Same Multiple, Different Fate

There’s a particular kind of unease that shows up when a market keeps paying more for the same handful of names, quarter after quarter, and the usual valuation arguments stop seeming to matter. It isn’t new. The specific companies change. The feeling doesn’t.

In December 1972, a group of roughly fifty blue-chip American growth stocks — Coca-Cola, IBM, Polaroid, Xerox, McDonald’s, Disney, Avon — traded at an average price-to-earnings ratio of 41.9, more than double the S&P 500’s 18.9. These were called the Nifty Fifty, and the idea behind owning them was simple: these were such obviously dominant, permanent businesses that price barely mattered. You bought them once and held forever. A handful carried multiples north of 75 or even 90 times earnings — Polaroid alone traded near 95x.

What actually happened next

The popular version of this story is that the whole group crashed and the lesson is “extreme multiples always end badly.” The real data is more interesting, and more useful, than that.

During the 1973–74 bear market, the S&P 500 fell about 45%. The Nifty Fifty fell harder across the board: Polaroid down roughly 91%, Avon down 86%, Xerox down 71%, McDonald’s down nearly 70%, even Coca-Cola fell by roughly 70%. By that measure, the popular story looks right.

But look at what happened over the following decades, not just the crash. Coca-Cola entered 1972 trading at roughly 46 times earnings — an eye-watering multiple by any normal standard. It went on to deliver about 16.2% annualized returns in the years that followed, beating the S&P 500’s 12.7%. McDonald’s, despite a 1972 multiple north of 75, still compounded at roughly 10.5% annualized over the next three decades. Xerox, trading at almost exactly the same multiple as Coca-Cola in 1972, did not deliver anything close — it simply failed to sustain the growth the price assumed. Polaroid, the most expensive name in the group, eventually went bankrupt in 2001 and was wiped out entirely.

Run the math across the whole basket and Jeremy Siegel’s research found something genuinely counterintuitive: an equal-weighted portfolio of all fifty stocks, bought at the top of the mania in December 1972, returned about 12.2% annualized through 1998 — essentially matching the S&P 500’s 12.7% over that quarter-century. The group, as a group, wasn’t actually the disaster the popular telling suggests. It just wasn’t one story. It was two.

The multiple told you almost nothing

Here’s the part that matters most: Coca-Cola and Xerox entered 1972 at essentially the same price-to-earnings ratio. One became one of the great long-term compounders in market history. The other did not. The P/E ratio, on its own, could not distinguish between them. What separated them was something else entirely — whether the underlying business could actually sustain the growth rate the price was demanding, year after year, for decades.

That distinction wasn’t only visible in hindsight. It was a question of business quality, competitive durability, and whether a company’s advantage was structural or circumstantial — the kind of thing a careful look at the actual business, not the multiple, could have surfaced at the time. The investors who did fine out of the Nifty Fifty weren’t the ones who correctly timed when to sell before the crash. They were the ones who happened to own the businesses whose moats turned out to be real.

Where this leaves AI, honestly

I’m not going to tell you which of today’s richly priced AI and technology names are this era’s Coca-Cola and which are its Polaroid. I don’t know, and treating a historical parallel as a forecasting tool is exactly the kind of thing this Letter exists to avoid. The technology being real doesn’t tell you that; the multiple doesn’t tell you that either — it didn’t in 1972, and there’s no reason it would now.

What the Nifty Fifty actually offers isn’t a prediction. It’s a reminder of which question was always the right one to ask, and it wasn’t “is the overall group too expensive.” It was, one business at a time: does this specific company have a durable enough advantage to actually grow into the price being paid for it, or is the price simply assuming a version of the future that the business itself has no particular claim to delivering. That’s precisely the kind of question our own Business Quality and Moat Durability checklists are built to force — not because they predict outcomes, but because they make you actually answer the question instead of deferring to the group’s valuation as if it were information.

Why this belongs under Quality and Risk

The comfortable version of valuation discipline is a single rule applied to a whole category — “this sector is expensive, avoid it” or “this sector is justified, buy it.” The Nifty Fifty’s own numbers argue against that comfort. The group average told you almost nothing useful. The individual business, examined on its own terms, told you everything that mattered — and still only after the fact, once growth either showed up or didn’t.

I don’t know how today’s AI-era multiples resolve, for any individual name, and I’m not going to pretend otherwise. What I can do is keep asking the business-quality question one position at a time, rather than letting a basket-level multiple make the decision for me.

Further Reading

Two books that shaped the thinking in this Letter:

The Most Important Thing by Howard Marks — second-level thinking and why the crowd’s price is information, not a verdict.

The Intelligent Investor by Benjamin Graham — the original case for margin of safety over narrative.

As an Amazon Associate I earn from qualifying purchases. Book links in this Letter are affiliate links; we may earn a small commission if you buy through them.

Never Miss a Letter

Real portfolio thinking, delivered monthly to start. No forecasts, no noise.