China CAPE Ratio
Current China CAPE
As of August 2026 · ratio
Rough estimate. This single-country CAPE is approximated by rolling the FXI ETF price from a recent anchor — free data has no per-country 10-year earnings — so treat it as a directional gauge, not a measured CAPE. It is bounded and is not used in the headline Global CAPE. See the methodology.
By Dominic Roe · Data Engineer & Business Intelligence Developer Updated August 2026
Is China cheap right now?
✓ Real · FXIOn 4 of 4 trailing multiples, China equities are priced below the all-world index (VT). These are current, fund-reported figures — not the estimated CAPE series charted below.
| Measure | China | World |
|---|---|---|
| P/E | 9.9 | 22.6 |
| P/B | 1.13 | 3.32 |
| P/S | 1.32 | 2.55 |
| Dividend yield | 1.87% | 1.59% |
| ROE (quality) | 11.4% | 14.7% |
Green = cheaper than the world index on that measure. ROE is shown as context, not cheapness: a lower return on equity means part of a cheap multiple may be deserved rather than a bargain.
Against its own history, China's estimated CAPE of 15.9 sits at the 15th percentile — historically cheap for China — below most of its own past readings.
Multiples as of August 2026.
- Historical average
- 23.6
- 262 months
- Historical median
- 23.1
- Percentile rank
- 15th percentile
- vs full history
- All-time range
- 10.6 – 35.4
- Last updated
- August 2026
- since November 2004
Understanding this metric
What is it?
The Cyclically Adjusted Price-to-Earnings ratio (CAPE), also called the Shiller P/E or PE10, measures how expensive China equities are relative to their inflation-adjusted earnings over the prior ten years. By averaging a decade of real earnings it smooths out the booms and busts of the business cycle, giving a steadier read on valuation than a one-year P/E.
How is it calculated?
CAPE is computed by dividing the real (inflation-adjusted) price of the China equity market by the average of its real earnings over the trailing ten years. Both price and earnings are expressed in today's currency using a consumer price index, so a CAPE of 25 means investors are paying 25 times a smoothed, inflation-adjusted measure of annual earnings.
Historical interpretation
Historically a high CAPE has been associated with lower subsequent long-run real returns, and a low CAPE with higher returns. It is a valuation gauge, not a market-timing signal: rich valuations can persist or grow richer for years. Compare today's reading to the metric's own history — its average, median and percentile rank — rather than to a single "fair value" number. China is the cheapest major market we track, and the most important case for reading a low multiple with care. Chinese large-caps trade at roughly half the all-world index's price-to-earnings and a fraction of its price-to-book — but that discount reflects real risks rather than a simple bargain: a multi-year property downturn, the 2021 regulatory crackdown on technology and private education, deflationary pressure, and the state ownership and geopolitical/delisting overhang that hang over the large state-owned banks and energy firms dominating the index. The CAPE percentile (how China is priced against its OWN history) is more useful here than the raw level, because a market can stay structurally cheap for years. Read it as "cheap for reasons", and weigh those reasons before concluding it is a bargain.
Limitations
CAPE has well-known limitations. Accounting standards, payout policies, profit margins, interest rates and sector composition all change over decades, which can shift the "normal" range upward or downward. The ten-year window still includes unusual periods (such as a deep recession) that distort earnings. CAPE says little about the next year and should be combined with other metrics rather than used in isolation. This series is an ESTIMATE, not a measured CAPE. Free data does not include long-history, per-country 10-year real earnings, so it is built by anchoring a recent CAPE level and rolling it through history using the inflation-adjusted price of the iShares China Large-Cap ETF (FXI) with a steady real-earnings trend removed. History only extends as far back as the country ETF (typically the late 1990s/2000s). Treat it as a directional valuation gauge. See the methodology documentation for sources and assumptions.
Frequently asked questions
Is the Chinese stock market cheap right now?
Yes, on the numbers: the large-cap China index shown above trades at well under half the all-world index's price-to-earnings and price-to-book — the lowest of the major markets we track. But cheap is not the same as a bargain. China's discount reflects a property-sector downturn, regulatory and geopolitical risk and weak investor confidence, so read the low multiple beside those risks and the CAPE percentile against the market's own history.
Why is the Chinese stock market so cheap?
Several overlapping reasons: a prolonged property-market downturn, the 2021 regulatory crackdown on big technology and private education, deflation and slowing growth, and the geopolitical and delisting risk plus state ownership that weigh on the banks and energy giants dominating the large-cap index. Investors demand a steep discount to hold those risks.
Is China a value trap?
That is the central debate. China is cheap on every price multiple, but a market can stay cheap for years if earnings disappoint or capital is not returned to minority shareholders — the classic 'value trap'. The CAPE percentile (versus China's own history) and the trend in earnings and shareholder returns matter more than the headline multiple for judging whether the discount will ever close.
Is the China CAPE figure a measured ratio?
No. Only the US CAPE is a measured series (from Robert Shiller). China's CAPE here is an estimate built from a free ETF price proxy (FXI) and a recent anchor, so treat it as directional. The trailing P/E, P/B, P/S and dividend yield shown above, however, are real, current figures reported by the large-cap tracker fund.
Embed or download this chart
Free to use with attribution. The embedded image updates automatically as the data does.