Pat Dorsey

A moat protects attractive economics; a reinvestment runway gives those economics somewhere to compound; capital allocation decides how much of the result reaches each share.

The Label Is Not the Evidence

Pat Dorsey is the founder and a portfolio manager of Dorsey Asset Management. The SEC's adviser registry identifies the firm as an SEC-registered investment adviser. Dorsey's July 2025 firm presentation identifies him as Morningstar's former director of equity research and describes a concentrated global equity strategy for largely institutional clients.

Dorsey helped turn Warren Buffett's moat metaphor into a repeatable research vocabulary at Morningstar. But "wide moat" is a communicated judgment, not a physical barrier. The underlying condition must appear in customer behavior, contracts, production cost, network utility, scarce capacity, intellectual property, or another constraint that prevents competitors from eroding attractive returns.

Historical return on invested capital is evidence to investigate, not the cause. A high number can come from a cyclical peak, under-recorded intangible investment, inherited assets, or acquisition accounting. The analyst has to explain why the next unit of capital can earn more than its opportunity cost after rivals respond.

Four Sources Became Five

The publisher record for Dorsey's 2008 book shows chapters on intangible assets, switching costs, network effects, cost advantages, mistaken moats, erosion, valuation, and selling. That book organizes the structural sources into four groups. Dorsey's current presentation still uses those four and calls pricing power the primary moat test.

Morningstar's current methodology uses five sources, separating efficient scale as its own category. It assigns wide, narrow, or no moat and describes a wide advantage as one expected to last more than 20 years and a narrow one as at least 10. Morningstar owns and governs that present rating system. Dorsey helped develop the earlier framework; he does not personally assign today's ratings.

Pricing power also needs interpretation. A cost leader may share savings with customers and preserve volume rather than raise price. A regulated license may restrict entry while capping price. The deeper test is whether the structure preserves attractive incremental economics against substitution and entry.

A Moat Needs a Reinvestment Runway

Dorsey's July 2025 presentation compares two stylized companies earning 20% on capital. One reinvests 30% of cash flow and allocates the other 70% through dividends, repurchases, or acquisitions. The other reinvests 70%, assuming enough opportunities exist. More capital then remains inside the known high-return operation instead of returning to an investor who must find another investment.

The arithmetic is straightforward; the assumptions are not. Total historical return is not automatically the return on the next customer, store, salesperson, factory, or acquisition. A mature installed base may be profitable while expansion requires discounts or expensive capacity. A large addressable market is not a runway unless the company can reach it with repeatable unit economics.

Some reinvestment is recorded on the income statement. Software development, sales hiring, and advertising may create benefits beyond the current period even when accounting rules expense them. That does not make every expense an asset. Customer cohorts, retention, payback periods, price realization, incremental margin, and cash conversion help separate funded growth from maintenance or waste.

The operational questions change by moat source. For switching costs, customers must describe the data migration, retraining, workflow disruption, and failure risk of leaving. For a network, utility must increase with participation and resist multi-homing. For a brand, price or volume must show that reputation changes behavior. For cost advantage, the saving needs a durable source such as process, location, purchasing, distribution, or scale.

Capital Allocation Transmits Business Value

As internal opportunities decline, the chief executive and board control more cash than the core business can productively absorb. Dorsey's presentation divides the choices into reinvestment, returning capital, and acquisitions. It argues for internal reinvestment when high-return opportunities remain and returning cash when they do not. It warns that passive repurchases can destroy value at excessive prices and that large, infrequent acquisitions often conceal strategic weakness.

The money moves through an organization. Operating teams request budgets; finance classifies spending; executives compare projects; the board authorizes major transactions; lenders and covenants restrict some choices. Shareholders receive what remains after wages, suppliers, tax, interest, compensation, acquisition consideration, and dilution. A strong product moat cannot guarantee a growing claim per share.

Management therefore matters even when it did not create the original barrier. Incentives determine whether executives widen the moat, protect a reported margin, pursue empire size, issue shares cheaply, or repurchase them expensively. An analyst needs both the business structure and the people authorized to allocate its cash.

Qualitative Work Tests the Numbers

Dorsey's firm says its process emphasizes primary research and qualitative insight. The presentation's practical claim is that a screen cannot reveal switching costs; customers have to explain the value proposition and the burden of changing. Likewise, a spreadsheet cannot prove that reinvestment is positive-value or that management will allocate cash rationally.

That does not license storytelling. Interviews are sampled, sometimes promotional, and vulnerable to confirmation bias. Customer statements should be checked against renewal, churn, price, usage, implementation time, and competitor wins. Management's claimed runway should be checked against cash consumption and returns on new cohorts. The recorded condition and the communicated claim need to converge.

Moats can erode through technology, standards, regulation, multi-homing, patent expiry, customer concentration, subsidies, or new capacity. Dorsey said in a current Morningstar interview that network effects are often overestimated. A network is not durable merely because many users are present; participants must receive increasing value and face a credible reason not to recreate the interaction elsewhere.

The Firm Record Is Not a Personal Return

As of July 1, 2025, Dorsey Asset Management described approximately $1.37 billion under management, seven employees, twelve positions, and 59% of capital in its five largest positions. Those are dated marketing disclosures about the firm and strategy. They are not an audited performance series.

The firm's SEC holdings filings are also incomplete for performance analysis. Form 13F is a delayed snapshot of qualifying US-listed holdings. It omits cash, many foreign securities, some derivatives, fees, transaction dates, client restrictions, and account differences. Market value in a filing does not reveal purchase price or return.

No complete audited personal or firm composite return for Dorsey appears in the public sources used here. Assets under management, book sales, rating influence, and concentrated holdings are not substitutes. A client result would belong to a defined account or composite after fees and cash flows and would reflect the full investment team, execution, currency, restrictions, and investee-company managers.

Using the Framework Without the Fund

The usable sequence is demanding. Normalize return on capital. Identify a causal barrier. Test it with customers, competitors, records, and unit economics. Estimate the amount and duration of attractive incremental reinvestment. Separate maintenance from growth spending. Inspect incentives and cash allocation. Translate enterprise gains into per-share value after debt and dilution. Then compare the uncertain result with the price.

Dorsey Asset Management's concentration, institutional access, research staff, fee arrangements, compliance capacity, and tolerance for tracking error do not transfer with the framework. Nor does a moat eliminate valuation risk. A durable company can produce a poor shareholder result when the runway is shorter than expected, allocation destroys value, or the purchase price already capitalizes an ideal future.

Inside CompanyGraph

CompanyGraph tracks the quality-compounder print live: companies whose earnings arrive backed by operating cash while revenue growth and cash margins read elevated together.