TOPIX vs Nikkei 225: What Each Japanese Index Actually Measures

A strong day for the Nikkei 225 is not a precise description of what happened to every Japanese share. Nor does buying a TOPIX fund recreate the portfolio implied by the country's most familiar stock-market headline. The two benchmarks organise Japanese equities differently, and that construction changes which companies drive the result.

Neither index is a direct measure of Japanese GDP. Both measure securities under an index rulebook. Our Japan market profile explains the wider economy; this guide focuses on what the benchmarks include and how they translate share-price movements into index returns.

TOPIX: equity size, adjusted for available shares

JPX defines TOPIX as a free-float-adjusted market-capitalisation benchmark. The adjustment matters because the amount of equity available to investors can differ from a company's total issued equity. A simple list of companies by total market value will therefore not reproduce the benchmark weights.

Conceptually, a larger eligible free-float equity value gives a company more influence, subject to the index rules. The practical question for a fund holder is how much of the portfolio sits in the largest constituents and sectors after those adjustments. A broad constituent universe does not, by itself, guarantee an evenly distributed portfolio.

Nikkei 225: price weighting with explicit adjustments

Nikkei selects 225 domestic common stocks from the Tokyo Stock Exchange's Prime Market, considering liquidity and sector balance. It conducts regular reviews in April and October. The index uses price adjustments and a divisor, so describing it as a simple average of today's unadjusted share prices would be inaccurate.

The consequence is still different from market-cap weighting: a company's influence is connected to its adjusted price rather than simply to the total value of its equity. A high nominal share price alone is not enough to calculate that influence. The current adjustment factors and divisor must also be considered.

Three comparisons that can produce the wrong answer

First, a price index excludes the reinvestment of cash dividends. A total-return index includes them according to its methodology. Comparing a price chart for one benchmark with a total-return chart for the other exaggerates or conceals differences that have nothing to do with the portfolio's stock selection.

Second, currency changes the result for an overseas holder. A yen return and a dollar return answer different questions. A hedged fund introduces another set of effects, including hedge implementation and costs. The trading currency printed beside an ETF is not sufficient evidence that the underlying yen exposure is hedged.

Third, a fund is not the index itself. Fees, tax treatment, cash holdings, trading costs and replication choices can create a gap between published benchmark returns and the return an investor receives. Compare a fund against the exact benchmark series named in its documents, using the same dates.

Constituent counts and rule changes need dates

The TOPIX documentation includes current rules and material concerning its next revision. Material prepared for October 2026 should not be described as already implemented in an August comparison. For this reason, we link the provider's documentation rather than freeze a potentially misleading TOPIX constituent count in the headline.

The same discipline applies when a company joins or leaves a benchmark. A recent country ranking, a fund factsheet and an index constituent file may describe different reference dates. Align them before attempting to explain a portfolio's exposure or a daily market move.

Start with the exposure you are trying to understand

To study the largest Japanese companies by market cap, use an issuer ranking. To analyse the behaviour of a particular benchmark, use its constituent weights and return series. To evaluate a fund, add its costs, domicile, tax documentation and currency policy. Each step answers a different question.

GMR's editorial preference is to put these definitions before performance tables. A five-year return comparison can be useful, but only after the measurement is consistent. Publishing two eye-catching percentages without specifying the series and currency gives an appearance of precision without a sound comparison.