Stocks for the Long Run Ch. 5: Valuation Matters — What the Starting Price Does to Your Return

Investment Background

The first four chapters may have left you with an impression: the long-run return on equities is 6.5%, and you need only hold long enough.

This chapter adds an important qualification — and it is the one qualification you can actually observe today.

The valuation level at which you buy has a material effect on your returns over the following ten to twenty years.

The 6.5% is an average across all starting points. Start from a cheap one and you receive more; start from an expensive one and you receive less.

The Wall Street Translation

The Relationship

The most common valuation tool is the cyclically adjusted price-to-earnings ratio, usually called CAPE, popularized by Robert Shiller.

It divides the current share price by the average inflation-adjusted earnings of the prior ten years.

Why a ten-year average? Because single-year earnings swing violently. In a recession year earnings collapse and the P/E looks extremely high — precisely when stocks are cheap. A ten-year average removes that distortion.

Historically, plotting starting CAPE against the subsequent decade's real annualized return produces a clear negative relationship:

Starting CAPE range	Subsequent 10-year real annualized return (historical range, approx.)
Below 10	8–15%
10–15	5–12%
15–20	3–9%
20–25	0–7%
Above 25	−3% to +5%

This is among the most robust long-horizon relationships in the historical data.

But Its Nature Must Be Stated Immediately

This is the most abusable point in the chapter, so the qualifications follow directly on the conclusion.

One, it has explanatory power over ten years and almost none over one to two.

A high CAPE does not tell you next year will fall. Markets can remain expensive for many years, and frequently do. Every approach that treats CAPE as a timing signal has performed poorly historically.

Two, it shifts a probability distribution; it does not produce a forecast.

A high CAPE means the entire distribution of ten-year returns shifts downward. It does not mean the outcome will be negative. Every row above is a range, not a point.

Three, the absolute level of CAPE drifts over time.

The modern average CAPE runs above that of the early twentieth century. Possible causes include accounting changes, lower transaction costs, broader market participation, and the interest rate environment. This means comparing today's CAPE directly against 1930's is problematic.

So the correct use is not "sell when CAPE exceeds 25," but "when CAPE exceeds 25, lower my planning return assumption."

This Chapter's Relationship to the Rest of the Library

The boundaries need drawing, because valuation is among the most-discussed topics here.

Book	What it does with valuation
The Intelligent Investor, Security Analysis	Select individual stocks — margin of safety, valuing a single business
Mastering the Market Cycle (Marks)	Locate the cycle — which end of the pendulum we occupy, and adjust posture
A Random Walk Down Wall Street	Essentially does not use it — it holds that valuation information is already in the price
This book	Adjust whole-market long-run planning assumptions — not selection, not timing, a planning parameter

The distinction is critical.

This chapter is not teaching you to judge the market high or low and move in and out. It is saying: when you type an expected return into a retirement calculator, that number should be informed by current valuations.

Marks teaches you to adjust your offensive-defensive posture. This book teaches you to adjust the number in the spreadsheet.

A Concrete Use

Let us turn it into an executable action.

Suppose you are doing thirty-year retirement planning and need an assumption for real equity returns.

The naive approach: use the historical average of 6.5%.

The better approach: look at where CAPE currently sits, then:

• CAPE at historical lows → use 7–8%, and you may still be conservative
• CAPE near the historical median → use 6–6.5%
• CAPE at historical highs → use 4–5%

Note the direction and magnitude of that adjustment. It does not liquidate you and it does not lever you. What it changes is how much you need to save, and how much you can safely withdraw.

That is the correct use of valuation information for a retirement investor.

Why This Is Far More Useful Than Timing

Because it operates on the variables you control.

You cannot control where the market goes next year. You can entirely control:

• Your savings rate
• Your retirement date
• Your withdrawal rate
• Your spending expectations

If current valuations imply lower returns over the next decade, the correct response is not to sell equities but to save a little more, or lower your assumed withdrawal rate.

That is an adjustment requiring no forecasting ability.

An Honest Note for Contemporary Readers

In presenting this framework, one thing must be conceded: for the past two decades and more, US market CAPE has sat persistently in its historically high range, and returns have nonetheless been substantial.

This has produced an unresolved debate:

One view holds that CAPE's mean has shifted upward structurally — lower costs, better corporate governance, higher corporate profit margins, and the greater weight of high-margin technology companies in the US market together justify a higher fair valuation.

The other view holds that mean reversion has merely been postponed, not cancelled.

The honest position is that we do not know which is right.

But that uncertainty itself points to the same practical conclusion: when valuations are high, lower your planning assumptions rather than change your holdings. If the first view is correct, you have merely over-saved — a survivable error. If the second is correct, you have avoided a serious planning shortfall.

This is an asymmetric choice, and under uncertainty you should take the asymmetrically favorable side.

Executable Trading Rules

1. Check CAPE once a year, and use it to set planning assumptions rather than trading decisions. Once a year is sufficient. If you find yourself checking monthly, you are already using it as a timing signal.

2. In a high-valuation environment, adjust the variables you control first. In priority order: raise the savings rate → lower the assumed withdrawal rate → consider delaying retirement one or two years → and only last, fine-tune allocation. The first three require no forecasting ability whatsoever.

3. Do not liquidate because valuations are high. Historically this is among the costliest errors available. Markets can keep rising from expensive levels for many years, and missing that compounding is usually more destructive than living through a decline.

4. Understand what dollar-cost averaging does within this framework. If you invest a fixed sum monthly, you automatically buy more shares when cheap and fewer when expensive. That mechanically offsets part of valuation risk, and requires no valuation judgment from you at all. It is also the deeper reason Chapter 6 of A Random Walk recommends the practice.

5. The correct response to high valuations is lowered expectations, not heightened vigilance. Psychologically these are very different. Lowering expectations is a planning action; heightened vigilance leads to frequent trading.

Relevance to a Retirement Portfolio

This is the most important chapter in the book for anyone approaching retirement.

The reason is sequence-of-returns risk: someone retiring at a high-valuation moment faces materially greater risk than someone retiring at a low-valuation one.

This aligns directly with Chapter 2 of Retirement Decumulation Mechanics — returns in the first few years of retirement have a disproportionate effect on the success of the whole thirty years.

Concretely:

If you are entering retirement at a clearly elevated valuation, consider:

• Setting the initial withdrawal rate at 3.5% rather than 4%
• Expanding the cash buffer from two years to three
• Using the rules in our Dynamic Withdrawal Guardrails tool so withdrawals adjust with the market

None of these is market timing. They add margin of safety at a starting point known to carry elevated risk.

And someone retiring at a low-valuation moment may reasonably use a higher withdrawal rate. That asymmetry is this chapter's most practically valuable output.

Chapter 6 takes on the book's most important question: the extent to which Siegel's central thesis may be wrong.