The spillover effect of real estate boom on stock market efficiency: evidence from China

ABSTRACT

We examine how the increase in real estate prices affects stock market efficiency. Using a sample of Chinese listed firms from 2010 to 2019, we find that high local real estate prices significantly exacerbate stock mispricing. Our findings remain robust after dealing with endogeneity concerns using an instrumental variable approach, an alternative sample choice, and alternative variable measurement. The positive effect of housing prices on stock mispricing is weaker for cross-listed firms and more pronounced for firms with more positive media coverage. Further analysis shows that the booming real estate market accelerates c orporate financial assets investment and inspires optimistic investor sentiment. Our results shed light on the negative spillover effect of the overheated real estate sector from the perspective of capital market efficiency.

KEYWORDS:

• Real estate prices
• stock market efficiency
• financial assets investment
• investor sentiment

JEL CLASSIFICATION:

• G14
• G30
• R30

Acknowledgements

Ge Yang acknowledges the financial support from the China National Natural Science Foundation (No. 72302048). Zeyu Sun acknowledges the financial support from the China National Natural Science Foundation (No. 71802047). Ximing Yin acknowledges the financial support from Hunan Provincial Natural Science Foundation (No. 2023JJ40196). Peng Bi acknowledges the financial support from the Jiamusi University Doctoral Special Fund Initiation Project (JMSUBZ2021-18) and Heilongjiang Province Philosophy and Social Science Research Project (22GLE374). We thank the editor, two anonymous referees, and Xiaojia Zheng for their valuable comments. Their insightful comments help us to improve the quality of this paper.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplemental data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/00036846.2024.2336884.

Notes

¹ Another strand of literature examines the determinants of housing market movements (e.g. newspaper-based economic sentiment (Balcilar et al. 2021)).

² For short-term financial assets (i.e. financial assets measured at fair value through profit and loss), the accounting rules (CAS for Enterprises No. 22) require to recognize the appreciation and depreciation of this type of assets as changes in fair value gains and losses on the income statement. The changes in fair value gains and losses, which do not involve with cash inflows and outflows, introduce accrual components into earnings.

³ Their overtrading behaviour resulted from excessive optimism could drive local listed firms’ stock prices above or below intrinsic values depending on the firms’ high-predicted/low-predicted long-term earnings growth rate. Investors may overprice firms with high long-term earnings growth rate and underprice firms with low long-term earnings growth rate. Overall, the excessive investor sentiment drives stock prices away from their intrinsic value and aggravates the stock mispricing.

⁴ Our sample period spans 10 years from 2010 to 2019. We start our sample period from 2010 to rule out the effect of the global financial crisis in 2008 and 2009, which may confound our measure of stock mispricing. Our sample period ends in 2019 due to data availability of the housing prices of all cities in China.

⁵ Frankel and Lee (1998) show that cross-sectional investigation using intrinsic value is not sensitive to the choice of costs of capital.

⁶ The database provides housing prices in 298 cities. To enlarge the sample, we replace the missing real estate price of a city with the real estate price of the province’s capital city. Results remain similar if we exclude cities with missing values in real estate prices. Our measure of city-level housing prices can not only control for common regional shocks to housing prices, but also the city-level idiosyncratic variations of local housing market.

⁷ The standardized macro market sentiment index is constructed based on the first principal component of six proxies: the closed-end fund discount, trading turnover, consumer confidence index, newly added stock accounts, the number and first-day returns on IPOs following Baker and Wurgler (2007).

⁸ 0.06 = 0.024 × 0.688/0.278.

⁹ See Berger and Ofek (1995) for a similar measure that aggregates imputed values for all segments of a firm.

¹⁰ For example, unexpected policy change for the next few years may affect both stock market and the housing market simultaneously. From the macro perspective, we may observe that the stock market has an immediate reaction in response to the policy effect, while followed by the later movements in the housing markets.

¹¹ Data on population density comes from the China Urban and Rural Development Database.

¹² The database takes the Full-Text Database of China’s Major Newspapers as the data pool, including more than 600 central or local newspapers and financial newspapers. It obtains firm’s annual media coverage through keywords such as stock code and abbreviation.

¹³ We are thankful to an anonymous reviewer for suggesting this alternative explanation.

Additional information

Funding

The work was supported by the National Natural Science Foundation of China [72302048, 71802047]; Hunan Provincial Natural Science Foundation [2023JJ40196]; Jiamusi University Doctoral Special Fund Initiation Project [JMSUBZ2021-18] and Heilongjiang Province Philosophy and Social Science Research Project [22GLE374].