak, breathing patterns. This data feeds the algorithms that make the AirSense 11, and whatever comes after it, progressively more effective. A competitor launching a CPAP device today faces not just a product quality gap but a data gap that will take years to close.

The second is the installed base and the recurring revenue attached to it. Thirty million active patients buying masks every three to six months is a durable annuity that does not depend on the next product launch. Patients who have established a functioning therapy routine are unlikely to switch ecosystems.

The third is the software layer. Brightree, the leading home medical equipment (HME) software platform in the US, and MEDIFOX DAN in Germany are deeply integrated into the administrative and clinical workflows of thousands of care providers. These platforms are not standalone software businesses, they are the connective tissue between device sales and the care delivery system. When a provider uses Brightree to manage billing, patient records, and inventory, switching to a competitor means disrupting those workflows entirely. That is a meaningful switching cost that goes well beyond product preference.

The fourth is regulatory and clinical credibility. ResMed’s devices are approved across virtually every major regulatory jurisdiction globally, supported by decades of peer-reviewed clinical evidence. New entrants face not just a commercial challenge but a multi-year regulatory validation process. That is a barrier that does not exist in consumer device categories.

The Wearables Tailwind

An often overlooked accelerant is the integration of OSA detection into consumer wearables. Apple Watch and Samsung Galaxy Watch now feature FDA (US Food and Drug Administration)-cleared breathing disturbance detection. These devices are identifying undiagnosed OSA patients at population scale and nudging them toward clinical care. ResMed has positioned itself as the natural destination for that flow, its myAir app integrates directly with Apple Health and Samsung Health, and CEO Mick Farrell has explicitly described the consumer wearables ecosystem as building the company’s “data lake.” The most powerful referral engine ResMed could imagine is a device worn by hundreds of millions of people, many of whom don’t yet know they need a CPAP machine.

The Numbers Don’t Lie

The financial record is the most reliable evidence of a moat. ResMed’s operating margin expanded from 24.4% in FY2015 to 32.8% in FY2025. ROIC (Return on Invested Capital) averaged 18.9% over the decade, well above the estimated cost of capital throughout. A business losing competitive position does not do that while tripling its revenue.

The comparison with Philips makes the point most cleanly. Before the recall, Philips operated at approximately 8% operating margin in its health technology division, a level ResMed was already far above and moving away from. After the recall, Philips fell into losses in 2022 and 2023. By 2025, it had only recovered back to that same 8% baseline. Over the same period, ResMed’s operating margin continued to expand. Two companies nominally competing in the same market, on very different structural trajectories.

3. Financial Performance

A Decade in Numbers

Revenue grew from $1.68 billion in FY2015 to $5.15 billion in FY2025, an 11.9% compound annual growth rate (CAGR) over ten years. That growth was primarily organic. Two meaningful acquisitions, the Brightree HME software business in FY2016 and MEDIFOX DAN in FY2023, added inorganic revenue, but the underlying Sleep and Breathing Health segment has compounded consistently without acquisition support.

Operating margin expanded from 24.4% to 32.8%, with a ten-year average of 26.5%. Gross margin averaged 57.6% across the decade, reflecting the premium positioning of ResMed’s clinical products relative to commodity device alternatives. Net margin in FY2025 reached 27.2%, up from 21.0% in FY2015, aided in part by a below-trend effective tax rate of 16.5% due to one-time items, a level that should not be assumed to repeat.

Diluted EPS (Earnings Per Share) grew from $2.47 in FY2015 to $9.51 in FY2025, a 14.4% CAGR. The share count has been essentially stable over this period, rising only 3.2% from 142.7 million to 147.3 million diluted shares. EPS growth here reflects genuine business improvement rather than financial engineering.

ROIC averaged 18.9% over the decade and reached 23.3% in FY2025. In every year of the available record, ResMed earned returns materially above its estimated cost of capital, a consistency that is more impressive than any single-year figure.

Free Cash Flow, the Real Story

Free cash flow (FCF) grew from $311M in FY2015 to $1,651M in FY2025, a 5.3x increase at an 18.2% CAGR. There was one significant dip: FY2022, when FCF collapsed to $195M despite strong revenue. The cause was working capital. The Philips recall demand surge required ResMed to draw down inventories and then rapidly rebuild them, creating a large transient cash outflow. Simultaneously, an elevated FY2021 effective tax rate of 46.3%, driven by a one-time $200M charge related to pre-acquisition earnings, further compressed cash in that period. Both were temporary. FCF recovered to $559M in FY2023, $1,286M in FY2024, and $1,651M in FY2025.

The FCF margin of 32.1% in FY2025 is the highest in the company’s history. Capex (capital expenditure) as a percentage of operating cash flow has fallen from 19% in FY2015 to just 6% in FY2025, reflecting the asset-light nature of a platform built primarily on software, data, and intellectual property rather than physical infrastructure.

Balance Sheet

Cash and equivalents stood at $1.21 billion at June 30, 2025, against total debt of approximately $670M, producing a net cash position of roughly $540M. Debt-to-equity has improved dramatically from the 0.38 reading in FY2023 (following the MEDIFOX DAN acquisition) to 0.14 at year-end FY2025. The revolving credit facility of $1.5 billion remains fully undrawn, maturing in 2027.

One genuine weakness worth naming: goodwill of $3.05 billion represents approximately 37% of total assets. This reflects the accumulated premium paid for acquisitions, primarily Brightree, Propeller Health, and MEDIFOX DAN. Tangible book value per share is substantially lower than reported book value. The intrinsic value of this business rests almost entirely on intangible assets, data, software, clinical relationships, regulatory approvals. That is a strength when the business performs and an accounting vulnerability if any major acquisition disappoints.

ResMed also pays a dividend of $2.12 per share in FY2025, representing a yield of approximately 0.95% at the current price, a signal of financial maturity that has grown steadily from $1.12 per share a decade ago.

4. Growth Levers

The Diagnosis Gap, the Single Biggest Lever

Here is the most important number in this entire report: fewer than 20% of OSA sufferers in the United States have been diagnosed and treated. In most international markets, the figure is closer to 10%.

ResMed does not need to take market share from anyone. It just needs the healthcare system to get better at finding the patients who already exist. Every percentage point improvement in diagnosis rates represents millions of people entering a care pathway that runs directly through ResMed’s ecosystem. This is the kind of structural growth driver that does not require macroeconomic tailwinds, product cycles, or competitive displacements. It just requires time.

ResMed is actively investing to accelerate this. NightOwl, a portable, cloud-connected, fully disposable diagnostic device that measures OSA severity overnight without requiring a clinic visit, launched across the US in April 2025. The acquisition of VirtuOx, an independent diagnostic testing facility, in May 2025 further expands the company’s ability to bring diagnosis into the home. The goal is to compress the traditional pathway, specialist referral, sleep lab, equipment pickup, device initiation, into something much closer to a single session at home.

International Penetration, COPD, and Software

The international opportunity mirrors the domestic one, large, underpenetrated, and growing as clinical awareness spreads and reimbursement coverage expands. International Sleep and Breathing Health revenue grew 9% in FY2025, and this segment operates at a fraction of US penetration levels across most geographies.

COPD (Chronic Obstructive Pulmonary Disease) affects approximately 480 million people globally and is the world’s third leading cause of death. ResMed’s non-invasive ventilation (NIV) and high-flow therapy (HFT) products for COPD patients represent a growing segment that the revenue breakdown does not currently call out separately, but it is a meaningful adjacency that deepens the platform’s clinical reach beyond sleep.

Residential Care Software, currently sold only in the US and Germany, grew 10% in FY2025 to $641M. Geographic expansion of this segment is a medium-term option that has not yet been exercised.

5. Management

The Farrell Family and Long-Term Alignment

ResMed is, at its core, a founder-influenced business. Dr. Peter Farrell built ResCare from a single licensed technology in Sydney in 1989 and steered it to become the global leader in sleep therapy. His son Mick Farrell has served as CEO since 2013, presiding over the company’s transformation from a device manufacturer into a connected care platform.

The family maintains meaningful skin in the game. Mick Farrell holds approximately 0.32% of outstanding shares, valued at roughly $105M at current prices. Dr. Peter Farrell, as founder and board director, directly owns a further 60,773 shares. While institutional investors including Vanguard and BlackRock collectively own the majority of the company, the Farrell family’s retained ownership signals genuine long-term conviction rather than founders who cashed out at the earliest opportunity.

Mick Farrell’s compensation structure reinforces that alignment. Approximately 91-92% of his total pay is performance-based, stock awards and options tied to specific financial metrics including revenue growth and EPS. Base salary accounts for only about 8% of total compensation. This is the kind of structure I look for in management: the CEO is economically motivated by the same outcomes that matter to long-term shareholders.

His communication style has been consistently direct on difficult topics. When the GLP-1 (glucagon-like peptide-1) panic hit in mid-2023 and ResMed’s stock fell more than 30%, Farrell engaged with the clinical evidence head-on rather than deflecting. He presented ResMed’s own patient data on GLP-1 users, made the probability-weighted case for ongoing demand, and continued investing in the business rather than cutting costs to manage short-term numbers. That is the kind of management behaviour that matters over a full market cycle.

Capital Allocation

R&D (Research and Development) spending of $331M in FY2025, representing 6.4% of revenue, has been sustained consistently across the decade. This is the right posture for a business whose competitive position depends on staying ahead clinically and technically.

