The Tollbooth

The distinction matters more than almost anything else in long-duration investing, and it is less obvious than it appears. Many businesses that look like tollbooths are not. Many that don’t look like tollbooths are. The error in either direction is costly, and the direction of the error shifts by era: in expansions, investors overpay for the appearance of pricing power; in contractions, they sell the real thing at exactly the wrong moment.

What follows is an attempt to define the property precisely — not to catalogue which industries possess it, but to identify what it actually is.

What pricing power is not

It is worth clearing the ground first.

Pricing power is not a great product. Great products attract competition, and competition eventually compresses margins. The competitive advantage of quality is real but time-limited unless quality is continuously renewed — and renewal is a cost. The best consumer hardware companies make products that are genuinely superior. They also spend billions sustaining that superiority. The moat, such as it is, requires constant maintenance.

Pricing power is not a strong brand. Brands erode. They erode through neglect, through scandal, through substitution by younger alternatives with stronger cultural resonance. A brand that once commanded a 20% premium over generic may find, a decade later, that it commands 8% — not because anything catastrophic happened, but because the distance between the brand and its alternatives narrowed while no one was watching. Brand is preference. Preference can shift.

Pricing power is not a monopoly position, strictly speaking. A regulated utility is often a monopoly but rarely has pricing power in the meaningful sense — regulators cap the return on capital, which imposes its own ceiling. Pricing power and the absence of competition are related but distinct properties. The confusion between them is where many investment theses go wrong.

What it is

Pricing power is the structural ability to raise prices without proportional loss of volume. The key word is structural — meaning it derives from the customer’s position, not the firm’s marketing, and it persists through economic cycles, competitive challenges, and management changes.

The customer who cannot defect is the foundation. A customer who chooses not to defect — because the product is better, the brand is preferred, the service more convenient — is a different and weaker thing. Preference can shift. Inability is architectural.

The inability to defect comes from one of three sources. The first is switching costs: the accumulated integration between the supplier and the customer’s own operations has made replacement not merely inconvenient but economically painful. Enterprise software is the canonical example — the system embedded in every workflow, every finance process, every operational data structure cannot be replaced without years of disruption and eight figures of cost. The price increase that triggers migration must be enormous. In practice it almost never comes, because the supplier knows this and prices accordingly: high enough to extract value, not so high as to make the migration calculation tip.

The second source is a network effect that locks out alternatives. A payment network derives its value from ubiquity — every merchant accepts it because every consumer has it, and every consumer has it because every merchant accepts it. A challenger must simultaneously recruit both sides of the market. This is not impossible, but the capital and coordination required are so large that the incumbent can price at the upper edge of tolerable, decade after decade, without triggering the defection that would otherwise discipline it.

The third is the absence of a credible alternative entirely — a geographic, regulatory, or physical chokepoint where the road runs through one gate. The port that serves a hinterland. The exchange that lists the contracts everyone must hedge. The rating agency that issues a credential a borrower must obtain. Here, pricing power is structural in the most literal sense: the customer pays because there is nowhere else to go.

The compounding mechanic

Why does this matter so much for long-duration investing?

Because pricing power transforms the relationship between time and value. A business without it must grow volume to grow revenue. Volume growth requires capital, marketing, new products, expansion into new markets — each a competition, each a risk, each a source of potential disappointment. Volume growth is earned, contested, and impermanent.

A business with real pricing power can grow revenue without growing volume. It grows by raising prices — at inflation, above inflation, at whatever rate the market will bear without triggering mass defection. It does this without proportional additional capital expenditure. The revenue increment falls, disproportionately, to the bottom line. Over ten or twenty years, the difference between a business that must fight for every dollar of revenue and one that compounds from a standing position is not incremental. It is the difference between two entirely different categories of investment.

Combined with high returns on invested capital — which structural pricing power tends to produce — the arithmetic becomes extreme. Capital that earns high returns on incremental investment, deployed inside a business that can raise prices annually without volume loss, compounds in a way that is genuinely non-linear. These are the businesses that look expensive on current earnings and prove, over a decade, to have been cheap. The price paid at entry is the smallest input to the outcome. Time and pricing power do the rest.

The tollbooth

The metaphor earns its place because it captures the essential geometry. A tollbooth sits on the only road. It does not negotiate. It does not compete for your business. You either pay and continue, or you turn around. The toll-taker has no marketing department, no product roadmap, no concern with your satisfaction — only with whether you need to cross.

Most businesses are not tollbooths. They are shops on a street where other shops also exist. They compete on price, quality, service, convenience. The customer exercises genuine choice. Any of these businesses can be profitable, but none can compound indefinitely without effort. The shop must continuously earn its customers.

What we are looking for in the core of the portfolio is the closest available approximation to the tollbooth: the business where the customer has crossed the river, built their operations on the other side, and now cannot afford to go back. The business that sets its price as a first-order decision and watches the volume follow, because there is no credible alternative the volume could go to instead.

The risk that does not look like risk

Every moat looks permanent until it doesn’t. This is not a counsel of despair — genuine moats do persist for very long periods, and the businesses that possess them compound value in ways that justify extended patience. But the failure mode of tollbooth businesses is distinctive and worth naming precisely: they are not eroded. They are circumvented.

The payment incumbent does not lose to a competitor that does the same thing better. It loses to a technology that does something different and renders the old road less relevant. The enterprise software moat holds perfectly until a platform emerges that the enterprise wants to build on top of — at which point the question of replacing the old system becomes not whether but when. The chokepoint loses its power not when a rival emerges at the gate, but when someone builds a different gate entirely.

The discipline, then, is not monitoring the pricing power of the existing business — that is usually visible, and deterioration within the existing structure is usually gradual. The discipline is monitoring the alternative-building activity happening outside the castle wall. Who is constructing the new road? Where is it likely to emerge? How long does the incumbent have before the question becomes relevant to the customer’s calculus?

This is work that most conventional valuation analysis cannot do. The earnings model prices the current road. It does not price the probability of a new one. The investor who asks only whether the current multiple is justified is asking the wrong question — or rather, only part of the right one.

What this means for the portfolio

At the core of our structure sit businesses we have judged to be genuine tollbooths, or close enough that the distinction does not materially affect the expected outcome over a decade or more. This is intentionally a small set. The claim we make about each of them is theory-grade: this business will compound regardless of which economic regime prevails, because its economics do not depend on the regime. That is a strong claim. We hold it with proportional seriousness, revisit it at every new piece of evidence about the road outside the wall, and accept that no position in the core is permanent — only conditionally earned.

The standard for remaining in the core is the same standard that placed a business there: not historical performance, not size, not reputation, but the structural relationship between price and the customer’s practical ability to leave. When that relationship changes — when a credible alternative emerges, when switching costs erode, when the network loses its closure — the thesis changes with it.

Most investors encounter genuine tollbooths rarely. They are rare. When one is available at a price that does not already reflect a century of compounding, the right response is concentration, not diversification. The diversified portfolio hedges against uncertainty about which businesses are exceptional. The investor who has done the work to identify the tollbooth has resolved that uncertainty — and should size accordingly.

— Shash Hegde