Why a Strong Moat Doesn't Excuse Paying Any Price

A wide moat is a claim about durability. It is not a claim about price. Those are two different judgments, made with two different tools, and the instinct to collapse them into one — "it's a great business, so just buy it" — is where a lot of otherwise careful investors quietly overpay for years without noticing they've done it.

The instinct isn't crazy. It comes from a true observation: cheap, no-moat businesses are frequently cheap for a reason, and chasing a low multiple into a structurally weak business is a well-worn way to lose money slowly. But the correction for that mistake shouldn't be "therefore price doesn't matter for a wide-moat business." It should be "therefore judge the moat and the price as two separate questions" — and a strong moat only answers the first one. What would it take to displace this company is a real question with a real answer. What you should pay for owning it anyway is a completely different calculation, and a genuine moat doesn't exempt you from running it.

The historical record already ran this experiment more than once, and one case is unusually clean because almost nothing about the underlying business — its competitive position, its customer relationships, its balance sheet — was actually in dispute at the time. The moat was real. The price was the problem. And the gap between those two facts cost investors a quarter of a century.

The Premium Instinct, and Where It Breaks

Paying up for quality is defensible reasoning almost everywhere else in life. A better-built house costs more and is usually worth it. A more reliable car costs more and usually saves you money on repairs later. The mental shortcut generalizes naturally to stocks: identify the best business, pay whatever it takes to own a piece of it, let compounding do the rest. For a narrow band of prices, that shortcut works fine.

It stops working the moment "whatever it takes" gets treated as a price-agnostic instruction rather than a price-sensitive one. A moat tells you how long a stream of earnings is likely to survive mostly intact — the switching costs, the network, the license, the brand that keeps a competitor from just showing up and taking share. It says nothing about how many of those future years the current price has already assumed. Pay for twenty years of durable earnings on a business that can plausibly deliver twenty, and the price is doing honest work. Pay for forty on the same business, and you're financing somebody else's optimism with your own capital, regardless of how real the underlying moat happens to be.

This isn't an argument that quality investors are making a rookie mistake. Most of them know, in the abstract, that price matters. The failure mode is subtler than that: a genuinely strong moat makes it easy to keep telling yourself the story is still intact — the switching costs are still there, the brand still resonates, the competitor still hasn't shown up — right through a period when the only thing that actually changed was the price you'd have to pay to own it. The moat gives you a reason to stay convinced. It doesn't give you a reason to stop checking what conviction is currently costing.

Cisco Systems: A Real Moat, a Twenty-Five-Year Price

What the Market Paid For

Cisco Systems closed at $80.06 on March 27, 2000, briefly making it the most valuable company in the world, with a market capitalization north of $500 billion. Cisco's fiscal 2000 10-K, covering the year ended July 29, 2000, reported net income of $2.67 billion on revenue of $18.93 billion. Run the arithmetic and the market was pricing Cisco at something in the neighborhood of 190 times trailing earnings and roughly 26 times revenue. Nobody was confused about what Cisco did. It made the routers and switches that were, quite literally, the plumbing of the internet buildout, and it held a dominant, durable economic moat in that market by any reasonable definition — switching costs baked into enterprise networks, scale advantages in R&D and distribution, a brand synonymous with "the internet works." Real moat. Not a story anyone made up.

What wasn't accurate was the price. A business can be exactly as good as its admirers believe and still be a bad investment, because "good" and "worth 190 times earnings" are unrelated claims that happened to get bundled together during a period when almost nobody wanted to separate them.

What Didn't Break

Here's the part that makes the case genuinely useful rather than just a cautionary headline: the moat mostly held. Cisco's stock fell roughly 88% from its peak to under $10 by 2002, and its fiscal 2002 10-K, covering the year ended July 27, 2002, reported pro forma net income of $2.9 billion — essentially in line with the $2.67 billion it had earned two years earlier at the top of the bubble. The business that investors were paying 190 times earnings for in March 2000 was, by the numbers, roughly the same size two years later. It didn't collapse. It didn't get disrupted by a scrappier competitor with a better switching chip. The moat was doing exactly what a moat is supposed to do — protecting a profitable, durable business — while the stock lost nearly nine-tenths of its value anyway, because none of that duration had anything to do with what the price had already assumed.