On acquisitions, the track record is mixed but acceptable. Brightree has been clearly value-creating, it transformed ResMed from a device company into a platform company with deep provider relationships. The MEDIFOX DAN acquisition at approximately EUR 975M in FY2023 is too early to assess definitively, though integration appears to be progressing. The Propeller Health acquisition in FY2019 at $225M has been harder to evaluate but is smaller in scale.

Shareholder returns have grown steadily: dividends increased from $1.12 to $2.12 per share over the decade, and buybacks of $300M were executed in FY2025. Together with dividends, that represents approximately 37% of FCF returned to shareholders, the remainder retained for investment. I find that balance reasonable given the available organic growth runway.

6. Valuation

What the Market Is Pricing In

At $224 per share, I estimate the market is pricing ResMed at approximately 23 years of discounted future cash flows. My valuation framework expresses price as the number of years of a company’s future cash flows, discounted at an appropriate rate, that are already embedded in today’s price. The assumptions I use are consistent with the company’s current and historical operating performance. Nothing heroic, nothing pessimistic.

23 years puts ResMed in the Hold zone.

To be direct about what that means for me: I don’t initiate new positions in the Hold zone. The Bearhold framework is built around the idea that when you are buying a stake in a business, you want the price working in your favour, not just the fundamentals. The Attractive zone (16–20 years) and the Exceptionally Attractive zone (below 15 years) are where I look to deploy capital. At those levels, the price embeds significantly less of the company’s future than the business’s quality justifies, and an investor captures both the fundamental compounding and the valuation correction as it plays out.

At 23 years, I am essentially paying a fair price for quality I can see clearly. The fundamentals of this business are genuinely exceptional, but there is limited margin of safety in the entry price. Every dollar of return has to be earned through the business’s ongoing performance, there is no valuation tailwind to help.

What I Am Waiting For

ResMed becomes interesting to me as a new position somewhere in the Attractive and the Exceptionally Attractive zones. That does not happen often for businesses of this quality. It tends to happen when the market becomes temporarily preoccupied with a specific risk, a GLP-1 study result, a reimbursement cut, an earnings miss driven by inventory timing, and the stock reprices faster than the underlying business warrants.

7. Risks

The GLP-1 Question

Since mid-2023, the dominant narrative around ResMed has been the GLP-1 risk. Glucagon-like peptide-1 agonist drugs, Ozempic, Wegovy, Zepbound and others, cause significant weight loss in a meaningful proportion of users, and obesity is a primary driver of OSA in a large patient population. If these drugs reduce OSA prevalence materially, the argument goes, ResMed’s addressable market shrinks.

The clinical evidence is real. The FDA approved tirzepatide (Zepbound) in December 2024 specifically for the treatment of moderate-to-severe OSA in adults with obesity, after clinical trials demonstrated meaningful reductions in apnea-hypopnea index among treated patients.

I take this risk seriously. But I believe it is more limited than the 2023 market reaction implied, for three reasons.

First, OSA is not purely mechanical. The condition has neurological and anatomical components, airway geometry, muscle tone, neural respiratory drive, that weight loss does not fully resolve. ResMed’s own connected patient data shows that a meaningful proportion of GLP-1 users who achieve significant weight reduction continue to require CPAP therapy.

Second, the undiagnosed population dwarfs the treated population. The 80% of OSA sufferers who have never received a diagnosis are entering the care pathway as awareness grows. This demand source is entirely independent of GLP-1 penetration.

Third, real-world GLP-1 adherence is substantially lower than clinical trial completion rates. Weight regain is common on discontinuation. Access constraints, particularly outside developed markets, further limit penetration.

This is a risk that deserves ongoing monitoring, particularly as longer-duration GLP-1 data accumulates. It is not a reason to avoid the business; it is a reason to track it carefully and incorporate it into the price I am willing to pay.

Reimbursement and Policy Risk

A meaningful portion of ResMed’s revenue flows through Medicare, Medicaid, and private insurers. The US DMEPOS (Durable Medical Equipment, Prosthetics, Orthotics and Supplies) Competitive Bidding Program has historically placed downward pressure on Medicare reimbursement rates. CMS (Centers for Medicare & Medicaid Services) rate changes and ACA (Affordable Care Act) enrollment policy are ongoing headwinds. The recently enacted One Big Beautiful Bill Act makes changes to Medicaid funding and ACA enrollment that could reduce the insured patient population over time, though the specific effect on sleep therapy demand remains uncertain.

This is a persistent risk that ResMed has navigated through multiple policy cycles. It acts as a headwind to pricing power but has not historically disrupted the underlying demand trajectory.

Goodwill and Acquisition Risk

Goodwill of $3.05 billion, 37% of total assets, reflects the cumulative premium paid for acquisitions. The MEDIFOX DAN acquisition at approximately EUR 975M is the largest and most recent at scale. If integration underdelivers, or if the German software market dynamics shift unfavourably, a material impairment charge would hit reported equity significantly. This is the honest weakness in an otherwise strong balance sheet.

Tariffs and Supply Chain

ResMed manufactures primarily in Australia and Singapore. The tariff environment as of early 2025 has created some input cost uncertainty, though medical device tariff relief has been confirmed through US Customs as of the FY2025 filing date. The situation remains dynamic and worth monitoring.

What Would Change My Mind?

If GLP-1 clinical data continues to accumulate showing high real-world adherence and sustained OSA resolution, not just single-study results, that would require a fundamental reassessment of the long-term volume outlook. If operating margins begin contracting despite revenue growth, that signals competitive pricing pressure and potential moat erosion. If a major acquisition integrates poorly or requires a goodwill writedown, that is a capital allocation warning I would take seriously.

8. The Verdict

ResMed is one of the clearest examples of a business with genuine, compounding structural advantages. The data platform took thirty years to build. The installed base generates recurring revenue that grows with each diagnosis. The software layer makes providers sticky. The clinical evidence is deep and growing. And the addressable market, driven by the vast undiagnosed population, has decades of runway left.

The financial record across ten years confirms the quality: revenue tripled, operating margins expanded by 8.4 percentage points, ROIC averaged 18.9%, and free cash flow grew fivefold. This is what a durable competitive advantage looks like in the numbers.

The GLP-1 risk is real. The goodwill load deserves monitoring. Reimbursement policy is a persistent headwind. These are honest weaknesses and they belong in any fair assessment of the business.

ResMed is Approved in the Bearhold Universe. It sits in the Hold zone at the time of this publication. I am not buying today. But if the price moves into the Attractive zones, whether through a market correction, a temporary earnings disruption, or another episode of fear about a risk the market overstates, I will be ready to act.

Disclaimer: This report reflects the author’s personal views and is not an investment advice. Investing carries the risk of permanent capital loss. Read the full disclaimer here

The author does not currently hold a position in $RMD at the time of publication.

Sources

• ResMed Inc. Form 10-K, Fiscal Year Ended June 30, 2025 (filed August 8, 2025)

• ResMed Inc. Form 10-K, Fiscal Year Ended June 30, 2024 (filed August 9, 2024)

• GuruFocus Financial Database — ResMed (NYSE: RMD), extracted April 4, 2026

• GuruFocus Financial Database — Philips (NYSE: PHG), extracted April 5, 2026

• Philips Annual Reports 2020–2025

• Lancet Respiratory Medicine (2019): “Estimation of the global prevalence and burden of obstructive sleep apnoea”

• FDA Drug Approval: Tirzepatide (Zepbound) for OSA, December 2024

• ResMed Press Releases: NightOwl US launch (April 2025); VirtuOx acquisition (May 2025)

• CMS DMEPOS Competitive Bidding Program documentation

• ResMed Proxy Statement (DEF 14A), filed October 2025 — executive compensation and insider ownership

There is a question I ask about every business before anything else: can I tell, with reasonable confidence, how this company will look in five years?

Not the exact numbers. Not the precise revenue figure or the margin to the decimal. Just the shape of it, the competitive position, the pricing power, the reason customers keep coming back regardless of what is happening in the world around them.

For most businesses I cover, the answer is grounded in something structural. An automotive parts distributor benefits from an ageing vehicle fleet that gets more complex and more expensive to repair every year. A salvage auction operator sits at the intersection of rising total loss frequency and a global buyer network that took decades to build. A payment network processes more volume every time a cash transaction moves to digital. These businesses have tailwinds I can reason about independently of consumer sentiment, cultural trends, or what happens to be fashionable this season.

Nike does not have that. And that is the reason it is not included in the Bearhold Universe.

What Nike Actually Sells

Nike is one of the most recognised brands on earth. The marketing is exceptional. The athlete relationships are unmatched. The distribution is global. None of that is in question.

But when you strip it back to what the customer is actually paying for, the answer is identity. The person buying a pair of Air Maxes or a Nike training kit is not primarily paying for a functional outcome. They are paying to be associated with something they find culturally relevant, a feeling, an image, an aspiration. That relationship is real. It is also fragile in a way that other business models simply are not.

Identity is subject to taste. And taste shifts without warning, without logic, and without giving management teams much time to respond.

The Problem with Taste as a Business Driver

The history of consumer brands is full of businesses that looked like compounders until the moment they didn’t. Brands that commanded premium pricing, built loyal followings, and generated strong returns for years, right up until something shifted in the culture and the loyal customer turned out to be loyal to the aesthetic, not the company.

This is not a management failure. It is a category characteristic. When your competitive advantage rests on cultural relevance, you are permanently exposed to a risk that no amount of operational excellence can fully insulate you from. Competitors do not need to engineer a superior product. They just need to feel fresher at the right moment.

Nike is experiencing exactly this right now. Revenue declined roughly ten percent in its most recent fiscal year. Operating margins have compressed significantly from historical levels. Competitors have established meaningful positions in running, a category that historically reinforced Nike’s performance credibility and pricing power.

