The CSI 300 and MSCI China answer different investment questions. The CSI 300 measures a basket of large, liquid mainland Chinese shares. MSCI China reaches across mainland and overseas listings, giving investors a different mix of companies, sectors and trading venues.
That distinction matters more than the shared “China” label. Choosing between them changes which businesses drive returns, how much influence the largest holdings have, and which parts of the Chinese equity market a portfolio leaves out. Neither benchmark represents the Chinese economy in miniature.
This comparison uses constituent and weight data dated September 30, 2026. Treat those figures as a dated snapshot, not permanent portfolio allocations. For the wider benchmark family, see our guide to Chinese stock market benchmarks.
What Each Index Includes
CSI 300: Mainland Market Exposure
The CSI 300 selects 300 securities from Shanghai and Shenzhen, using liquidity and market capitalization criteria. Its methodology screens out the bottom half of eligible securities by average daily trading value before ranking the remaining candidates by average daily total market capitalization. Portfolio weights then use adjusted market capitalization, with free float handled through inclusion bands rather than a simple equal allocation. These selection and weighting rules appear in the CSI 300 index methodology.
The practical distinction is between selecting a company and deciding its weight. A large company can qualify for the index without receiving a weight proportional to every share it has issued. Shares held by strategic owners affect the amount considered available to the market.
For an investor, the CSI 300 is therefore a mainland equity allocation, not a complete portfolio of Chinese businesses wherever they happen to list. A familiar Chinese brand does not automatically belong in it.
MSCI China: Several Share Classes and Listing Venues
MSCI China covers large and midsize companies across A shares, H shares, B shares, red chips, P chips and foreign listings, including ADRs. It targets approximately 85% of its eligible Chinese equity universe. Eligible large and midsize A shares enter at 20% of their free float adjusted market capitalization under the current MSCI China index coverage rules.
That 20% figure is easy to misread. It is an adjustment applied to eligible A shares before final index weights are calculated. It does not mean MSCI China always allocates 20% of its portfolio to mainland shares.
Consider a simplified example. Suppose an eligible A share has an otherwise applicable free float market value of $10 billion. Applying a 20% inclusion factor produces $2 billion for weighting purposes, before comparison with the rest of the index. Its final percentage depends on the adjusted value of every other constituent.
This makes MSCI China broader by listing venue, but not a neutral combination of the mainland and offshore markets. For the ownership and listing distinctions behind these labels, read our comparison of Chinese A shares, H shares and ADRs.
Holdings: Different Companies Drive the Results
The difference becomes tangible in the largest positions. At September 30, 2026, the CSI 300’s leading holdings included Zhongji Innolight, Kweichow Moutai and CATL. Its ten largest positions together represented approximately 23.11%, calculated by adding the published weights in the CSI 300 September 2026 factsheet.
A concentrated top position can make an index behave differently from a more evenly distributed basket, even when both contain hundreds of securities. The relevant question is not simply how many names appear in the constituent file. Ask how much of each invested dollar reaches the largest businesses.
| Rank | CSI 300 holding | Weight | MSCI China holding | Weight |
|---|---|---|---|---|
| 1 | Zhongji Innolight | 3.64% | Tencent | 13.86% |
| 2 | Kweichow Moutai | 3.18% | Alibaba | 9.37% |
| 3 | CATL | 3.11% | China Construction Bank H | 4.33% |
| 4 | Ping An Insurance A | 2.30% | ICBC H | 2.63% |
| 5 | Eoptolink Technology | 2.19% | Xiaomi | 2.26% |
MSCI China contained 576 constituents, but its ten largest holdings accounted for 41.60%. Tencent and Alibaba alone represented 23.23%, calculated from their published weights. Those holdings and allocations are recorded in the MSCI China September 2026 factsheet.
The implication is straightforward: more constituents did not produce less concentration at the top. In this snapshot, two MSCI China positions had roughly the same combined weight as the CSI 300’s ten largest holdings.
As a hypothetical sensitivity test, a 10% fall in a position weighted at 13.86% would subtract about 1.39 percentage points from an index, assuming every other price and relevant exchange rate stayed unchanged. That is arithmetic, not a forecast. It shows why the largest positions deserve more attention than the constituent count.
Sector Exposure: “Technology” Is Too Broad a Description
The September factsheets also show different sector allocations. MSCI China assigned 22.56% to consumer discretionary, 20.98% to financials and 17.94% to communication services. The CSI 300 allocated 21.6% to information technology, 20.8% to financials and 15.8% to industrials.
Financials had similar headline weights, but that does not make the portfolios equivalent. The remaining allocations, underlying companies and share classes still differed. Nor should sector labels be treated as interchangeable between providers without checking their classification methods.
For portfolio analysis, separate three questions: what sector a company occupies, how it earns money, and what could change its valuation. A single “technology” label is too blunt to answer all three.
