Does Cross-Listing Really Enhance Market Efficiency for Stocks Listed in the Home Market? The Perspective of Noise Trading in the Chinese Stock Market

ABSTRACT

The investor recognition hypothesis and the bonding hypothesis, which help us understand the market quality of stocks that are cross-listed on different stock markets, imply improved market efficiency after cross-listing because of increased investor participation. However, the noise trading of inexperienced investors in the Chinese stock market negatively affects market efficiency. By employing propensity score matching and multivariate regression analysis, we show that the increased individual investor participation actually lowers market efficiency in their home market after cross-listing. This effect is more evident for stocks that were either listed first on the Chinese stock market or listed on the Chinese stock market and the Hong Kong stock exchange (SEHK) on the same date than for stocks that were listed first on the SEHK.

KEY WORDS:

• cross-listing
• market efficiency
• noise trading
• order imbalance
• propensity score matching

JEL CLASSIFICATION:

• G12
• G14
• G15

Acknowledgments

We would like to thank the editor, Ali M. Kutan, and two anonymous referees for their helpful comments.

Notes

^1. The qualified foreign institutional investor (QFII) policy allows foreign institutional investors to trade A-shares within a limited quota system granted by the Chinese government. The qualified domestic institutional investor (QDII) policy, introduced in April 2006, allows qualified domestic institutional investors to invest in fixed-income products and money markets. On May 11, 2007, the investment scope was widened to stock-related products.

^2. The quota granted to QDII is more highly regulated by the government.

^3. Chordia and Subrahmanyam (2004) argue that “the concept of order imbalance over an interval makes sense only in a paradigm of an intermediated market, wherein market makers accommodate buying and selling pressures from the general public” (p. 486). The Chinese stock market is an electronic continuous auction market without market makers. Thus, the order imbalance method may not be appropriate in the Chinese stock market. However, Chordia and Subrahmanyam (2004) also indicate that “order imbalances can signal excessive investor interest in a stock, and if this interest is autocorrelated, then order imbalances could be related to future returns” (p. 486). Thus, the use of the order imbalance method is justified by this statement. Empirical findings verify the correlation between order imbalance and future stock returns in the Chinese stock market (Narayan, Narayan, and Westerlund 2015).

^4. More specifically, this method measures the speed of information integration into prices during the 5 minutes after new information arrives if the change in order imbalance and stock returns is considered new information.

^5. The variable definitions are provided in the Appendix.

^6. As the post-treatment values of covariates are used in our PSM setting, the PSM analysis does not intend to show the causal effect of cross-listing on market efficiency, trading costs, individual noise trading level, and institutional trading level. It is used simply to identify an appropriate control group by controlling for the covariates related to market efficiency (Heinrich, Maffioli, and Vázquez 2010). The key dimensions are the main control variables in multivariate regression. For a detailed description, see the section “Methods of Analysis and Data.”

^7. Definitions of the variables are provided in the Appendix.

^8. OIBNUM is the estimated number of buyer-initiated trades minus seller-initiated trades divided by the total number of trades for the time interval.

^9. The difference in the total capitalization mean is significant between the two groups at the 5% level. This result remains unchanged while a different matching technique (composite match score) is used. Moreover, the total capitalization mean for stocks in the control group is also much higher than that of the full sample. Therefore, we do not consider that this significant difference means inferior matching quality.

^10. Based on their lab data, Bloomfield, O’Hara, and Saar (2009) argue that noise trading reduces bid-ask spreads and improves market liquidity. However, Wang (2010) indicates that while the informed traders’ arbitraging activity against noise traders diminishes, market liquidity actually deteriorates. Our empirical results are consistent with the analysis by Wang (2010).

^11. The Shanghai-Hong Kong Stock Connect is a cross-boundary investment channel that connects the SHSE and the SEHK. Under the program, investors in each market are able to trade shares on the other market using their local brokers and clearinghouses.

^12. Some of the cross-listed stocks are for state-owned enterprises with a large portion of nontradable shares retained by the enterprises themselves or other state-owned enterprises and institutions. In our sample period, some of these shares became tradable, which created a sudden increase in the proportion of institutional holdings. However, these shares, which ensure a majority of holding by the State (because these firms are “state-owned”), are not actively traded and distort our measure of institutional trading. Chen, Harford, and Li (2007) balance the costs and benefits of institutional ownership and indicate that long-term investments, such as the corporate holding of state-owned enterprises and institutions in our case, specialize in monitoring rather than trading. Their holdings enhance the quality of company management instead of influencing price efficiency by trading actively on information. Therefore, we omit these shares from total institutional trading. Insholding represents all institutional holdings, including mutual funds, securities firms, QFIIs, insurance companies, social security funds, trust companies, banks, and private funds, except general corporate holdings.

Additional information

Funding

Yingyi Hu acknowledges financial support from the Humanity and Social Science Fund of the Ministry of Education of China (approval number 10YJC790087). The authors are responsible for any remaining errors.