Inflation-hedging properties of regional Chinese real estate market: evidence from 35 cities in China

Abstract

The housing markets in China have been gaining considerable interest from investors, but the inflation-hedging characteristics of housing remain ambiguous. Based on Chinese city-level data, this study evaluates different inflation-hedging properties in eastern, middle and western real estate markets using panel vector autoregressive (PVAR) models. Findings suggest middle real estate markets afford the best hedging opportunities for expected inflation, which is robust considering housing market heterogeneity, financial crisis and the 2010 purchase restriction order. Moreover, hedging efficacy of anticipated inflation differs between markets with low and high supply–demand ratio.

Keywords:

• inflation-hedging
• Chinese regional real estate
• PVAR model
• difference-in-difference

JEL Classification:

• O18
• P25
• R11
• C33

Acknowledgement

The author wishes to express appreciation to Prof. Norm Miller and two anonymous reviewers for their valuable comments.

Notes

¹ Data: Chinese Household Finance Survey (CHFS).

² Data: American Community Survey (ACS).

³ City samples are geographically divided into three regions according to the official partition method of the statistical bureau in China, namely the eastern region, the middle region and the western region.

⁴ Data source: WIND database. We use domestic loans for regional real estate investment rather than loan to value (LTV) ratio to explore the monetary policy factors in influencing the ability of hedging inflation among regional real estate markets. Adjusting LTV ratio is one of the targeted macro prudential policies of the central bank to contain risk of a real estate boom. But LTV ratios are relatively constant, ranging from 50% to 60% on residential real estate loans.

⁵ PVAR models are selected because housing prices and inflation potentially share an endogenous relationship.

⁶ Ten OECD countries: Australia, Denmark, Finland, France, Germany, Japan, Netherlands, Spain, the UK and the US.

⁷ The quarterly housing price index released by the National Bureau of Statistics covering 2000–2010 is only available for 35 cities. This is the earliest publicly available housing price index for China. See. Fu et al. (2008) for more discussion.

⁸ The eastern regions include 16 cities (Beijing, Tianjin, Shijiazhuang, Shenyang, Dalian, Shanghai, Nanjing, Hangzhou, Ningbo, Fuzhou, Xiamen, Jinan, Qingdao, Guangzhou, Shenzhen, Haikou), the middle regions include 8 cities (Taiyuan, Changchun, Haerbin, Hefei, Nanchang, Zhengzhou, Wuhan, Changsha) and the western regions include 11 cities (Hohhot, Nanning, Chengdu, Chongqing, Guiyang, Kunming, Xi’an, Lanzhou, Xi’ning, Yinchuan, Urumqi). See Appendix for detailed descriptions.

⁹ Data source: http://www.stats.gov.cn/

¹⁰ We use the Dickey–Fuller test to examine the inflation rate and find that there is no unit root for the inflation rate, which means the inflation rate is stationary. We use AIC and Schwarz criterion (SC) to determine the ARMA model orders at city level. Different cities have different orders of ARMA.

¹¹ This process was first proposed by Arellano and Bover (1995) and adopted by Love and Zicchino (2006).

¹² The estimator augments Arellano and Bover (1995) by making an additional assumption that first differences of instrument variables are uncorrelated with fixed effects. This allows more instruments and can dramatically improve efficiency. It builds a system of two equations – the original equation and the transformed one – and is known as system GMM. See Oikarinen and Engblom (2015) for discussion.

¹³ We experiment with various possible orderings among the three variables and find results regarding different relationships between regional real estate returns and inflation among the three regions are robust to orderings selected.

¹⁴ The cities that implemented the policy before January 2011 are Beijing, Tianjing, Taiyuan, Shanghai, Nanjing, Hangzhou, Hefei, Fuzhou, Xiamen, Jining, Qingdao, Zhengzhou, Guangzhou, Shenzhen, Haikou and Kunming.

Additional information

Funding

This work was supported by the Overseas Study Program of Guangzhou Elite Project [JY201414]