What the Stock Market Isn't Telling You About Fair Value

Number goes up. That’s it. That’s the entire thesis for a generation of investors who’ve known nothing but ZIRP (zero interest-rate policy), stimulus checks and the mother of all liquidity waves.

And honestly? I get it.

When your first decade in the market is defined by V-shaped recoveries, you start to believe that valuation is just the annoying lecture that boring old grandpa gives before the fireworks. You were never shown a single discounted cash flow model, were never handed a copy of The Intelligent Investor, and were told instead to “buy the dip.”

That’s not an accident. The dip always gets bought, until one day… it never does.

Here’s the kicker: the tools for rational pricing have not disappeared. They’ve been ignored.

I’ve spent the better part of twenty years inside institutional asset management—estimating intrinsic value for pension capital, endowments and, on occasion, my own undiversified anxiety. The pattern I keep stumbling across is the same one you’d probably guess: retail portfolios are managed with price charts, hot tips, and a whole lot of hope. Meanwhile, the quiet, patient discipline of asking “what is this business actually worth?” has been pushed to the margins.

That’s a structural disadvantage. And it’s also, quite conveniently, fixable by about 5 pm this evening.

But first, we need to talk about why the gap between price and value keeps getting wider.

The Uncomfortable Question Nobody Wants to Model

Rational valuation models do one thing very well, in all their unglamorous glory: they force you to admit what you’re paying for.

Think about it. When you buy a stock at 40 times earnings, you’re not buying a company—you’re buying a promise, spread over roughly forty years, that profits keep growing at a rate the market currently imagines. Did you run those numbers when you hit the buy button? Most people didn’t. And it’s rare for them to have been asked to, either.

The question “what is this asset worth on its own merits?” is the founding question of the entire value tradition—Graham, Fisher, Buffett, Munger, and everyone who’s ever sharpened a pencil at a write-down. It has been asked with far too little frequency in recent years. In its place came momentum, narratives and the single most dangerous word in modern markets: because.

“Because AI.”
“Because rate cuts.”
“Because everyone else is buying it.”

Institutional investors aren’t much better, by the way. Their careers are measured against benchmarks, and the active managers who survive are those that perform a perfect impersonation of the index — but with slightly more expensive lunches. Rational valuation is acknowledged, padded around, then quietly forgotten by morning.

At the end of the day, we all know the machinery still works. Dividend discount models still click. DCFs still compute. The Shiller CAPE ratio is still sitting there like a well-organized lighthouse, projecting data on the cliffs while the boats below race toward the rocks.

Which is why this whole topic deserves a proper defense — not of the mathematics, mind you, but of the mindset.

Hypothesis: The Spreadsheet Beats the Story

Here’s my claim, plainly stated:

Rational valuation models don’t tell you when to buy. They tell you what you’re buying, and they nudge you away from the transactions that require tomorrow to be a miracle.

Back in 2008, one of my partners made a comment that has stuck with me for years. He said, “The models were right. They were just early. And early in a crisis feels identical to wrong.”

That’s the tragedy of rational valuation. It gets mocked during manias and mourned during panics, when in reality it was the one steadfastly reasonable voice at both parties.

So let’s be clear about the hypothesis before we dig into the evidence:

• Investors who anchor decisions to intrinsic-value estimates are rewarded over multi-year windows.
• The reward mostly arrives as loss avoidance — the overlooked cousin of profit.
• The degree of discipline matters more than the precise choice of model.

That last point is worth repeating, because I see people get bogged down in terminal value assumptions and WACC debates when simple, back-of-the-napkin math would do.

The Evidence: Starting Price Predicts Ending Destiny

The academic rubber hits the road when you look at cyclically adjusted price-to-earnings data, originally popularized by Robert Shiller and John Campbell in the late 1990s.

Shiller’s CAPE ratio smoothes earnings over ten years, adjusted for inflation, precisely to strip out the cyclical noise that makes single-year P/E ratios so unreliable. And what the long-run data shows is harder and harder to debate as the dataset grows.

Here’s the table that keeps me disciplined through bulls, bears, and everything in between:

Does that table look like investment advice? It’s not meant to be. It’s a menu of lenses.

But notice the deeper pattern. All four models ask the identical question in slightly different costumes: What am I paying, and what do I get back over time?

That question has been answered, time and again, with a century of data.

Historical analysis from Shiller’s team and numerous follow-up studies paints a fairly consistent picture: when the CAPE ratio has been below 15, subsequent 10-year real equity returns have tended to land in the high single digits to low double digits. When CAPE was stretched above 25, the following decade delivered something closer to a wash.