The company is rebuilding its wholesale relationships after a strategic overcommitment to direct-to-consumer that did not deliver on its promise. China, which represents a meaningful share of the business, has been declining for several consecutive quarters.

Some of this is execution. But not all of it. Some of it is simply what happens when taste moves on and the brand has to work harder to stay relevant than it used to.

Subscribe now

What I Need to See Instead

When I build a position in a business, I want to be able to answer one question clearly: why will this company’s customers still be here in five years?

For the businesses that pass the quality filter, the answer is almost always structural. The switching costs are high, the network effects are real, or the service is so deeply embedded in the customer’s operations that replacing it is more painful than paying for it. These businesses do not need to be culturally relevant. They need to be necessary.

Nike is not necessary. It is desired. And desire, unlike necessity, is something that has to be earned back every season.

The status

Nike is Rejected in the Bearhold Universe. Not because it is a bad company, it is not. Not because the brand is broken, it is not. But because the uncertainty embedded in its revenue model is not the kind of uncertainty I am willing to hold through cycles.

The quality filter exists precisely for moments like this. A well-known name, a strong brand, a long operating history, none of that is sufficient if the fundamental question cannot be answered with confidence.

I want to own businesses where the answer to “why will customers still be here in five years” is obvious. For Nike, it is not.

Subscribe now

This report reflects the author’s personal views and is not an investment advice. Investing carries the risk of permanent capital loss. Read the full disclaimer here

The Outlook

Copart is one of the most quietly exceptional businesses in the United States. It operates in a market that many investors have never thought about, the online auctioning of salvage vehicles, and it has built a position in that market that is almost impossible to challenge. The business requires more land than almost any competitor can afford to acquire, it improves as it gets bigger, and its primary customers have no viable alternative. The stock price at the time of writing reflects a business priced for moderate expectations, which is unusual for a business of this quality.

The two risks that matter, concentration among insurance sellers, and the long-term question of autonomous vehicles, are real and deserve honest treatment. Neither, in my view, overrides the fundamental quality of what has been built here. This report addresses both directly.

At a Glance

Company: Copart, Inc

Ticker: $CPRT · NASDAQ

Sector: Industrials

Industry: Online Vehicle Auctions & Remarketing

Market Cap: $31.9 billion (at $33.06)

Status: Approved

First Coverage: April 2026

Valuation Zone: Attractive (last updated in April 2026)

Disclosure update: The author now holds a position in CPRT. The position was initiated after the original publication date of this report. All analysis and conclusions remain unchanged. This report reflects the author’s personal views and is not an investment advice. Investing carries the risk of permanent capital loss. Read the full disclaimer here

1. The Business

What it Does

Copart operates online auctions for salvage and total-loss vehicles. When a car is involved in an accident and the insurance company determines that the cost of repair exceeds the vehicle’s value, the car is declared a total loss. The insurance company pays the policyholder the car’s pre-accident value, takes title to the vehicle, and then needs to dispose of it. That is where Copart comes in.

Copart takes possession of the vehicle, transports it to one of its storage facilities, photographs and documents it thoroughly through its proprietary technology, and lists it for sale on its VB3 online auction platform. Buyers from around the world, vehicle dismantlers, rebuilders, used car dealers, exporters, bid against each other in real time. Copart collects fees from both the seller and the buyer. The vehicle sells. The process repeats.

This description makes the business sound simple. It is operationally complex at scale, and that complexity is the foundation of the moat.

How the Company Makes Money

Copart earns fees at multiple points in the process. Sellers, primarily insurance companies, pay processing fees, listing fees, transportation fees, and storage fees. Buyers pay transaction fees, title fees, and loading fees. Because a significant portion of fees are tied to the final auction selling price under Copart’s Percentage Incentive Program, the company has a direct financial incentive to maximise the value each vehicle achieves. This alignment between Copart’s revenue and its sellers’ outcomes is a structural differentiator.

In fiscal year 2025 (ended July 31, 2025), Copart generated total revenues of $4.65 billion, of which $3.97 billion was service revenue and $678 million was vehicle sales revenue, the latter from markets like the UK, where Copart purchases and resells vehicles on a principal basis.

History and Origin

Copart was founded in 1982 by Willis Johnson in Vallejo, California with a single salvage yard. The founding insight was straightforward but consequential: insurance companies needed a professional, efficient, and geographically distributed way to dispose of total-loss vehicles. The fragmented industry of local salvage dealers was not meeting that need well.

The business grew through acquisitions of regional salvage yards over the following decade and went public in 1994. For most of its early history Copart held physical auctions, buyers would travel to its yards to bid on vehicles in person. The transformative moment came when the company began migrating its auction process online in the early 2000s, completing the transition across its US operations by the mid-2000s. This was not an incremental improvement. It was a fundamental reshaping of who could participate in each auction.

By opening every sale to buyers anywhere in the world with internet access, Copart dramatically expanded the pool of bidders competing for each vehicle. More bidders means more competition. More competition means higher prices. Higher prices means insurance companies get better returns on their salvage. Better returns means insurance companies want to work with Copart. The flywheel that would define the business for the next twenty years was set in motion.

Willis Johnson stepped back from day-to-day operations and the company is now led by Jeffrey Liaw, who joined Copart in 2016 as CFO. He was promoted to President in 2019 and became Co-CEO in 2022 before taking over the sole CEO position.

Scale and Footprint

Copart operates in the United States, Canada, the United Kingdom, Germany, Brazil, Spain, Ireland, Finland, the UAE, Oman, and Bahrain. The US business generates approximately 83% of total revenues and is the operational core of the company.

Copart operates over 21,000 acres of land globally, they own more than 90% of their operational land outright, a key differentiator from competitors who typically lease their facilities. In the US alone, Copart owns or leases facilities in every state. The company owns approximately $2.39 billion worth of land on its balance sheet, a figure that significantly understates fair value given how long much of this land has been held and how dramatically land prices around major population centres have appreciated.

The international business is less mature and operates differently in some markets, particularly the UK, where Copart buys vehicles outright. But the global buyer base is central to the model: Copart now maintains a database of approximately 1 million registered members across every continent. The scale of that global buyer pool, and the competitive pressure it places on every auction, remains central to the financial returns Copart delivers to its insurance sellers.

2. The Moat

Source of Competitive Advantage

Copart’s competitive position rests on three interlocking advantages that have strengthened over time. Understanding each one individually understates how they reinforce each other.

The first is physical infrastructure. Copart requires large parcels of land near major population centres to store the high volume of vehicles it processes before they sell. This land is expensive, increasingly scarce, and subject to complex local zoning restrictions that make new entrants face a fundamentally different cost environment than the one Copart faced when it was building its network. The company has spent decades and billions of dollars assembling a footprint that a competitor would need to replicate entirely, at today's land prices, with today's zoning restrictions, competing for the same properties in the same markets. No rational capital allocator would attempt it. This is not a moat that can be disrupted by software or a new technology. It is a physical asset base that took forty years to build.

The second is the global buyer network. Copart has accumulated hundreds of thousands of registered buyers across every continent. The value of this network is directly proportional to its size: more buyers means more competition for each vehicle, which means higher selling prices for insurance companies, which means insurance companies prefer Copart. A new entrant cannot simply build a marketplace without supply. It cannot build supply without having already demonstrated it can achieve competitive selling prices. The chicken-and-egg nature of this problem is the classic marketplace moat, and Copart has been building its side of it for four decades.

The third is the contractual relationships with insurance companies. While no single insurance company accounts for more than 10% of Copart's revenues, an important point to which we will return, the company has established long-term supply agreements with the major carriers. These agreements are built on trust, track record, and the demonstrable financial returns Copart generates for its sellers. Switching costs are not contractual so much as they are practical: an insurance company that moves its business to a smaller competitor will likely see lower auction returns, a worse buyer experience, and disruption to its operations. The cost of switching exceeds the benefit in almost every realistic scenario.

Evidence of the Moat

The financial record is unusually clean on this point. Copart has maintained ROIC consistently above 20% in every year of the past decade, averaging approximately 28% over the last ten years. Operating margins have ranged between 32% and 42% throughout this period. Gross margins have held in a narrow band around 45% for a decade. These are not the numbers of a business that competes on price. They are the numbers of a business that sets the standard.

The most revealing evidence of moat is what has happened to the competition. Insurance Auto Auctions (IAA), Copart's primary US competitor, was acquired by Ritchie Bros. Auctioneers in 2023 after years of underperformance relative to Copart. The combined entity has not threatened Copart's position in any meaningful way.

Moat Trajectory

The moat is strengthening, not weakening. Each additional facility Copart opens makes the network more valuable to insurance sellers because of improved geographic coverage. Each additional buyer who registers on the platform increases the competitive pressure on every auction. Each year of accumulated auction data improves the company’s IntelliSeller tool, which uses machine learning to help sellers optimise their pricing decisions. These are all self-reinforcing dynamics that become harder to replicate as they compound.

Competitive landscape

Copart’s meaningful competition is effectively limited to IAA (now part of Ritchie Bros.) in the US. This is a two-player market for insurance-company salvage vehicles at national scale. Local and regional operators exist but cannot replicate the global buyer network or the geographic coverage that national carriers require. The fact that this market has not attracted more serious competition in four decades is itself informative. The capital requirements are enormous, the zoning challenges are real, and the buyer network takes decades to build.

3. Financial Performance

Revenue Growth

Copart has grown revenue from $1.15 billion in fiscal 2015 to $4.65 billion in fiscal 2025, a compound annual growth rate of 15% over a decade. For a business operating in physical infrastructure with significant real-world constraints, sustaining that rate over ten years is exceptional.

