Bubble Or Riddle? An Asset-Pricing Approach Evaluation On China’s Housing Market
Qu FENG and Guiying Laura WU
EGC Report No: 2015/01
H/T Barry Ritholtz
Rapid house price growth and high price-to-income ratio in major Chinese cities have aroused a hot debate on whether there is an asset bubble in China’s residential housing market. To investigate this question, we employ an equilibrium asset-pricing approach, which suggests an non-arbitrage condition on the rent-to-price ratio. This ratio should be equal to the difference between the user cost of housing capital and the expected appreciation in house prices. Using a novel micro-level data set on pair-wise matched price-to-rent ratio collected in the fourth quarter of 2013, and forecasting the expected house price appreciation based on fundamental factors, our empirical exercises do not suggest the existence of a house price bubble at the national level. However, this conclusion highly depends on the expected income growth rate and may not apply to individual markets.
Bubble Or Riddle? An Asset-Pricing Approach Evaluation On China’s Housing Market – Introduction
In the past decade China has been experiencing a surge of house prices at an unprecedented rate. Figure 1 plots the average residential house prices in the 35 major Chinese cities. These cities represent all municipalities, provincial capital cities and quasi-provincial capital cities in China, whose house prices have been closely watched by policy makers, researchers and investors. On average their residential house prices have steadily increased from 2426 yuan per square meter in 2003 to 7718 yuan per square meter in 2012. This implies a more than tripled property value in 9 years, or a 13.7% nominal compound annual growth rate. During the same period, the average CPI of these cities only rose by 30%. Figure 2 depicts China’s average residential house price-to-income ratio, a common measure of housing affordability. At the national level, this ratio has sharply increased from 6.6 in 2003 to 8.1 in 2009, and gradually declined to 7.3 in 2013 after a series of house price regulations. The 35 major Chinese cities witness an even higher price-to-income ratio, which reached 8.5 in 2013 (E-house China, 2014). By contrast, the price-to-income ratio was around 4 in the US, 5 in the UK and 6 in Australia right before the recent financial crisis (Reserve Bank of Australia, 2008). Such rampant house price growth and unusually high price-to-income ratio have aroused great interest and concern on whether China has an asset bubble in its housing market.
To assess whether the house prices are too high or too low, an equilibrium asset-pricing approach has been offered by the housing literature. According to this approach, neither accelerating house price growth nor the remarkable price level itself is the intrinsic signs of a bubble, let alone the anecdotal price fluctuations in single property or casual observations on the housing markets. In contrast, the golden rule of the evaluation boils down to an non-arbitrage condition on the rent-to-price ratio, which is equal to the difference between the user cost of housing capital and the expected appreciation in house prices at equilibrium. Himmelberg, Mayer and Sinai (2005) is one leading example in applying this approach to assessing the house prices in the US.
This paper aims to address whether there is a house price bubble in China using this asset-pricing approach. We argue that the expected house price appreciation, instead of high house price growth or price-to-income ratio, is central to the debate on the existence of a house price bubble. There are, however, three significant challenges in implementing the approach to China. First, there are no readily available data on rent-to-price ratio that have properly controlled for house characteristics. Second, little is known on each component in the user cost of housing capital for a nascent market like China. Last and most importantly, although economic theory provides useful suggestions on the fundamental factors that determine the house prices, there is no prior information on, either their own expected growth rates, or their elasticities in house price growth accounting.
See full PDF below.