When Is a Great Business Too Expensive?

Why quality is not a price

Whether a business is great and whether its stock is worth buying at today's price are two different questions, and the second one is where investors get hurt. Quality is a property of the business: the durability of its advantages, the returns it earns on capital, the cash it generates. Price is a property of this afternoon's market. Confusing them leads to the most expensive sentence in investing: it is a great company, so the price does not matter.

Buffett built the second half of his record on paying fair prices for wonderful businesses rather than wonderful prices for fair ones, and that framing cuts both ways. The wonderful business deserves a premium. The question is always how much premium, because at some price the mathematics of ownership stop working no matter what makes a business great.

Quality does provide a second layer of protection, since a compounding business grows into moderate overpayment over time; that idea is developed in margin of safety. But the operative word is moderate. No growth rate rescues every entry price, as the arithmetic below shows.

The two judgments even rest on different evidence. Quality shows up across a decade of filings: margins holding, returns on capital staying high, customers staying put while prices rise. Price adequacy shows up in a single comparison, today's multiple against the future the business can plausibly deliver. Investors happily spend forty hours on the first question and four minutes on the second, and the second is where the money is actually made or lost.

When is a great business too expensive by the numbers?

A great business becomes too expensive when even its best plausible decade produces a poor return from today's price. Run a clean hypothetical. A superb company earns $4 per share and will compound earnings at 12 percent a year for ten straight years, a feat few large companies manage. In year ten it earns $12.42 per share, and the market then values it at a healthy 25 times earnings, about $310 per share.

The business performs identically in every scenario below. The only thing that changes is what you paid on day one.

Hypothetical example; earnings grow 12 percent a year from $4.00 to $12.42, dividends ignored, figures rounded.

Read the bottom row slowly. The company executed flawlessly for a decade, tripling its earnings, and the buyer at 60 times earnings compounded at 2.6 percent a year, less than they would likely have earned in bonds. Nothing went wrong with the business. The price had simply collected the whole decade of success in advance, plus a premium the future had to refund.

That is the working definition of too expensive: not that the company will stumble, but that even the good outcome pays you badly.

What does a high multiple already assume?

A high multiple is a forecast wearing a number. Pay 60 times earnings and you are asserting that years of strong growth are close to certain; the market has written that scenario into the price already. Your return then depends on what happens relative to the expectation, a mindset unpacked in market expectations. Growth that merely matches the script earns you the script's modest arithmetic. Growth that misses it gets punished twice, once through earnings and once through the multiple.

History has run this experiment at scale. In the early 1970s, the Nifty Fifty, a set of premier American growth companies, became known as one-decision stocks: buy them at any price and hold forever. By 1972 many traded between 40 and 80 times earnings. Most of the underlying businesses kept right on growing, yet after the 1973 to 1974 crash, buyers from the peak waited years, in some cases well over a decade, just to break even. The companies were as great as advertised. The prices were not. That episode is a caution about entry prices, not a prediction of anyone's future returns, but the mechanism it demonstrates is permanent.

A useful reflex: whenever a multiple is high, translate it back into the growth it implies, then ask how many companies in history have delivered that. The list is always shorter than the market's enthusiasm suggests.

The double hit deserves numbers of its own. Suppose a stock trades at $240 because the market expects $4 of earnings next year, sixty times the forecast. The company grows, but only to $3.20 of earnings, a 20 percent miss. A market that paid 60 times for perfection now pays 40 times for very-good, and the price lands near $128, down 47 percent, in a year the business itself got bigger. Only the expectations shrank. That is what priced for perfection means in practice.

The Costco tension

The clearest live illustration of this whole tension is Costco, a business almost nobody criticizes at a price almost nobody finds comfortable. The quality case is real: membership renewal rates around 90 percent, deliberate low-margin pricing that widens the moat every year, decades of disciplined execution. Our full breakdown is in how we analyze Costco.

The tension is that the market has admired Costco for years and priced it accordingly, usually at a large premium to the broad market and to its own retail peers. That leaves an investor with exactly the trade-off this article is about. Wait for a cheap price and you may wait forever, missing a compounder because it never went on sale. Pay the premium and your return quietly depends on another decade of near-flawless execution that is already written into the price. We walk through that specific judgment, with current figures, in whether Costco is worth the premium.

Reasonable, disciplined investors land on both sides. What separates them from the crowd is that both sides can state the growth assumptions the price contains and have judged them plausible or not, rather than voting on the company's reputation.

Notice what the disagreement is not about. Neither side disputes the renewal rates, the cost discipline, or the durability of the model. The whole argument is the price of admission, which is exactly the separation of questions this article is urging. Whenever you catch yourself defending a multiple by praising the company, the conversation has changed subjects without telling you.

How do you decide without fooling yourself?

Write the forecast down before you pay for it. Take today's price, your required return, and compute what the business must be worth in ten years for the numbers to work; then ask whether that future is plausible rather than merely possible.

Required year-10 value = entry price x (1 + required return)^10
$240 x (1.10)^10 = $622

For the 60x buyer above to earn 10 percent a year, a $240 entry must become about $622. Even after ten years of 12 percent growth delivers $12.42 of earnings, $622 means the market of that day must still pay roughly 50 times earnings for a larger, slower company. Seeing the assumption written out is usually enough to judge it honestly. Comparing today's multiple against the company's own historical band, the method in relative valuation, supplies the context for how unusual the current premium is.

Then give the decision a form you can audit. Write the growth you assumed, the exit multiple you allowed, and the return that combination implies, before you commit. A year later, results and price will both have moved, and the written version lets you judge the bet you actually made rather than the one you remember making.

The decision this leaves is genuinely yours: how much of the future you will pay for in advance, and how much margin for disappointment you demand.

Where to go from here

Keep the two judgments separate, and make the second one explicit: decide what the business is worth to you, then compare the market's asking price against the growth it presumes. The flip side of this article, finding good businesses the market underestimates, is covered in when a stock is undervalued. To watch the tension in a live setting, open Costco's statistics tab and see what the market is currently asking you to believe.