The growth has been driven by four forces operating simultaneously: an increase in total loss frequency rates as vehicles have become more technologically complex and therefore more expensive to repair; market share gains at the expense of smaller regional competitors; geographic expansion both within the US and internationally; and an increase in revenue per transaction driven by higher average vehicle values and an expanded global buyer network.

In fiscal 2025, total revenues grew 9.7% year over year to $4.65 billion. Service revenues, the higher-quality, fee-based component, grew 11.4% to $3.97 billion. The recent moderation from earlier rates reflects the natural effect of a larger base and some normalisation following the elevated total-loss activity of the pandemic years.

Profitability

Copart’s profitability record is one of the clearest expressions of the structural strength of its competitive position. Operating margin in fiscal 2025 was 36.5%, above the ten-year average of 36.3% and meaningfully higher than the 30.1% the business generated a decade ago in fiscal 2015. The direction of travel matters as much as the level: a business that has expanded operating margins by 6.4 percentage points over ten years while growing revenue at 15% annually is not merely maintaining its competitive position, it is strengthening it. Operating income compounded at 17.3% annually from fiscal 2015 to fiscal 2025, growing from $344 million to around $1.70 billion. This outpaced revenue growth, which is precisely what margin expansion looks like in the financials.

Net income in fiscal 2025 was $1.55 billion, up 13.9% from $1.36 billion in fiscal 2024. The net margin of 33.4% is above the ten-year average of 29.6% and almost double the 19.2% the business earned in fiscal 2015. The steady expansion in net margin reflects operating leverage, the benefit of a larger and more efficient network, and the growing contribution of interest income, the company earned $178.9 million in net interest income in fiscal 2025 on its $4.79 billion net cash position, a meaningful source of earnings that did not exist a decade ago.

On earnings per share: diluted EPS grew from $0.21 in fiscal 2015 to $1.59 in fiscal 2025, a compound annual growth rate of 22.4%, ahead of both revenue and net income growth. This performance was bolstered by an overall reduction in diluted share count, which fell from approximately 1.05 billion in fiscal 2015 to 978 million in fiscal 2025. While the share count has drifted slightly higher since 2021 due to stock-based compensation, the long-term decline reflects the cumulative impact of Copart’s historical share buybacks (especially the aggressive buybacks between 2015-2017) and disciplined capital allocation.

Free Cash Flow

Free cash flow grew from $186 million in fiscal 2015 to $1.23 billion in fiscal 2025, a compound annual growth rate of 20.8% over a decade. That represents a 6.6-fold increase in the business’s capacity to generate cash, which is a meaningful claim about the quality of the underlying economics.

The capex-to-operating-cash-flow ratio tells a particularly revealing story about where the business stands today. Over the past decade, this ratio averaged approximately 44%, meaning Copart reinvested roughly 44 cents of every operating dollar back into the physical infrastructure that generates its competitive moat. In fiscal 2025, that ratio fell to 31.6%, its second lowest level in ten years. The business is indeed entering a higher-harvest phase in the US, but it remains an investment-heavy business globally.

Return on Invested Capital

Over the past decade, ROIC has averaged 28.4% and reached as high as 32% in fiscal 2022. In fiscal 2025, ROIC was 29%, essentially in line with the long-run average. This consistency is more impressive than a single peak figure. A business that has compounded at 15% annually in revenue while sustaining high returns on incremental capital is, by the standard definitions, a genuine compounder. Every dollar reinvested in the business has generated returns well above any reasonable estimate of the cost of capital in every year of the available record.

Balance Sheet

Copart’s balance sheet at July 31, 2025 was exceptional. The company held $2.78 billion in cash and restricted cash plus $2.01 billion in held-to-maturity securities, a combined liquid position of $4.79 billion against total financial debt of effectively zero. Total liabilities of $883 million were overwhelmingly operational in nature. This is a fortress balance sheet. It provides management with complete flexibility to act, whether to acquire land, weather a severe recession, respond to catastrophic weather events, or return capital to shareholders. The net cash position also generates meaningful interest income that did not exist in prior cycles, adding a new dimension to earnings quality.

Capital Expenditure

Copart spent $569 million on capex in fiscal 2025, the majority directed toward land acquisition and facility development. Importantly, the majority of Copart’s capex is growth-oriented, acquiring land and building new facilities that generate long-term structural barriers, rather than maintenance capex spent on preserving the existing asset base. This distinction matters when evaluating the quality of the free cash flow figure.

4. Growth Levers & Addressable Market

Copart is not a business that has exhausted its runway. The US network is maturing, but that maturation is precisely what frees capital and management attention for the next phase of growth. There are four distinct levers that can drive the business forward from here, and they are not speculative: each has evidence of traction today.

A) The structural Tailwind in Total Loss Frequency

The single most important driver of Copart’s domestic volume is total loss frequency, the percentage of accident-involved vehicles that insurers choose to declare a total loss rather than repair. In fiscal 2025, CEO Jeff Liaw described the full-year total loss frequency rate of 22.2% as an all-time annual high. This is not an anomaly. It is the result of forces that have been building for decades and show no sign of reversing.

Modern vehicles are fundamentally more expensive to repair than their predecessors. Advanced driver assistance systems, cameras, sensors, integrated infotainment, and complex structural materials mean that even moderate collision damage frequently triggers repair estimates that exceed the vehicle's value. Electric vehicles add a further dimension: EVs require approximately four additional labour hours per repair compared to internal combustion vehicles and carry roughly 30% higher repair costs. As EV penetration grows and as ADAS technology becomes standard across more model lines and price points, the economics of repair versus total loss continue to shift in Copart's favour. The average age of vehicles on US roads has also reached 12.8 years, meaning a large portion of the fleet consists of older vehicles where even modest damage tips the repair-versus-salvage calculation toward salvage. Each of these trends compounds the other, and none of them is cyclical.

B) International Expansion, the Majority of the Opportunity

The US salvage auction market (calculated based on the Auction Fees paid to the auction house) is estimated at $3.8 billion in 2025 and is projected to reach $7.2 billion by 2030, growing at around 13.6% CAGR. Copart’s domestic service revenue in fiscal 2025 was approximately $3.4 billion (including buyer fees, seller fees, transportation fees, and title processing) meaning the company is already capturing the largest share of a market that is still growing. The domestic runway is real but finite.

The international opportunity is a different order of magnitude. The global online salvage auction market was estimated at $12.4 billion in 2025 and is projected to reach $27.2 billion by 2030 at a 17% CAGR, a faster growth rate than the US market and from a base where Copart's penetration is a fraction of what it has achieved domestically. In fiscal 2025, international service revenue growth reached 18.9% for the full year, significantly outpacing the 10.4% growth in U.S. service revenue. The direction of travel is clear.

The specific international market opportunities are substantial. Germany's online salvage market was worth approximately $1.1 billion in 2025 and is projected to grow at a 21% CAGR through 2030, a market where Copart has been building infrastructure since 2017 and where the transition from a principal-based model to a consignment model is already delivering margin improvement. Brazil sits at approximately $480 million with mid-teens growth. India, which Copart briefly entered and then paused to wait for the market to develop further, is estimated at $230 million and growing at approximately 23% annually, a market that Copart is well-positioned to re-enter as formal insurance penetration and salvage regulation matures. In the UK, Copart already holds approximately 60–70% of the insurance-customer market, making it the dominant operator in Europe's most developed salvage market. Asia-Pacific is the fastest-growing region globally, expanding at a 16–24% CAGR depending on the source.

Copart's international buyer pool is a direct competitive advantage in these markets. The most recent available data indicates that international buyers purchasing US vehicles were acquiring vehicles significantly higher in value than comparable vehicles purchased by domestic US buyers, a reflection of the quality and purchasing power of the global member base Copart has assembled over four decades. That same buyer network can be directed toward inventory in Germany, Brazil, or any market Copart enters, providing an immediate advantage that a local competitor could not replicate. This is network effects working across geographies, not just within them.

C) Expanding Beyond Insurance, Blue Car and Adjacent Categories

Approximately 81% of Copart’s vehicle volume originates from insurance company sellers. That concentration is a risk, but it also reveals the size of the untapped opportunity in non-insurance channels.

Blue Car, Copart’s service offering aimed at banks, rental car companies, and fleet operators, delivered strong double-digit growth in fiscal 2025, as reported in the annual report. This is not a minor product line. Banks and financial institutions need to liquidate repossessed vehicles and lease maturities efficiently; fleet operators and rental companies cycle through vehicles on a regular cadence and need the same combination of geographic reach, global buyer access, and transparent pricing that insurance companies value. As Copart's platform becomes better known outside the insurance vertical, the addressable supply base expands materially without requiring any new infrastructure investment.

Purple Wave, Copart's heavy equipment and agricultural machinery auction platform, delivered continued growth in fiscal 2025. Heavy equipment and farm machinery represent a large and fragmented global market that benefits from exactly the same dynamics that made Copart successful in salvage vehicles: a global buyer pool competing for local inventory, transparent price discovery, and specialist services for sellers who lack efficient alternatives. This is early innings, but the model is proven.

D) The EV Transition as a Structural Accelerator

Electric vehicles are widely discussed as a long-term threat to Copart through the autonomous vehicle channel. The near-term reality is almost exactly the opposite. EVs are more expensive to repair than internal combustion vehicles by a meaningful margin. Battery damage, which occurs in a significant proportion of EV collisions, is frequently uneconomical to repair at all, either because the battery module is deeply integrated into the vehicle’s structure or because replacement costs are prohibitive. This drives EV total loss rates materially above ICE vehicle rates at comparable damage severity levels. As EV penetration of the vehicle fleet grows, still in its early stages globally, it adds a structural tailwind to Copart’s volume that did not exist a decade ago and that will intensify over the coming years before autonomous vehicle adoption becomes a countervailing force.