The stock didn't close above that March 2000 high again until December 2025. Twenty-five years, on a business that was never structurally broken, run by a management team that never lost the plot, sitting on top of a competitive position that, by most reasonable moat classifications, never stopped being wide. I'll admit that's a longer stretch than almost any other case in this category — most overpriced-quality stories resolve faster, in five or ten years rather than twenty-five — but the mechanism is the same one that shows up in the milder cases. It's just easier to see clearly here because so little else about the underlying business changed along the way.

Duration Still Has a Price

Every dollar of future earnings a moat protects is still a dollar you're buying today, and the price of that dollar depends on how many other dollars are lined up in front of it, waiting to be discounted. A wide moat extends the visible runway of a business's earnings. It doesn't extend it infinitely, and it certainly doesn't make the price you pay for that runway irrelevant. Two businesses with identical, genuinely wide moats can be a good investment and a bad one simultaneously, purely as a function of what an investor paid to own each — this isn't a hypothetical, it's what happened to two cohorts of Cisco shareholders separated by nothing more than which March they bought.

This is exactly the distinction our Quality × Valuation grid is built to keep visible instead of collapsing into a single buy-or-don't verdict. Moat strength and price sit on separate axes on purpose, and a companion piece on judging moat quality without leaning on a valuation model goes deeper into why those two judgments need to happen in a specific order — moat first, price second — rather than getting blended into one gut call. Our earlier piece on paying a premium for durable moats made the case that a persistent premium is often the correct price for genuine duration. Cisco is the other half of that same argument: a premium stops being a duration bet and starts being pure speculation once it's pricing in more years of dominance than any moat, however wide, can plausibly deliver.

I'm less confident than I'd like to be about exactly where that line sits in real time — it's a great deal easier to draw it looking back at 2000 than it would have been sitting in a brokerage account watching the stock keep going up through early March of that year. Nobody rings a bell at 190 times earnings and tells you to stop. That's the honest, uncomfortable part of this whole exercise.

Cisco is the cleanest version of this pattern because so little else about the business changed along the way, but it isn't the only one. Coca-Cola's investors lived through a milder version of the same mechanics after 1998, when the stock traded near 50 times trailing earnings on the strength of a brand and a bottling network nobody seriously disputed was a real moat — and then spent a long stretch of years going essentially nowhere while the underlying business kept growing earnings the whole time. Different industry, different decade, same root cause: the moat wasn't the mistake. The price paid for it was.

A Practical Check Before You Pay Up

Short of building a full fair value model every time, a few questions catch most of the damage before it happens.

• Ask what growth rate, sustained for how many years, the current price actually requires — then ask honestly whether this specific business, with this specific moat, has ever grown at that rate for that long.
• Compare today's multiple against the business's own trading history, not against the sector average; a moat doesn't announce itself getting more durable just because the multiple expanded.
• Separate "the story is true" from "the price is fair." Cisco's internet-buildout story in 1999 was true. That had nothing to do with whether $80 a share was a fair price for it.
• Watch for anyone — including yourself — treating moat quality as a reason to stop asking about price altogether. That's usually the moment the two questions have already been quietly merged into one bad one.

None of this argues for treating every rich multiple with suspicion. A genuinely durable business can sustain a real premium indefinitely, and plenty of wide-moat companies bought at fair-to-generous prices have gone on to compound for decades without incident. The argument is narrower than that, and it's the same one whether the business in question makes routers or memberships or search results: moat quality tells you the earnings are likely to keep showing up. It was never designed to tell you what those earnings are worth today, and treating it as though it does is how a genuinely wide moat still manages to produce a genuinely bad investment.

Key Takeaways

• A moat measures durability, not price. Cisco Systems held a genuinely wide moat throughout the dot-com collapse and the two decades that followed — the moat was never the problem.
• Cisco closed at $80.06 on March 27, 2000, near 190 times its fiscal 2000 net income of $2.67 billion (year ended July 29, 2000). It fell roughly 88% to under $10 by 2002, even as pro forma fiscal 2002 net income of $2.9 billion (year ended July 27, 2002) held roughly flat versus 2000 — the business barely wobbled while the stock did almost all the damage.
• The stock didn't close above its March 2000 high again until December 2025 — twenty-five years for a business that, by most measures, never stopped being wide-moat.
• Judge moat quality and price as two separate questions, in that order. A wide moat justifies paying a premium. It doesn't justify paying any premium, and the difference between those two ideas is where a lot of quiet, multi-decade underperformance actually starts.