ValueScreener

Lesson 4 of 7

Naming your equity premium

Three routes, run today. The historical leg compounded at 10.5 percent since 1965; the implied leg lands near zero before a growth assumption is added.

Javier Sanz, founder

By Javier Sanz

Founder of ValueScreener. Built Ninety Nine, a securities services provider, and sold it to a tier-one global bank; ran operations and internationalization at Alpaca, the brokerage API.

Who writes thisEditorial processLinkedIn (opens in a new tab)X (opens in a new tab)

In one minute

  • The three routes to a premium are not alternatives to choose between on quality. They measure different things and disagree by design.
  • Run today, the implied route on the market's own median multiple lands near zero before any growth is assumed, and above two points after.
  • The deliverable is not a premium. It is a premium with its route, its date and its inputs written beside it.

An equity risk premium cannot be looked up, because it is not a fact about the world. It is an estimate, and the useful skill is producing one you can defend rather than finding the right one.

Where the market's own cost of capital sits

  • 10th percentile: 5.57%
  • 25th percentile: 6.51%
  • median: 8.26%
  • 75th percentile: 10.35%
  • 90th percentile: 12.98%
Estimated weighted average cost of capital across 2,746 US-listed common equities carrying one, read 2026-09-16.Source: The product's own stored estimates, built from a 10-year Treasury par yield stored 2026-09-15, measured betas, and an assumed 5.5 percent premium.

Route one: historical

Take the realised return on equities over a long window and subtract the return on government bonds over the same window.

The equity leg is readable from a published table. Berkshire Hathaway's 2025 annual report prints the S&P 500 with dividends included compounding at 10.5 percent a year from 1965 to 2025, against Berkshire's own 19.7 percent.

The bond leg over those same 61 years is not held here, so this route cannot be completed on this page. That is the honest state of it: a historical premium needs two series, and quoting the equity leg alone is not a premium. Whoever publishes one should be able to name both series and the window.

Route two: survey

Ask investors what return they require. It measures opinion, and opinion moves with the last three years of returns, which is the opposite of what a long-run input should do. No survey figure is quoted here because none was read for this lesson.

Route three: implied, computed today

The second row is the whole lesson. A no-growth version of the implied route says the market is priced at a premium of about nothing over government bonds today. Adding a 2.5 percent perpetual growth assumption moves it by more than two and a half points, which is larger than the premium itself.

That is not an argument against the implied route. It is the reason the route has to be quoted with its growth assumption attached, and the reason two analysts using "the implied premium" can be three points apart.

What this product does, and says it does

ValueScreener assumes 5.5 percent, applies it to every company, and labels it "our assumption" in the source trail of every figure it touches. That is higher than the implied number above and inside the range historical estimates usually produce.

The label is the point. A reader who disagrees can see exactly which input to replace, and the discounted cash flow workbench takes a rate directly rather than building one from a premium at all.

The deliverable

Write four things next to any rate you use: the route, the date, the inputs, and what you would have to see to change it. A premium without them cannot be audited a year later, including by you, and a rate nobody can audit is the part of a valuation that quietly does the most work.

Check yourself

Why can the historical route not be completed on this page?

  1. The equity return is unavailable
  2. The government bond return over the same window is not held here, and a premium needs both legs
  3. Historical premiums are no longer used

Show the answer

Answer: B.

The S&P 500 leg is printed in a published table. Without the matching bond series over the same 61 years, quoting it alone would not be a premium.

The implied premium moved from about zero to plus 2.53 points. What changed?

  1. The Treasury yield
  2. A 2.5 percent perpetual growth assumption was added to the earnings yield
  3. The cohort of companies

Show the answer

Answer: B.

The growth assumption is worth more than the premium itself, which is why an implied figure means nothing without it stated.

What makes a premium defensible?

  1. Using the highest of the three routes, for conservatism
  2. Stating the route, the date, the inputs, and what would change your mind
  3. Matching whatever the market consensus is

Show the answer

Answer: B.

No route is authoritative. What survives review is a number whose derivation someone else can follow and disagree with specifically.

Sources

  • Berkshire Hathaway 2025 annual report, the S&P 500 with dividends table: 10.5 percent compounded annually 1965 to 2025.
  • The product's public distribution endpoint: median price to earnings of 20.37 across 1,964 US companies, built 2026-09-15.
  • The product's stored reference rate: 10-year Treasury par yield of 5.00 percent, stored 2026-09-15.
  • Damodaran, Investment Valuation: the three estimation routes and the case for the implied premium as the forward-looking one.

What this leaves out

  • The implied calculation here uses a median trailing multiple as a stand-in for the market's aggregate forward earnings yield, which is a simplification, not the published method.
  • Cohort figures were built on 2026-09-15 and are rebuilt on a cadence.
  • No premium is recommended here, and the product's own 5.5 percent is an assumption rather than a measurement.
  • Nothing here is advice about any security.

The model, not your inputs

Lesson 4 of 7 of What a discount rate can be. It teaches how this part of a valuation is built. The inputs are still yours, the page supplies none of them, and arithmetic on invented numbers is a worked example rather than a value.

Free audit

Research tool, not investment advice. See our methodology for how the product reads a filing and what it refuses to compute.

Once a week, what changed in the filings of US-listed companies, why it can matter to a thesis, and what deserves a closer review. Grounded in filings, not market noise.