What the TAM Picture Tells us

A business that controls roughly half of the US domestic market, that is in the early stages of penetrating a $10.6 billion global market growing at 15% annually, and that is actively expanding into adjacent vehicle categories through Blue Car and heavy equipment, is not running out of room. The presence of a massive growth runway is undeniable; the real variable is Copart’s ability to scale operations and deploy capital with its trademark efficiency. Given its historical financial performance, the company has already proven it possesses the discipline to execute.

5. Management

Leadership and Tenure

Copart is effectively a founder-influenced business. Willis Johnson, who founded the company in 1982, remains Chairman of the Board. His operating philosophy, own the land, build the infrastructure, focus on seller returns, expand the buyer network, is deeply embedded in how the company operates.

Liaw joined in 2016 as CFO. He brought a sophisticated private equity background (formerly at TPG Capital) that sharpened the company's approach to capital allocation, ROIC, and international M&A. He served as Co-CEO with Jay Adair before becoming the sole CEO on April 1, 2024.

The transition from founder Willis Johnson to his son-in-law Jay Adair (now Executive Chairman), and then to Jeff Liaw, represents one of the most successful leadership successions in the industrial sector. The team's ownership mindset is reflected in their lack of a dividend; they prefer to reinvest every dollar into land or opportunistic buybacks.

Skin in the Game

Insiders, including Willis Johnson and his affiliates, collectively own a significant portion of Copart’s outstanding shares, exceeding 11% of shares outstanding as of the most recent proxy. Executive compensation is weighted heavily toward equity, tying financial outcomes directly to long-term share price performance. Options issued to senior executives include market conditions requiring the stock to trade above a specified price threshold before exercise, a feature that further aligns incentives with shareholder value creation.

Capital Allocation Track Record

Copart’s capital allocation record reflects a management team that has consistently prioritised long-term value over short-term return metrics. The primary deployment of free cash flow has been into land acquisition and facility development, investments that have generated ROIC averaging 28% over the past decade and that have deepened the structural barriers protecting the business.

Copart has historically been a reluctant repurchaser, preferring to build a massive cash pile for internal reinvestment. Aside from a defining $739 million buyback in 2017, the company remained largely dormant on this front until early 2026, when it deployed $500 million to take advantage of recent share price volatility. Combined with a total absence of dividends since its 1994 IPO, this reflects a management team strictly focused on long-term compounding over immediate distributions.

The primary reservation is the accumulation of $4.79 billion in liquid assets earning treasury rates when the business historically generates returns above 20% on invested capital. This is capital that could be working harder.

Communication and Transparency

Copart’s management communicates with shareholders in a manner that reflects genuine confidence in the business and a willingness to engage with difficult questions. Earnings calls are substantive. The risk factor disclosures in annual filings acknowledge real challenges rather than papering over them. The 10-K filings are unusually detailed about the operational mechanics of the business.

6. Valuation

Current valuation

Our valuation framework measures how much of the sum of a company’s future cash flows, discounted back to today at an appropriate rate, is already embedded in the current stock price. Rather than expressing this as a precise figure, we express it as an approximation in years.

At the time of this report, with the stock trading at $33.06, the market is pricing Copart at approximately 16 years of discounted future cash flows. This is an approximation, not an exact calculation. The assumptions embedded in our model are consistent with the company’s current and historical operating performance, they are not heroic, and they are not pessimistic. This figure is monitored and updated on a monthly basis as the stock price and business performance evolve. This places the business in the Attractive zone of the valuation gauge.

What 16 years means

What 16 years means in practical terms is this: the current price assumes that Copart will stop generating meaningful cash flows after year 16. Every year of cash generation beyond that point comes to you as an investor for free.

For a business with the structural characteristics described in this report, physical infrastructure that compounds in value, a buyer network that deepens with each passing year, contractual relationships with insurance companies that have persisted for decades, and a non-discretionary market that has existed since automobiles did, the question of whether Copart will still be generating substantial cash flows in year 17 and beyond seems like a reasonable bet.

Hold, Add, and Exit Logic

The valuation framework is not a buy-and-sell signal generator. It is a tool for thinking about price relative to value, and for making disciplined capital allocation decisions across the Bearhold Universe.

The current Attractive zone suggests the price does not demand excessive optimism about the future. For a business of this quality, that is a reasonable entry point.

The general logic across all five zones is as follows. In the Exceptionally Attractive and Attractive zones, the price is working in your favour, you are receiving more embedded future cash flows per dollar deployed than the market typically offers for a business of this quality. In the Hold zone, the business is fairly recognised, neither compelling to initiate a new position nor a reason to exit one already held. In the Expensive and Exceptionally Expensive zones, the price is embedding a level of optimism that the historical record does not automatically support.

My approach is straightforward: when a position moves into the Expensive zone, I sell and reallocate the capital to other quality businesses in the Bearhold Universe where the price sits in the Attractive and Exceptionally Attractive zones. The logic is simple, if you own a collection of exceptional businesses, your capital should always be working in the most attractively priced names available to you. Holding an Expensive position when an Attractive alternative exists is an opportunity cost that compounds against you over time. Discipline on the exit is as important as discipline on the entry.

Growth Engines

Every stock price is driven by one of two engines, or both simultaneously. Understanding which is working in your favour, and which might work against you, is as important as understanding the business itself.

The first engine is fundamental growth. Over time, a stock price tracks the growth of free cash flow per share. If a business compounds its FCF per share at 15% annually, that is roughly what the fundamental engine contributes to investor returns over a full holding period. It is steady, it is predictable, and it is the engine that quality businesses run on indefinitely.

The second engine is valuation re-rating. When a stock is mispriced, when the market is embedding fewer years of future cash flows than the business's quality and durability justify, the price tends to correct upward simply to reach fair value. This engine can produce returns that dwarf what fundamentals alone would deliver. It can also run in reverse: when a stock sits in the Expensive zone, the market is embedding too many years, and any mean reversion subtracts from returns even as the business continues to perform. A good business at the wrong price is still a poor investment.

For Copart at the time of writing, both engines appear to be working in the investor's favour. Over the past decade, Copart has compounded FCF per share at approximately 21% annually, one of the strongest rates among large businesses in any sector. That is the fundamental engine, running in full. The valuation engine has room to contribute as well: at 16 years of embedded cash flows in the Attractive zone, the market is not fully pricing the duration and quality of what Copart has built. An investor entering at this price is not relying on optimism, they are being paid by both the business performing and the price catching up.

The risk scenario is equally clear. If the stock moves into the Expensive zone, driven by price appreciation that outpaces fundamental growth, the valuation engine begins working against the position. In that scenario, the approach is to sell and reallocate capital to a quality business in the Bearhold Universe.

7. Risks

Risk 1 - Insurance Company Concentration

Copart obtains approximately 81% of its vehicle volume from insurance company sellers. While no single insurer accounts for more than 10% of consolidated revenues, the collective dependence on a relatively small number of large carriers is a genuine structural risk. If the major US insurers were to consolidate their salvage operations, pursue vertical integration, or develop a credible alternative platform, Copart’s supply would be at risk.

The probability of this materialising is low. Insurance companies are not in the business of operating salvage yards. The operational complexity, the land requirements, and the need to build a global buyer network from scratch are formidable disincentives. Moreover, the financial returns Copart delivers to its insurance partners, through higher auction prices enabled by its global buyer network, are difficult to replicate. An insurer attempting to internalise this function would almost certainly achieve worse financial outcomes than it does today working with Copart.

The risk is real but its probability of materialising is low, and the mechanism by which it would occur requires insurance companies to act against their own financial interests.

Risk 2 - Autonomous Vehicles and Declining Accident Rates

This is the risk that deserves the most honest treatment in any Copart analysis, and the one with the longest time horizon.

The thesis is straightforward: if autonomous vehicles eventually reduce accident rates materially, or eliminate them entirely, the supply of total-loss vehicles that feeds Copart’s business would decline in parallel. A 50% reduction in accident rates would, all else being equal, reduce Copart’s volume significantly.

There are several important qualifications a serious investor must hold simultaneously.

First, the timeline for meaningful autonomous vehicle adoption is deeply uncertain. The promises of full self-driving technology have been pushed out repeatedly over the past decade. Even optimistic estimates place widespread adoption of truly autonomous vehicles well into the 2030s or beyond in the United States.

Second, vehicle complexity has historically been a tailwind for total loss frequency, not a headwind. Modern vehicles, with advanced driver assistance systems, cameras, sensors, and complex structural materials, are more expensive to repair than older vehicles. This has driven total loss frequency higher over the past thirty years even as vehicle safety technology has improved. The transition period from human-driven to autonomous vehicles is likely to involve even more complex and expensive-to-repair vehicles, maintaining or increasing total loss frequency in the near and medium term.

Third, even in a world where US domestic accident rates eventually decline, Copart’s international business and its global buyer network provide a degree of diversification that pure domestic exposure would not.

The autonomous vehicle risk is real, horizon-dependent, and does not present itself as an imminent threat to the business. An investor with a ten-year horizon should have it on the watch list. An investor with a twenty-year horizon needs to weigh it more carefully as part of any thesis.

What Would Change my Mind?

