« Back to Results
Loews Philadelphia, Washington C
American Real Estate and Urban Economics Association
International House Pricing
Friday, Jan. 5, 2018 8:00 AM - 10:00 AM
- Chair: Piet Eichholtz, Maastricht University
A Tale of Two Countries: Comparing Booms, Busts and Bubbles in the United States and Chinese Housing Markets
AbstractThis paper compares the recent evolution of property values in the U.S. and China across cities and time, by estimating similar models for the two countries, to compare long run fundamental levels and short run dynamics. We find some similarities between the countries in terms of long run fundamentals, but major differences in adjustment. In particular, we find that the U.S. adjustment process is prone to “bubbles” in the sense of strong momentum in the short run while slowly reverting to trend over time. On the other hand, Chinese prices have been strongly mean reverting, with nothing like momentum. In short, our results suggest that the U.S. house prices have tended to chase past house prices, at least for a while; whereas the counterpart in China have tended to chase past rents. We find differences across cities, especially in China, but differences inside countries are smaller than differences between them.
Analyzing the Changes in the Distribution of House Prices in Beijing
AbstractUtilizing a comprehensive housing transaction dataset in 2012 and 2015 from Beijing, China, we analyze the changes in the distribution of house prices over time through regression and decomposition methods. The OLS and quantile regression results suggest that housing characteristics are valued differently between 2012 and 2015 and at different locations of the house price distribution. The result of Oaxaca-Blinder mean decomposition suggests that the proportion of the price gap between 2015 and 2012 due to altered returns over time is greater than the proportion caused by changes in housing characteristics. The quantile decomposition shows that the decomposition effects are heterogeneous across the complete distribution of house prices. We also find that low-priced homeowners are exposed to a decline in house prices, while there is a deteriorating house condition from 2012 to 2015. The medium-low-priced homeowners also experience a decrease in living conditions, but a rise in house prices. However, high-priced home buyers pay more for housing but enjoy better living conditions.
Locally Weighted Quantile House Price Indices and Distribution in Japanese Cities, 1986
AbstractLocally weighted regression, or loess, is a non-parametric approach taken through a multivariate smoothing procedure; it allows the coefficients of hedonic house price regressions to vary over space. Using data on residential condominium sales in 15 Japanese cities over the 1986–2015 period, we construct house price indices based on loess and quantile regressions. The estimates show how the appreciation rate of house prices changed over time in the Greater Tokyo and Kansai areas, especially during the boom and the burst. One advantage of taking the locally weighted quantile approach is that it facilitates comparisons of the change in a full distribution for various large metropolitan areas, each of which contains multiple cities. During the Japanese asset bubble of the 1980s, house prices in the Greater Tokyo area rose more rapidly in the boom (and declined more slowly in the burst) than those in the Kansai area. The house prices in the Kansai area had a high appreciation rate during the boom and declined more afterwards, compared to the Greater Tokyo area. The distribution of house price changes was larger in the Kansai area, which saw a small degree of variation before the boom but had a large one at the peak of the boom. During the boom, the appreciation rate of high-priced houses was larger than that of low-priced houses in the Kansai area; meanwhile, in the Greater Tokyo area, the prices of high-priced houses appreciated less than did those of low-priced houses.
University of Cambridge
University of Southern California
Massachusetts Institute of Technology
University of Amsterdam
- C2 - Single Equation Models; Single Variables
- R3 - Real Estate Markets, Spatial Production Analysis, and Firm Location