KWEB vs SMH: Correlation
How closely do KraneShares CSI China Internet ETF (KWEB) and VanEck Semiconductor ETF (SMH) trade together? Their weekly returns over three years give a correlation of 0.32, which is moderate.
Data as of 2026-08-27 · refreshed every trading day · weekly returns · methodology
How correlated are KWEB and SMH?
On 3 years of weekly data the KWEB/SMH correlation comes out at 0.32, moderate. Recent behaviour matches the longer record: 0.32 over 1 year against 0.32 over 3. The 5-year figure is 0.31, and annualized covariance runs at 370.3 %².
Among the 40 assets we track against KWEB, SMH ranks #27 by 3-year correlation. Correlation aside, the last 12 months split them widely, with SMH ahead by 119.2 points (-26.1% versus +93.1%). Across three years, the rolling one-year figure varied moderately, from 0.24 to 0.66.
How is this computed?
Pearson correlation on weekly returns: ρ(A,B) = cov(rA, rB) / (σA · σB), over windows of 52, 156 and 260 weeks. Covariance is annualized (×52) and expressed in %². Full definitions on the methodology page.
KWEB vs SMH: side by side
| KWEB (KraneShares CSI China Internet ETF) | SMH (VanEck Semiconductor ETF) | |
|---|---|---|
| 1-year return | -26.1% | +93.1% |
| 5-year return | -36.2% | +332.8% |
| Volatility (ann.) | 33.9% | 33.7% |
| Beta vs S&P 500 | 0.90 | 1.91 |
| Max drawdown (3Y) | -41.6% | -35.7% |
| Sector / category | ETF · Thematic | ETF · Thematic |
Year-by-year returns
| Year | KWEB | SMH |
|---|---|---|
| 2022 | -17.2% | -33.5% |
| 2023 | -9.1% | +73.4% |
| 2024 | +12.0% | +39.1% |
| 2025 | +23.5% | +49.2% |
| 2026 | -23.3% | +59.1% |
Calendar-year price returns; the current year is year-to-date as of the data date above.
Are KWEB and SMH good diversifiers for each other?
Yes, to a useful degree: a correlation of 0.32 leaves real independence between the two, which historically damped combined volatility.
FAQ
What is the correlation between KWEB and SMH?
As of 2026-08-27, the correlation of weekly returns between KWEB and SMH is 0.32 over 3 years, 0.32 over 1 year and 0.31 over 5 years.
Is SMH a good diversifier for KWEB?
Yes, to a useful degree: a correlation of 0.32 leaves real independence between the two, which historically damped combined volatility.
What does a correlation of 0.32 mean?
On the −1 to +1 scale, 0.32 describes how much the two returns move together: +1 is lockstep, 0 is independence, negative values mean opposite directions. It says nothing about which performed better.
Use this data
$ curl https://www.pairbook.io/api/v1/pairs/kweb-vs-smh.json
Drop this badge in a README or notebook; it updates with the data:
[](https://www.pairbook.io/pair/kweb-vs-smh/)
Free with attribution; caching and terms are described in the API documentation.
Related comparisons
Hubs: KWEB correlations · SMH correlations