Meaningful and sustained market share losses to IAA or a new entrant, over two or more consecutive years, would be an early warning signal that the moat is weakening. If total loss frequency begins a sustained multi-year decline attributable to measurable reductions in accident rates from driver assistance technology, not just a single year of mild weather, that would require a fundamental reassessment of the long-term volume outlook. If a major insurance carrier publicly announces plans to internalise salvage operations, that would warrant immediate re-evaluation.

8. The Verdict

Copart is the kind of business that takes decades to build and is almost impossible to replicate once built. It occupies a structural position in a non-discretionary market, behind barriers that compound in strength each year, run by a management team whose interests are aligned with long-term shareholders and whose operational record demonstrates consistent execution across three decades.

The two risks discussed, insurance concentration and autonomous vehicles, are not trivial. They are real, they deserve ongoing monitoring, and they prevent this from being a thesis where no adverse scenario is imaginable. But neither risk, assessed honestly against the current time horizon, overrides the fundamental quality of what has been built.

At a valuation of 16 years of embedded cash flows, the market is asking for reasonable performance from a business with an exceptional track record. That is a combination worth owning.

Copart is the kind of business that quietly makes you wealthy if you give it time and leave it alone.

Disclosure update: The author now holds a position in CPRT. The position was initiated after the original publication date of this report. All analysis and conclusions remain unchanged. This report reflects the author’s personal views and is not an investment advice. Investing carries the risk of permanent capital loss. Read the full disclaimer here

There is a specific kind of discouragement that hits investors with small accounts, and it has nothing to do with their actual performance.

It happens when they see someone else post a $50,000 gain. Or read about a fund that returned $200 million last year. Or watch a portfolio tracker tick upward by $800 on a good day and feel, despite knowing better, that $800 is not enough to matter.

This is the dollar sign trap. And it quietly destroys more good investors than bad decisions ever will.

The Comparison That is Not a Comparison

When you see a $50,000 gain, you are looking at an output. You have no idea what the input was. If that gain came from a $2 million account, it represents a 2.5% return, a number that trails inflation in most years and would be unremarkable by any professional standard. If your $800 gain came from a $5,000 account, it represents a 16% return. You outperformed every major index, most hedge funds, and the overwhelming majority of retail investors, and you felt bad about it.

This is not a minor perceptual error. It is a fundamental miscalibration of what investing success actually means.

Dollars are an output of two things: percentage returns and capital deployed. In the early stages of building wealth, you have limited control over the second variable. You have complete control over the first. The only rational scorecard, at any account size, is percentage return, full stop.

What Dollar Thinking Actually Costs You

The danger of anchoring on dollar amounts is not just psychological discomfort. It produces a specific and predictable chain of bad decisions.

When progress feels too slow because the balance is small, the instinct is to accelerate. To take more risk. To concentrate more heavily. To look for the one trade that closes the gap between where you are and where you feel you should be. This is the moment investors reach for leverage, pile into speculative positions, or abandon a disciplined process that was, by any honest measure, working.

The account that blows up is almost never the one that made bad decisions from the start. It is the account with a solid process that got abandoned because the investor compared their dollar returns to someone else’s and concluded, incorrectly, that they were falling behind.

Comparison is not motivating in investing. It is corrosive. The person you are comparing yourself to has different capital, different time horizons, different risk tolerance, and different life circumstances. Their dollar number tells you nothing useful about your performance.

Subscribe now

The Professional Standard

Consider what the investment management industry considers exceptional performance.

A fund that consistently returns 15% per year, net of fees, is considered elite. Institutions allocate billions to managers who deliver 12%. Warren Buffett’s long-run average at Berkshire Hathaway, widely considered the greatest investment track record in history, is approximately 20% per year compounded over decades.

These numbers are percentages. Nobody in the professional investment world measures manager quality in dollar terms, because dollar terms tell you nothing without knowing the size of the capital base. The entire infrastructure of institutional investing, benchmarks, Sharpe ratios, alpha, information ratios, is built around percentage returns precisely because they are the only comparable unit of measurement.

The retail investor who abandons a 15% annual process because the dollar amounts feel small is walking away from elite level performance because they forgot to look at the right number.

The Mathematics of Patience

The reason percentage thinking matters so much in the early years is that compounding is not linear. The returns do not feel significant when the capital base is small. They become significant when the capital base is large, and the capital base only becomes large if the percentage returns are preserved and compounded consistently over time.

A $10,000 account returning 15% per year becomes $163,000 in twenty years. The same account returning 15% per year but losing half its value once, because the investor abandoned the process and took a catastrophic risk, recovers to roughly $80,000. The single moment of abandoning discipline costs more than the entire first decade of compounding.

This is the mathematics the dollar sign hides from you. The impulse to chase a faster dollar return feels like ambition. In practice it is the most expensive thing an investor can do to their long term wealth.

The Only Comparison Worth Making

There is one comparison that is genuinely useful, and it has nothing to do with other investors.

Compare your percentage return to the market. If the index returned 10% and you returned 13%, you created alpha. You did something that most professionals, most algorithms, and most institutional capital failed to do. That is the only external benchmark that tells you anything real about the quality of your process.

Everything else, the dollar amounts other people made, the account sizes you see on social media, the feeling that your progress is too slow, is noise that will cost you if you let it reshape your decisions.

A small account with a sound process and a correct scorecard is not a problem to be solved. It is the beginning of something that compounds.

Subscribe now

Let’s Start with the comparison most people make, and watch it collapse under scrutiny.

Boston Dynamics Spot: 33 kilograms, 14 kilogram payload capacity, top speed of 1.6 meters per second, IP54 weather resistance. Price: $74,500. Unitree B2: 60 kilograms, 40 kilogram payload capacity when walking, top speed of 6 meters per second, nearly four times faster, IP67 weather resistance, which means it handles full water immersion, not just splashing. Price: approximately $25,000.

These are not a toy and a professional machine. They are both industrial quadruped robots designed for the same category of work: facility inspection, perimeter security, hazardous environment monitoring, emergency response. The B2 is heavier, faster, carries nearly three times the payload, has superior weather resistance, and costs one third as much.

This is the starting point that makes what follows so interesting. Because in a normal market, that comparison would end the conversation.

What a Normal Market Would Do

When a competitor offers a comparable product at one third the price, one of two things happens. Either the premium product justifies its cost through some combination of superior reliability, better software, stronger enterprise support, or a proven track record, and commands a loyal customer base willing to pay for that, or it loses market share until pricing corrects.

Boston Dynamics Spot does have genuine advantages. It has a longer deployment history, more third party integrations, better enterprise support infrastructure, and software autonomy that is more mature. These are real things worth paying for in mission critical applications. The 3 to 1 price ratio is not entirely irrational.

But here is what the normal market logic misses entirely: a significant portion of the customer base for these robots cannot buy the cheaper product. Not because they don’t want to. Not because the B2 is inferior for their purposes. But because the law prohibits it.

The Wall, Updated

The 2025 National Defense Authorization Act, Section 1078, mandated a study on the Department of Defense’s use of unmanned ground vehicles manufactured in China and established a mechanism by which a procurement ban could automatically spring into effect without further Congressional action. The FY2026 NDAA expanded the legislative architecture further, increasing defense spending to $855.7 billion for the Department of Defense and broadening the scope of restrictions on Chinese manufactured autonomous systems.

The legislation also extended to components. Effective June 2026, DoD is prohibited from procuring items from entities on the 1260H list, a list of companies with alleged ties to the Chinese military. By June 2027, the prohibition extends further: no items in the supply chain may contain components from those entities. The net is getting wider, and it is doing so automatically, with each annual defense bill adding another layer.

This is no longer a targeted measure aimed at a handful of named companies. It is becoming a structural feature of how the US government procures technology, a standing presumption against Chinese-origin hardware in sensitive applications, with the burden of proof reversed.

Meanwhile, in December 2025, Congress encouraged the DoD to designate Unitree, the manufacturer of the B2, as a Chinese military company. In its own filings with the Shanghai Stock Exchange, Unitree executives acknowledged receiving funding from PLA connected programs.

The wall is not a diplomatic position that might soften with a change of administration. It is becoming code. It is being written into procurement law, supply chain regulations, and component-level traceability requirements. Each NDAA adds another brick.

Subscribe now

The Mirror

China’s side of this equation is less visible to Western observers but operates on the same logic in reverse.

China’s defense procurement naturally flows to domestic manufacturers. State investment in robotics has accelerated dramatically, the government views humanoid and industrial robots as a strategic industry, not merely a commercial one. Chinese firms developing hardware for PLA programs receive preferential access to government contracts that American competitors simply cannot reach regardless of product quality.

The result is two parallel customer bases, each reserved for its domestic manufacturer. The US government’s $855 billion defense budget flows to companies that can clear American procurement requirements. China’s state budget flows to companies embedded in its own industrial ecosystem. Neither pot of money is accessible to the other side’s competitors.

This is what the price ratio actually reflects. It is not purely a quality gap. It is a market structure gap. American manufacturers are not competing against the best product in the world for government contracts, they are competing among themselves, for a buyer that cannot go elsewhere.

The Technological Consequences

This is where the long-term implications become genuinely serious, and where I think most analysis stops too early.

The first consequence is the loss of competitive pressure as a forcing function for improvement.

In open markets, you must be better than the best competitor in the world to win. That pressure produces a specific kind of discipline: constant benchmarking, aggressive iteration, an acute sensitivity to where you are falling behind. When legislation removes the most aggressive competitor from your addressable market, that pressure relaxes. The American robotics industry now competes primarily against itself for its largest contracts. self competition is less brutal than global competition. Over time, that matters.