The relationship isn’t clockwork. It has failed to be precise in virtually every cycle. But the signal has been shown to be genuine, replicable and economically meaningful.

Compare that to the mania playbook. Nobody has found a reliable way to predict what narrative-driven retail money will do next week. Nobody has built a model that accurately prices momentum until the momentum stops. Zero.

But rational valuation models can tell you, right now, what the market is pricing in — and decide whether you want to live with that assumption.

Unpacking the Models (Without the Business-School Creep)

Let’s walk through the core logic of the most durable frameworks, because knowing why they work is what stops you from turning them into superstitions.

Dividend Discount Models: Small, Steady and Dismissed

Imagine you buy a stock for the sole purpose of collecting its dividends forever. Your return, logically, comes from two sources: the dividends you receive and the future sale price. In the DDM, the sale price at infinity shimmers into the past and you’re left with the present value of an endless stream of dividend cash.

It’s an elegant idea. The Gordon Growth Model, its most famous variant, is just:

Value = Dividend × (1 + Growth) ÷ (Required Return − Growth)

Crude? Yes. Useful? More than you’d imagine.

The beauty of the Gordon model is that it explodes when you demand more growth than returns can justify. That’s the point. When the required return is 9% and the assumed growth rate gets pushed to 9.5%, the denominator turns negative. The algebra is screaming something important: “stop assuming the unassumable.”

That lesson gets lost when everyone’s chasing companies that pay no dividend at all.

Discounted Cash Flow: The Workhorse With a Bad Reputation

DCF analysis is the model that undergraduate finance courses beat into you, knowing full well you’ll abandon it by the time you land your first internship.

The mechanics are straightforward, even if the practice is fiddly. You project free cash flow over five to ten years, discount it back to today using a required rate of return that compensates for risk, and then add a terminal value for everything beyond the forecast window.

The output is an intrinsic-value estimate. The inputs are judgment.

And here’s the uncomfortable truth most people ignore: a DCF model is only ever as good as the assumptions it’s fed. It has been said a thousand times that you can make a DCF justify any price you already want to justify.

That’s true, but it’s also the point. The model isn’t meant to give you certainty—it’s meant to make your assumptions visible. The act of building one forces you to ask how fast you think this business can genuinely compound, how wide the moat really is, and how much margin of safety you’re demanding as compensation for being wrong.

Most investors find that exercise considerably more difficult than opening a trading app.

Graham’s Ladder of Valuation

Benjamin Graham, the patron saint of value investing, spent most of his later career trying to simplify. He watched as Wall Street analysts built enormous, intricate models that collapsed in practice.

His response became the Graham formula:

Value = EPS × (8.5 + 2 × Expected Annual Growth)

It’s not subtle and it’s not sophisticated. But it has survived, and not just because it was written by a legend. It survives because it works as a screen. It guards you against the single most expensive error in all of investing: paying a premium multiple for a company whose growth is merely average.

Graham himself warned that formula was designed for moderate growth companies with predictable earnings. He never meant it to be applied to every ticker on your watchlist. But the instinct behind it—keep your required growth modest, and treat anything above that as speculation—is worth building an entire portfolio around.

The CAPE and Market-Level Valuation

Individual stocks get DCFs. Whole markets get CAPE.

The cyclically adjusted earnings multiple uses ten years of inflation-adjusted earnings, which smooths away the booms and busts of a single business cycle. The resulting ratio can be compared to historical averages that go back well over a century.

At the moment of writing, U.S. market CAPE readings have sat meaningfully above their historical median for extended stretches. History would suggest the following decade’s returns from such starting points have often been muted.

But again—and I cannot stress this enough—that does not mean sell everything. It means adjust expectations. It means demand a little more margin of safety on what you buy. It means overweighting the parts of your portfolio that don’t require a roaring bull market to succeed. It means shutting your ears when the cheerleaders tell you “this time is different.”

Because every time has been different. The differences have never — not once — made future returns more generous from expensive starting points.

Where the Models Fail (Take This Seriously)

Now, before you print out this article and glue it to your monitor, let me level with you about what rational valuation cannot do.

First, it cannot tell you the day, the week, or the quarter. The stock market, as the old industry joke goes, has predicted nine of the last five recessions. Likewise, your DCF will sit there quietly, correct and useless, while the insanity runs for another eighteen months.

The analyst’s joke about their own models is even darker: “The model is right — it’s the market that’s wrong.” Nobody laughs at these conferences because everyone in the room