Author Details

Naoya Nishimura

Naoya Nishimura

Executive Director, MSCI Research

Social Sharing

Extended Viewer

Best practices for Japanese equity indexes

  • Planned reforms in Japan’s stock exchange have sparked a debate on investability and market representativeness of TOPIX, the country’s leading domestic benchmark index.
  • A market-capitalization-weighted index with a fixed cut-off may result in unexpected fluctuation of market coverage and the number of index constituents.
  • MSCI’s Global Investable Market Index (GIMI) methodology is designed to balance investability and market representativeness considerations.

The Tokyo Stock Exchange’s (TSE) plan to revamp its market structure and domestic equity index benchmark construction rules has sparked a debate on how to provide investability and representativeness in its leading index, the Tokyo Stock Price Index (TOPIX).1 While JPX, which owns the Tokyo exchange, has agreed to consolidate its five markets into three, the future index methodology for the TOPIX has not been announced.

Observers have said that the TOPIX has too many illiquid companies and is not representative of the market,2 making it less attractive to some domestic and international investors. One of the biggest problems: The index is tied to the current market structure of TSE, requiring inclusion of many small companies with illiquid shares. To address the investability issue, some market participants cited index methodologies that limit the number of constituents to a fixed number.3 Some other market participants as well as an expert study group created to recommend changes in market structures4 suggested setting a fixed minimum market capitalization for selecting securities.

But these are not the only options. One can seek to cover a fixed percent of the equity market in terms of free-float-adjusted market capitalization while providing minimum size and liquidity requirements.5

We investigated these three approaches to market-cap-weighted equity indexes that could potentially achieve high levels of investability and broad market representation:

  • Fixed number of constituents. We create a hypothetical simulated index by selecting the largest 1,192 Japanese stocks by free-float-adjusted market capitalization to analyze a market-cap-weighted index with a fixed number of constituents.6
  • Minimum market-cap. We create a hypothetical index that includes all Japanese equity securities with free-float-adjusted market capitalizations equal to or above JPY 7.3 billion to analyze a minimum market-cap cut-off index.7
  • Market coverage approach. We take the MSCI Japan IMI, which balances broad market coverage with minimum security size integrity, using the MSCI Global Investable Market Index (GIMI) methodology.


Comprehensive and consistent representativeness

Among the three indexes in our study, the minimum market-cap approach had both the largest market-cap coverage and highest number of constituents throughout our study period. But these two metrics fluctuated more in this approach than the other two because additions to and deletions from the index were based on a fixed market-cap size.



Data from June 2007 to December 2019, quarterly data based on effective dates of MSCI Quarterly Index Review. Float-adjusted market-cap (or Mcap) coverage is calculated as index-level Mcap divided by total Mcap of all Japanese public equity securities in MSCI Equity Database.



Data from June 2007 to December 2019, quarterly data based on effective dates of MSCI Quarterly Index Review.

On the other hand, fixing the number of constituents would have kept market coverage stable until 2015, when it declined due to an increase in the market capitalization of the smallest companies. These micro-cap stocks were not included in the index. Thus, when their market cap increased more than the rest of the market, the market cap coverage of the index decreased.



Data from Nov. 30, 2007 to Dec. 31, 2019. Data of MSCI Japan Micro-Cap Index starts from November 30, 2007. MSCI Micro Cap Indexes include securities which are not part of MSCI Investable Market Indexes and meet special requirements on size and liquidity.

In contrast, the market coverage approach consistently covered around 96% of free-float-adjusted market capitalization throughout the study period from June 2007 to December 2019. In addition, the number of constituents was less affected by market movements, because the rebalance of index constituents reflected the liquidity and size movements of all investable securities in the market.


Is size the only proxy for liquidity?

The larger the market-cap coverage becomes, the more illiquid names the index potentially includes. One way to seek to manage liquidity is to exclude micro-cap stocks, which have less impact on market coverage. Another approach is to look at how actively a stock is traded relative to its size.

Fixing the number of constituents and setting a minimum market-cap size effectively bars illiquid stocks. On the other hand, the market coverage approach, which is based on the MSCI GIMI methodology, incorporates both approaches: It sets a minimum market-capitalization requirement comparable with those of other developed-market indexes and screens out stocks based on average traded value ratio (ATVR), MSCI’s liquidity measure which shows the percentage of securities’ free-float-adjusted market capitalization traded per year.

Historically, the first two approaches in our analysis included more illiquid stocks than the market-based approach in terms of ATVR.8 The weight and number of illiquid stocks in the fixed number of constituents index were higher than that of the market-based approach.



Data from June 2007 to December 2019, quarterly data based on effective dates of MSCI Quarterly Index Review. ATVR is based on the latest month-end data preceding each QIR date.



Data from June 2007 to December 2019, quarterly data based on effective dates of MSCI Quarterly Index Review. ATVR is based on the latest month-end data preceding each QIR date.

The future of TOPIX methodology remains a subject of debate. A broad equity market benchmark is expected to have a large market representation and be investible, but these two objectives could be in conflict once market coverage goes beyond a certain point. Our analysis comparing the MSCI GIMI methodology with other hypothetical methodologies, however, showed that one could strike a balance by carefully incorporating market coverage and investability considerations into a methodology.



1“Review of TSE Cash Equity Market Structure.” JPX, March 31, 2020. TOPIX is a free float-adjusted market-capitalization-weighted index which consists of all domestic stocks listed on the first section of Tokyo Stock Exchange.

2Lee, M. J. and Kawamato, S., “Goldman says Japan stock index may still be too big after re-size.” Bloomberg, Feb. 10, 2020 (republished in Japan Times).

3“Comments Received from Market Participants in Response to the Review of the TSE Cash Equity Market Structure.” Tokyo Stock Exchange, March 2019.

4“Final Report by the Expert Study Group on the Structure of Capital Markets in Japan.” Financial System Council Capital Market Working Group, Dec. 27, 2019.

5For more detail, see “MSCI Global Investable Market Indexes Methodology.” MSCI, February 2020.

6We used 1,192 because it represents a historical average of the number of constituents in the MSCI Japan IMI. This index is rebalanced quarterly.

7We used JPY 7.3 billion because it represents a historical average of the minimum free float-adjusted market capitalization of MSCI Japan IMI constituents. This index is rebalanced quarterly.

8Illiquid securities here are defined as having ATVR less than 20%, following DM Minimum Liquidity Requirement of the GIMI Methodology.



Further Reading

MSCI Global Investable Market Indexes Methodology

Selected geographic issues in the global listed equity market

Is Europe more than the sum of its parts?

Small cap allocations may not be that straightforward

Building Best Practices Benchmarks for Global Equities