The second consequence is diverging technical standards, and this one is underappreciated.

When two large ecosystems develop in isolation, they do not merely produce different products, they eventually produce incompatible ones. Communication protocols. Software interfaces. Sensor specifications. Navigation architectures. The longer the two ecosystems develop separately, the harder integration becomes, even in a hypothetical future where the political environment shifts. We have already seen this play out in telecommunications with Huawei. What took a decade to build in that industry is now structurally very difficult to unwind. Robotics is following the same trajectory.

The third consequence is a distortion in how we measure quality.

When a manufacturer’s primary customer cannot buy from anyone else regardless of price, quality, or performance, the normal feedback loop between product and market breaks down. Customers who cannot walk away are not the same as customers who choose to stay. The former will tell you what works within the constraints they have. The latter tell you what is genuinely best. A robotics industry that serves a captive government buyer loses access to the harshest and most honest signal the market provides.

The fourth consequence, and perhaps the most consequential over a decade, is what happens when the Chinese commercial market matures.

Right now, Chinese manufacturers can sell freely into the global commercial market while American manufacturers cannot access the Chinese market and Chinese manufacturers cannot access the American government market. If you are a Chinese robotics company with PLA funding, a cost structure shaped by domestic manufacturing economics, and access to every non aligned commercial market in the world, you are competing aggressively in the 80% of the market that is unrestricted. American manufacturers, by contrast, are competing fiercely in the government market and also in that same unrestricted commercial market, but without the structural advantage of a captive home buyer funding their fixed costs.

In other words, the American government market funds American manufacturers’ balance sheets. It also, subtly, insulates them from the most brutal form of commercial competition. That insulation may feel like strength. Over twenty years, it may prove to be fragility.

What this means for investors

I want to be precise here, because there is an easy mistake to make.

The legislative protection is real. It is durable over any reasonable investment horizon. A business that sells industrial inspection robots to the US government, and whose Chinese competitors are legally prohibited from doing the same, has a structural advantage worth paying for. That is not in dispute.

What is in dispute is whether that structural advantage is the same thing as a durable competitive moat. They are related but not identical.

A moat, properly defined, is an advantage that is self-reinforcing, one that gets harder to overcome as time passes. The legislative wall has some of that character, because each NDAA adds restrictions rather than removes them, and because supply chain compliance infrastructure becomes more embedded over time. But it also has a vulnerability that a pure product moat does not: it depends on a political environment that, while stable, is not guaranteed.

More importantly, it does not force the innovation that a pure product moat requires. A company protected by legislation does not have to be better than its Chinese competitor to win contracts. It only has to be compliant, reliable, and good enough. That is a different standard. And businesses that optimize for a different standard tend, over time, to produce a different quality of product.

The robotics companies I find most interesting in this environment are not necessarily the ones with the most government contracts. They are the ones that are winning in the unrestricted commercial market, where Chinese competitors are present, where price matters, and where the product has to earn its position on merit alone. Those companies are being sharpened by a harder benchmark.

A company that thrives only behind the wall is not the same as a company that thrives despite having a wall available.

A Note on Honesty

The robotics sector presents a recurring challenge for the kind of investing I try to practice. The structural dynamics I have described above are real and investable. But most robotics companies, even excellent ones, do not yet pass the quality filter I apply before any serious analysis: high certainty of being larger and more profitable in ten years.

The capital intensity is high. The competitive dynamics are still resolving. The technology is advancing faster than any individual company’s moat can calcify. And the geopolitical foundations of the structural advantage, while durable, are a different kind of durability than the kind produced by decades of customer switching costs or brand loyalty.

I am watching this sector carefully. I am not yet ready to put anything on the Approved List. When I am, you will know.

Subscribe now

I have been a student of the Adobe business model since their pivotal shift to the SaaS (Software as a Service) model in 2013. On paper, Adobe is the quintessential quality compounder. It is the kind of business that usually sits at the very top of my universe of names.

The fundamentals are, quite simply, a dream for any disciplined investor:

• Net Profit Margins: Consistently hovering around 30%.

• Cash Flow Dominance: A Free Cash Flow (FCF) margin above 40%, meaning the company generates more actual cash than its accounting profit suggests. This is rare and indicative of a capital light masterpiece.

• The Standard Moat: 90% of the world’s creative professionals use Photoshop. It isn’t just a tool; it is a language.

However, as we look at the landscape in 2026, a new variable has entered the equation: Generative AI, while the market is currently in a state of panic selling, the real risk isn't the noise, it’s the lack of structural visibility. Here is why I am hitting the pause button on Adobe, despite the attractive price tag.

1. The Casual User and the Erosion of the Low End Moat

Adobe’s moat has always been protected by a high wall of complexity. If you wanted professional results, you had to invest hundreds of hours into learning the Adobe suite. That learning curve created immense switching costs. You didn’t just buy the software; you invested your career in it.

AI is lowering that wall. For the casual or non professional user, the small business owner, the social media manager, the student, AI native tools like Canva (integrated with Magic Media) or Midjourney are now good enough.

We don’t yet know what percentage of Adobe’s 30 million+ subscribers are these casuals. If 20% of their base realizes they no longer need a $60/month Creative Cloud subscription because a $10 AI tool does 90% of what they need, the valuation of the entire company must be permanently rerated lower.

Subscribe now

2. The Threat of Seat Contraction and Workflow Deflation

In the institutional world, Adobe makes money by the seat. If an advertising agency has 100 designers, they pay for 100 licenses. This has been a steady, predictable stream of income for a decade.

But AI introduces workflow deflation. If AI integrated Firefly tools make a designer 50% more efficient, that agency faces a mathematical choice. They don’t necessarily produce 50% more work; they might just realize they only need 60 designers to produce the same output.

This leads to seat contraction. Even if Adobe attempts to raise the price per seat to capture some of that AI value, they are fighting a structural headwind they’ve never faced: A shrinking customer base within their own power users.

3. Is Firefly a Bridge or a Pier?

Adobe’s management argues that their proprietary AI, Firefly, is the solution. They believe that by embedding AI directly into Photoshop, they make their moat even deeper.

But here is the Risk Analysis view: Is Firefly a bridge to a more profitable future, or a pier that lead to nowhere? If AI becomes a commodity, where every software has a generate button, then Adobe loses its unique selling proposition. If the magic is in the AI model and not the software interface, then the value shifts from Adobe to the model providers (like OpenAI or Midjourney).

In this scenario, Adobe becomes a shell for someone else’s technology. Their 40% FCF margins would likely collapse as they pay massive compute fees to stay relevant.

4. Pricing Power vs. The Good Enough Substitution

A true compounding machine relies on the ability to raise prices without losing customers. Historically, Adobe did this with the confidence of a monopoly.

But substitution is the enemy of pricing power. In a world where specialized AI native startups are popping up every week, Adobe’s all in one bundle starts to look expensive. If a marketing team can get 80% of Adobe’s value from a suite of smaller, cheaper AI tools, Adobe loses the leverage to hike prices by 5–10% annually. Without those price hikes, the revenue growth story slows down significantly.

Investor vs. Gambler: The Certainty Gap

The vast majority of fund managers dumping $ADBE today are doing so out of a lack of technical understanding. They see a headline and sell. They are reacting to price, not value.

But there is a flip side. Buying the stock just because it is down 66% from its high without understanding the 10 year structural impact of AI isn’t value investing, it’s gambling.

My philosophy for Bearhold Research is built on the margin of safety. This margin isn’t just found in the stock price; it is found in the predictability of the business.

Adobe is one of the best businesses ever built. But being a great business isn’t enough to trigger a buy. I need to be able to model the next decade with high confidence. Right now, the AI Fog has reduced that visibility to near zero.

The Verdict: Why I am Waiting

I would rather miss a cheap stock that eventually recovers than own a business whose core competitive advantage is being structurally reshaped in ways I cannot quantify.

Adobe remains a magnificent machine, but until we see how seat contraction and casual substitution play out in the future, it stays on the watchlist, not in the Bearhold Universe.

Subscribe now

Let’s start with a simple question.

If someone offered to sell you a small business, a café, a laundromat, a car wash, how would you decide what it’s worth?

You’d probably ask how much money it makes. How reliably it makes it. Whether it’s likely to make more or less in the future. And then you’d figure out how much you’d be willing to pay today for that future stream of cash flow.

That’s it. That’s valuation. Everything else is just a more complicated version of that same question.

The problem is that somewhere between that simple café example and the modern financial industry, valuation became a discipline designed to produce precise answers to questions that don’t have precise answers. And in doing so, it gave investors false confidence in numbers that are largely fictional.

Let me show you what I mean.

The Standard Approach, and Why it Often Fails

The most widely taught method of valuing a business is the Discounted Cash Flow model, the DCF. You forecast the business’s future cash flows, apply a discount rate, and arrive at a “intrinsic value” that tells you whether the stock is cheap or expensive.

In theory it’s elegant. In practice it has a fatal flaw.

The terminal value.

In a standard DCF, the terminal value, the value assigned to all cash flows beyond your forecast period, typically accounts for 70 to 90 percent of the total calculated value. Seventy to ninety percent. Which means that most of the number you’re working so hard to calculate is just a disguised assumption about what happens in year five or six and beyond.

Change one assumption, the long-term growth rate, by a single percentage point, and your “intrinsic value” changes by 30, 40, sometimes 50 percent. The precision is an illusion. You’re not calculating the value of a business. You’re laundering your assumptions through a spreadsheet.

This doesn't mean DCF is useless. It means you have to use it honestly, as a framework for thinking, not a machine for producing answers.

