Skip to content
4 min readBy Dominic Roe · Data Engineer & Business Intelligence Developer

What Is a Good CAPE Ratio?

There's no single 'good' CAPE ratio — it depends on the market and the era. Here are the rough guideposts, why they shift, and the better way to judge any reading.

There is no single "good" CAPE ratio. As a rough guide, the long-run average across major markets is around 16–17, so readings in the mid-teens have historically been reasonable-to-cheap, and readings of 30 or more have been historically expensive. But that rule of thumb hides more than it reveals — what counts as a "good" CAPE depends heavily on which market you are looking at and which era. The only reliable way to judge any reading is against that market's own history.

(New to the metric itself? Start with what is the CAPE ratio? — this post is about how to interpret the level.)

The rough rule of thumb

If you want ballpark guideposts, history gives you these:

CAPE level Rough historical read
Below ~12 Cheap — often a generational buying point
~13–18 Around the long-run average
~19–25 Somewhat expensive
~26–33 Expensive
Above ~33 Very expensive — rare, seen near major peaks

For context, the US market's CAPE has averaged about 17.4 since 1881 (median 16.1), and has ranged from under 5 to a peak around 44. The blended Global CAPE has averaged close to 16.8. So a CAPE in the teens is genuinely middling; the high-20s and beyond is the expensive end of the historical range.

Why there is no single "good" number

Two things make a universal threshold misleading:

Different markets have different "normal" CAPEs. The US habitually trades at a higher CAPE than the UK, Europe or emerging markets — it holds more high-growth, high-margin technology businesses, which command richer multiples. A CAPE of 16 is unremarkable for the UK but would be extraordinarily cheap for the US. (More on this in global vs country CAPE.) You can see the spread across markets on the country comparison.

The era matters. Interest rates, profit margins, accounting rules and index composition all drift over decades, and they shift what "normal" looks like. The US CAPE has averaged meaningfully higher over the last 30 years than over the previous century. That doesn't mean "this time is different" — but it does mean a fixed line in the sand is the wrong tool.

The better question: where does it sit in its own history?

Instead of asking "is this a good CAPE?", ask "where does this reading sit within this market's own record?" That is what a percentile tells you. A market in the 90th percentile has been more expensive only 10% of the time; one in the 20th percentile is cheaper than it has been four years in five. (See how to read valuation percentiles.)

This sidesteps the cross-market problem entirely: a UK CAPE and a US CAPE can't be compared directly, but their percentiles can. Every market on this site is scored exactly this way — against itself.

What the numbers look like today

As of mid-2026, the guideposts above put things in stark relief:

  • The US CAPE sits around 41 — its 99th percentile, more than double its own long-run average of ~17. By any historical standard, that is very expensive. (Is the S&P 500 overvalued?)
  • The Global CAPE is around 29 — its 97th percentile, against an average near 17. (Is the stock market overvalued?)
  • Several markets abroad still trade in the teens — closer to, or below, their own norms.

What a "good" CAPE means for returns

The reason a low CAPE is "good" for a buyer: historically, the lower the starting CAPE, the higher the subsequent long-run return — and vice versa. Buying when a market sits well below its own average has, on average, paid off over the following decade; buying at the top of the range has not. You can explore that relationship, and the considerable uncertainty around it, on the expected returns page.

As always, CAPE is one input among several, and it is not a market-timing signal — a "bad" (high) CAPE can persist for years. Judge it against a market's own history, alongside other gauges, and use it to set expectations rather than to predict turning points.

See where things stand

CAPEvaluationexplainer