Consider two hypothetical businesses. One sells advertising and transaction services through an online platform. Another manufactures components for overseas equipment makers. Both might feature in an investor’s technology thesis, yet advertising budgets, export orders and production costs would affect them differently.
This is why an investment view needs a second step. “Chinese innovation will grow” does not tell you which benchmark expresses that view. Identify the businesses expected to benefit, check whether they are included, then examine their actual weights. Otherwise the index name is doing more work than the research.
Mainland Exposure Is Not the Same as Domestic Revenue
Listing geography answers where a security trades. Revenue geography answers where its issuer earns money. Do not substitute one for the other when evaluating either benchmark.
A mainland allocation should not automatically be treated as a bet solely on Chinese household spending. Equally, an overseas listing should not automatically be treated as an export business. The useful portfolio question is how much revenue, profit and cash flow depends on the economic activity behind your investment view.
For a domestic consumption thesis, examine the relevant companies’ customer exposure and earnings drivers. For an export thesis, examine overseas sales, production locations and sensitivity to trade restrictions. Those checks are more informative than sorting every company into “onshore” or “offshore” and stopping there.
The distinction also prevents misleading comparisons with GDP. Neither index should be expected to match the growth rate of the economy: an equity investment has a purchase price, an earnings claim and a changing valuation. Economic growth alone does not settle the return calculation.
Overlap: Owning Both Does Not Remove Duplicate Exposure
When comparing funds tracking these benchmarks, run two overlap checks. Start with identical securities, then group different share classes belonging to the same issuer. Our guide to diversifying China exposure across companies and markets covers the wider portfolio decisions.
The holdings above illustrate why the second check matters: Ping An appears through its A shares in the CSI 300 and its H shares among MSCI China’s largest positions. Different securities can still expose a portfolio to the same business.
Use a weighted calculation rather than counting repeated names. In a hypothetical portfolio split equally between two funds, an issuer weighted at 4% in one and 2% in the other contributes 3% to the combined allocation: half of 4%, plus half of 2%.
Repeat that exercise at sector level. A blend can reduce dependence on one large holding while leaving exposure to a shared industry largely intact. Adding a second fund is a transaction; achieving diversification requires checking what changed.
Also compare the proposed allocation with existing holdings. Before adding either benchmark, inspect any broad emerging market or global fund already in the portfolio. The useful number is the combined exposure after the purchase, not the size of the new position viewed alone.
Compare Performance on the Same Basis
A return comparison needs matching dates, currencies and dividend treatment. The standard CSI 300 price series should not be placed beside an MSCI China total return series and treated as a fair contest. CSI publishes separate total return and net total return versions, while MSCI’s factsheet identifies the return convention used.
For a dollar based investor, use the same dollar reporting basis for both benchmarks. Comparing one series in renminbi with another in dollars mixes stock performance with a currency difference.
A hypothetical example shows the effect. If a renminbi asset rises 10% while the renminbi falls 5% against the dollar, the dollar return is approximately 4.5% before costs: 1.10 multiplied by 0.95, minus one. Simply subtracting the two percentages gives only an approximation.
Before accepting a performance chart, check four items:
- The start and end dates match.
- Both returns use the same reporting currency.
- Both series use comparable dividend and withholding tax assumptions.
- The comparison separates index returns from actual fund returns.
Then ask what drove the gap. Different holdings, sector weights and currency exposures are reasons for returns to diverge. A recent winner does not establish that one benchmark is permanently better suited to a portfolio.
Choose the Benchmark Before Comparing ETFs
An index defines the intended exposure; a fund implements it. Index funds may hold every constituent or use a sample, and fees, trading costs and tracking differences can cause returns to depart from the benchmark. These distinctions are covered in the SEC’s investor guidance on index funds.
Start by deciding which market exposure serves the portfolio. Only then compare funds tracking that exposure. A lower fee does not fix a benchmark mismatch: a cheap allocation to the wrong companies remains the wrong allocation.
Check the full benchmark name in the prospectus. Do not assume that an index carrying “China A,” “All Shares,” “ESG” or another modifier is interchangeable with MSCI China. Request the methodology and holdings file when the name leaves room for doubt.
Keep the implementation review separate from the benchmark decision. Compare the fund’s published objective, replication approach, costs and historical tracking against the exposure you selected. Our guide to China ETF fees, liquidity and tracking differences develops those checks without confusing fund quality with index composition.
Which Exposure Fits the Intended Role?
Use the CSI 300 as the starting benchmark when the intended role is a mainland large company allocation. Use MSCI China as the starting benchmark when the intended role spans mainland and overseas Chinese listings, accepting its partial A share inclusion and the concentration shown in its largest positions.
A combination can express a deliberate preference for more mainland exposure alongside a broader listing universe. It should not be selected on the assumption that two China funds automatically provide twice the diversification.
The decision is not “300 stocks versus 576.” It is which companies, share classes and weights belong in the portfolio. Write down the exposure you want, measure what each benchmark supplies, and choose a fund only after that comparison is complete.