A Better Way to Think About it

Here’s the mental model I use instead, and the one that sits at the heart of everything I write about on Bearhold Research

When you buy a stock, you’re buying a claim on a business’s future cash flows. The question isn’t “what is this business worth?” in some abstract sense. The question is: how many years of today’s free cash flow are already embedded in the current price?

Let me make that concrete.

Imagine a business generating $100 million in free cash flow this year. The stock is trading at a market capitalisation of $2 billion.

That’s 20 times free cash flow. Which means at today’s earnings power, the market is asking you to pay for 20 years of cash flows upfront.

Now ask yourself: is this business likely to still be generating at least this much free cash flow in 20 years? Is it likely to generate more? And what could go wrong?

If the business has genuine competitive advantages, a long reinvestment runway, and operates in a growing market, 20 years of cash flows might be a reasonable price to pay. If it’s a cyclical business in a declining industry with a weak balance sheet, 20 years of cash flows is terrifying.

The number of years embedded in the price tells you how much has to go right for the investment to work out. The higher the number, the less room for error.

Why This Framework Matters for Beginners

If you’re new to investing, here’s what to take from this.

Don’t let anyone convince you that valuation is too complex to understand. The fundamentals are simple: you’re buying future cash flows, and the price you pay determines how good a deal you’re getting. Everything else, the models, the ratios, the financial jargon, is just different ways of answering that same question.

Start by getting comfortable with one number: free cash flow. Not earnings, free cash flow. The actual cash the business generates after maintaining and growing its operations. That’s the number that matters. Everything else can be dressed up.

Subscribe now

Why this Framework Matters for Experienced Investors

If you’ve been investing for a while, here’s the harder truth.

Most of the time when you think you’re doing rigorous analysis, you’re actually doing elaborate rationalisation. You’ve decided you like a business, and you’re building a model that confirms it. The terminal value assumption you chose, the one that makes the DCF work, is the assumption that happens to justify the price you’re already willing to pay.

The years-of-cash-flows framework forces honesty because it starts with the current price and works backwards. Instead of asking “what is this worth?”, which invites you to construct a narrative, it asks “what does the current price assume?” That’s a much harder question to answer dishonestly.

A Word on The Simplicity of This Approach

Before we go further, an important clarification.

Dividing market capitalisation by free cash flow is a sanity check, not a complete valuation. It gives you a quick, honest starting point for understanding what the current price is assuming, but it deliberately ignores two factors that significantly affect the real answer.

The first is the growth rate of free cash flow. A business growing its free cash flow at 15% annually is worth considerably more than one generating the same cash flow today with no growth. If free cash flow is growing, the “years embedded in the price” calculation overstates how long it actually takes to justify the price, because each future year generates more cash than today.

A Practical Example

Let’s use a real business to make this tangible, Visa.

Visa generates extraordinary free cash flow. In its most recent fiscal year it produced roughly $19 billion in free cash flow on a market capitalisation of around $560 billion at the time of writing.

That’s approximately 29 times free cash flow. Nearly three decades of today’s earnings power embedded in the current price.

Is that expensive? It depends entirely on what you believe about Visa’s next 30 years.

If you believe Visa will continue displacing cash globally, growing its network, expanding into new payment categories, and compounding at high returns on capital, then 29 years of cash flows might be a reasonable price for that outcome. The business has structural tailwinds, a genuine network effect, and near-zero capital requirements.

If you think the payments landscape will be disrupted, margins will compress, or growth will slow, then 29 years of cash flows is a lot to pay for an uncertain outcome.

The framework doesn't tell you what to think. It tells you what you need to believe for the investment to make sense. That's its value.

What to Do With This?

Next time you’re looking at a stock, before you build a model or read an analyst report, do one thing:

Divide the market capitalisation by the free cash flow.

That number tells you how many years of today’s free cash flow the market is asking you to pay for. Then ask yourself honestly: does the quality of this business, its competitive position, and its growth runway justify that number?

If yes, you might have found something worth owning. If no. move on.

If you’re not sure, that’s where the real work begins.

Subscribe now

Nobody tells you this when you start investing.

They teach you how to read a balance sheet. How to calculate a PE ratio. How to build a discounted cash flow model. What they don’t teach you is what happens inside your head when the stock you just bought falls 30%, and you have to decide what to do next.

I learned that part the hard way. Several times.

The Big Name Trap

Early in my investing journey I bought a stock because it was cheap. Not cheap in the way a rigorous analyst means, not cheap relative to intrinsic value, not cheap relative to the quality of the business. Just cheap because the price had fallen and the name was big.

A well-known company with a recognisable brand. I figured the name alone was a kind of safety net. Big names don’t disappear, right?

What I missed was that the fundamentals were deteriorating. The price had fallen for a reason. I held on, not because my analysis supported it, but because I didn’t want to admit I’d made a mistake. The stock kept falling. I eventually sold at a loss, not because I had a clear reason to sell, but because I couldn’t take it anymore.

The lesson came later, quietly: price going down is not a reason to hold. Only the fundamentals are.

Waiting to Break Even

This one is more embarrassing to admit.

I bought a stock. It fell. I waited — not because I believed in the business, but because I needed the price to come back to where I bought it so I could sell without feeling like I’d lost.

Eventually it did. The price recovered. I sold immediately, relieved.

Then I watched the stock double over the next two years. It was the beginning of a bull market and I had just sold my position at breakeven because my ego needed to feel like it hadn’t made a mistake.

The stock didn’t know what I paid for it. The market doesn’t care about your entry price. But I was making decisions based on a number that only existed in my own head.

Subscribe now

News as noise

When I first started investing, every headline felt like information.

A disappointing earnings report. A negative analyst note. A macro scare. Each one triggered a response, sometimes panic, sometimes doubt, sometimes the urge to do something just to feel in control.

It took years to understand that most news is noise. That the market’s reaction to a headline is rarely proportional to the actual impact on a business’s long term value. That a great business reporting one bad quarter is almost never a reason to sell, and a falling stock price on heavy volume on a day when nothing fundamental has changed is almost never a reason to panic.

Learning to distinguish signal from noise is not an analytical skill. It’s a psychological one. And it takes time, usually measured in mistakes.

The loneliness of conviction

This one surprised me the most.

When you make a decision based on your own analysis and hold it through volatility, success feels strangely lonely. You start looking for affirmation. Did anyone else buy this? Are other investors talking about it? Is anyone else holding through this drawdown?

The crowd becomes a comfort blanket. If other people are in the same position, the decision feels safer, even if their reasons for owning it have nothing to do with yours.

But here’s what I eventually understood: the right decision has nothing to do with the crowd being with you.

In fact, the most interesting opportunities, the ones that compound the most over time, are almost always the ones where you’re making a decision the crowd hasn’t fully made yet. Waiting for consensus before acting is just a sophisticated way of buying high.

Conviction by definition means you've done the work independently. The crowd agreeing with you later is nice. It's just not relevant to whether you were right.

What Changed

None of this clicked overnight. It accumulated slowly, through losses, through missed opportunities, through regret, through watching positions I sold at breakeven turn into multi-baggers.

What eventually changed was simple: I stopped asking “what is the market telling me?” and started asking “what does the business tell me?”

The business doesn’t know what the stock price is. It doesn’t care about the macro environment. It just keeps doing what it does, compounding value if it’s genuinely exceptional, or slowly deteriorating if it isn’t. Your job as an investor is to understand the business well enough that the price movements stop feeling like information.

That’s the whole game. Everything else is just noise you have to learn to ignore.

Subscribe now

In 1957, Charles O’Reilly opened a single auto parts store in Springfield, Missouri. He had no venture capital, no grand vision of market domination, and no particular reason to believe he was building anything exceptional. He just wanted to sell car parts.

Sixty nine years later, that store is one of the greatest compounding businesses in America, and almost nobody outside of serious investing circles knows it exists.

O’Reilly Automotive operates over 6,300 stores across the US, Mexico, and Canada. But the store count misses the point. What O’Reilly has actually built is a logistics and service machine so efficient that it has achieved something almost impossible in retail: 33 consecutive years of comparable store sales growth. Through recessions, financial crises, a pandemic, and the supposed existential threat of Amazon and electric vehicles, the business just kept growing.

The auto parts business looks simple from the outside. Buy parts cheaply, sell them at a markup, repeat. But the real business is distribution, getting the right part to the right place at the right time. A mechanic waiting on a brake caliper isn't going to wait two days for an Amazon delivery. They need it in an hour. O'Reilly's dense distribution network, hub stores, superhubs, overnight lanes, means they can fulfill an extraordinary percentage of orders same day or next day. That's the moat. Not the brand, not the price, not the loyalty card. The speed.

Subscribe now

The numbers that matter:

• ROIC consistently above 20% since 2014.

• Gross margins above 50%

• Over $25 billion returned to shareholders through buybacks since 2011

• 33 consecutive years of comparable store sales growth

The average American vehicle is now 12.5 years old, the oldest fleet in recorded history. Older cars break down more. They need more maintenance. They require more parts. O’Reilly’s best customer isn’t someone buying a new car accessory, it’s a mechanic keeping a 2009 Honda Civic running for another three years. That customer is becoming more common, not less.

O’Reilly is the kind of business that makes you question what “exciting” means in investing. No viral product launches. No disruptive technology. No celebrity CEO. Just disciplined execution, a culture built around service, and a distribution network that gets better every year.

That’s what compounding actually looks like in practice.

I write about businesses like this every week, quality compounders that reward patience and punish hype. If this kind of thinking resonates, please subscribe below.

Subscribe now