ETFIQetfiq.com · independent ETF data

Data as of . Every figure is the themes desk's own, on published data.

BAI vs KWEB

iShares A.I. Innovation and Tech Active ETF and KraneShares CSI China Internet ETF, side by side on the themes desk.

0% of the two portfolios are the same securities at the same weight.

What they hold in common

By the books each fund has filed, BAI and KWEB hold 0% of their money in the same securities at the same weight.

Largest positions each one holds and the other does not
Only in BAIOnly in KWEB
Broadcom Inc 5.97%Tencent Holdings Limited 10.47%
NVIDIA Corp 5.43%PDD Holdings Inc. 7.94%
SK hynix Inc 4.93%Alibaba Group Holding Limited 7.72%
Taiwan Semiconductor Manufacturing Co Lt 4.64%NETEASE, INC. 6.99%
Alphabet Inc 4.26%MEITUAN 6.84%
Tower Semiconductor Ltd 4.22%BAIDU, INC. 4.20%
Lam Research Corp 3.99%Full Truck Alliance Co Ltd 4.17%
Advanced Micro Devices Inc 3.84%KUAISHOU TECHNOLOGY 4.09%

Weight overlap is an ETFIQ calculation: for every security both funds hold, the smaller of the two weights, summed. Above 50%, holding both is close to holding one of them twice. Holdings as filed for Apr 30, 2026 and Jun 30, 2026.

Performance, window by window

BAI and KWEB over each window. Every figure is an ETFIQ calculation with distributions reinvested. Source: ETFIQ.
WindowTotal returnvs the S&P 500vs the Nasdaq-100
BAIKWEBBAIKWEBBAIKWEB
3 months−2.1%−1.2%−6.8 pts−5.9 pts−4.2 pts−3.3 pts
6 months+36.4%−12.3%+21.2 pts−27.5 pts+16.2 pts−32.5 pts
1 year+43.1%−27.2%+23.1 pts−47.2 pts+17.5 pts−52.8 pts
3 yearsn/a+0.5%n/a−77.4 ptsn/a−93.0 pts
Since launch+81.2%+27.7%+46.5 pts−435.6 pts+34.9 pts−918.5 pts
Open the live comparison on ETFIQ
BAI and KWEB on the themes desk fields, as of Sep 4, 2026. Source: ETFIQ.
BAI
iShares A.I. Innovation and Tech Active ETF
KWEB
KraneShares CSI China Internet ETF
IssueriSharesKraneShares
ThemeArtificial intelligenceInternet and e-commerce
In the S&P 50050.0%0.0%
Active share vs the S&P 50077.9%100.0%
In the Nasdaq-10043.9%7.9%
Holdings4533
Top ten, share of the fund43.8%60.2%
Expense rationot published0.69%
Fee for the differing partnot published0.69%
Total return, 1 year+43.1%−27.2%
vs the S&P 500+23.1 pts−47.2 pts
vs the Nasdaq-100+17.5 pts−52.8 pts
Below its high−17.8%−69.8%

BAI in plain words

By weight, 50% of BAI's portfolio is stocks that are also in the S&P 500; its active share against the S&P 500 is 78%. The top ten holdings are 44% of the fund across 45 positions, as filed for Apr 30, 2026. Over the year to Sep 4, 2026 the fund returned +43.1% with distributions reinvested against +20.0% for the S&P 500, so a holder was ahead by 23.1 pts. It sits 17.8% below its all-time high of Jun 22, 2026.

KWEB in plain words

By weight, 0% of KWEB's portfolio is stocks that are also in the S&P 500; its active share against the S&P 500 is 100%. The top ten holdings are 60% of the fund across 33 positions, as filed for Jun 30, 2026. Over the year to Sep 4, 2026 the fund returned −27.2% with distributions reinvested against +20.0% for the S&P 500, so a holder was behind by 47.2 pts. It sits 69.8% below its all-time high of Feb 17, 2021.

Questions people ask

Do BAI and KWEB hold the same stocks?
By their latest filings, 0% of their books are the same securities at the same weight, which is mostly different names. Overlap is the sum of the smaller weight of every security they share.
Which is more different from the S&P 500, BAI or KWEB?
BAI has 50% of its weight in S&P 500 names and KWEB has 0%, so KWEB differs more from the index.

Other comparisons

Where these figures came from

ETFIQ links to the documents behind every figure. It is not affiliated with any issuer, and a link is not an endorsement.

BAI against KWEB, ETFIQ, data as of Sep 4, 2026. Every figure is arithmetic on a named public source; the method is at etfiq.com/methodology. A comparison is not a recommendation.

Cite this page. ETFIQ, BAI against KWEB, data as of Sep 4, 2026. https://etfiq.com/compare/themes/BAI-KWEB.html Free to use with attribution; the underlying files are at Open data.

How every figure is computed · Standards and sources