Abstract
In this paper we examine house price synchronization in 15 global cities using real house price data from 1995:Q1-2020:Q2. We find that although there is evidence for bilateral positive phase synchronization, there is no evidence for an integrated global housing market for our sample of cities. Using a hierarchical clustering approach, we identify three clusters of cities with similar housing price cycles that are not solely determined by geographic proximity. We interpret this finding as suggestive of a rather segmented housing market for the global cities in our sample. Using a dynamic factor model with time-varying stochastic volatility, we decompose a city’s real housing price growth into a global component, a cluster-based component, and an idiosyncratic component. For most cities in our sample, the global component plays a minor role, whereas the cluster-based factor explains a large fraction of the observed variation in real house price growth, with its contribution peaking during the Great Recession of 2007-09.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
In our discussion, we use housing market synchronization and house price synchronization interchangeably although co-movement in house prices can occur even if markets are segmented. For instance, it is possible that a common factor such as global liquidity drives house prices in two otherwise segmented markets.
There are various channels through which housing markets may synchronize across national borders. For instance, if a group of countries share strong trade linkages and exhibit synchronization in real economic activity, then it is reasonable to expect a high degree of synchronization in the housing market as well. Similarly, a higher degree of synchronization in interest rates across different countries may induce similar cycles in housing markets.
Gyourko et al. (2013) use “Superstar Cities” term for major cities in the U.S. However, many business news platforms commonly use labels such as “global cities” or “superstar cities” to include major cities across the globe. For example, see https://www.mckinsey.com/featured-insights/innovation-and-growth/superstars-the-dynamics-of-firms-sectors-and-cities-leading-the-global-economy.
See, https://www.economist.com/graphic-detail/2016/03/31/global-house-prices. The discussion on global city housing price can be found at: https://www.economist.com/graphic-detail/2019/03/11/global-cities-house-price-index
Knight Frank also provides prime global cities index. See https://content.knightfrank.com/research/323/documents/en/prime-global-cities-index-q2-2021-8341.pdf.
A recent example can be found at: https://www.cnbc.com/2020/09/30/these-cities-are-at-risk-of-a-housing-bubble-as-home-prices-inflate-worldwide.html
Some of the media attention on housing market in the global cities is focused on the prime locations within those cities. Our paper does not focus on these prime locations but on the overall market in global cities. It is possible that housing markets of prime locations are more synchronized globally than city-level housing markets.
Our methodology also allows for endogenous treatment of clusters and more general treatment of time-varying volatility. See details in “A Latent Factor Model with Time-varying Stochastic Volatility” section.
Our sample of cities include Amsterdam, Chicago, Copenhagen, Hong Kong, London, New York City, Oslo, Paris, San Francisco, Singapore, Stockholm, Sydney, Tokyo, Toronto, and Vancouver.
Phase implies expansion and contraction of housing market in this study.
Note that there is some evidence that geography may not be the primary factor driving housing price cycles for the U.S. For example, Hernández-Murillo et al. (2017) found that different clusters of cities do not necessarily share geographic similarity and exhibit their own idiosyncratic volatility.
Hale (2012) documents state-dependence for banking networks with a decline in the number of bank relationships in the global banking network during episodes of recessions in the U.S. For the housing market. Claessens et al. (2010, 2012) document that housing price cycles are highly synchronized and the degree of co-movement is higher during periods of synchronized recessions. See Huidrom et al. (2016) for state-dependence in emerging market business cycles.
Singapore is a city-state and Hong Kong is a metropolitan area that is officially a special administrative region of the People’s Republic of China.
Note that data for only 6 out of 15 cities in our sample is sourced from a single source, namely BIS Detailed Residential Property Prices. For other cities we use different official sources to collect data on residential properties. As noted by Hirata et al. (2013) comparing house price indices across different countries is problematic due to different methodologies and concepts used in the measurement of these prices. There are housing price indices available from private entities such as Knight Frank, The Economist, etc. These measures are subject to the same measurement issue and are also not freely available to the public, unlike the official sources used in this paper.
To compute real housing price index of a city we deflate the nominal index by city or metropolitan area level consumer price index where available. For Amsterdam, Copenhagen, London, Paris, Oslo, and Stockholm we use national consumer price index data as the city level data was not available for these cities in our sample.
Note that our approach is similar to the measure of business cycle coherence based on a binary transformation proposed by Harding and Pagan (2006). Similar methodology has been adopted to study international synchronization in business cycles, credit-GDP gaps, and credit gaps (see Mink et al. (2012) and B. et al. (2016), and Meller and Metiu (2017)).
Meller and Metiu (2017) apply similar approach to credit cycles for 14 advanced economies.
A similar approach for identifying number of clusters is used by Meller and Metiu (2017) where they investigated clusters in aggregate credit gaps between 14 developed nations.
Note that these clusters are dependent on the cut-off value of 1 we used in our algorithm as a stopping rule. For instance, if we use a cut-off point of 0.7, then we will get four clusters, with Cluster 2 splitting into two clusters containing Chicago and New York in one cluster, and Amsterdam and Oslo in another separate cluster.
We also implemented a cluster analysis at the country level to investigate the degree of housing price growth synchronization at the national level. The results are presented in Fig. 10 of the Appendix A. We find that housing markets at the city-level show a greater segmentation of the housing market than at the national level. In addition, we use data on real housing price growth for 19 cities in the U.S. and investigate the degree of synchronization in the U.S. housing market. The results are presented in Fig. 11 and we observe that using the same cutoff point of 1 there is one common cluster that encompasses all cities in the U.S. This gives us confidence that our finding of a segmented housing market for global cities is not merely an artifact of our clustering approach.
In Section 3.2 we identified three clusters in our sample. : The first cluster has 4 cities: Hong Kong, Singapore, Tokyo and Vancouver. The second cluster also has 4 cities: Chicago, NYC, Amsterdam and Oslo. The third cluster has 7 cities: London, Paris, Sydney, Toronto, Copenhagen, San Francisco and Stockholm.
We follow the conventional dynamic factor models and assume that the common factor, the cluster-based factors and the idiosyncratic factors are independent of each other. A richer set of models may allow correlation among the shocks of different factors. At the same time, these richer classes of models suffer from a lack of parsimony and identification problems.
Note that loadings for different factors are not reported when they are restricted to be zero in the model presented in the previous section. Further, we did not impose sign constraints on loading parameters in our estimation.
We also conducted a principal component analysis and found that the first two principal components capture only 43 percent of the overall variation in the housing price growth, with first component accounting for only 29 percent. These results are not included in the paper for brevity and are available upon request.
Meller and Metiu (2017) use a similar measure to define a real credit cycle.
We would like to thank an anonymous referee for making this point.
Using a very long-sample, Jorda et al. (2019) have found that since 1980, house prices have tended to be much less connected globally than equity markets.
Factors like population growth (Oikarinen et al., 2018), local variations in credit supply (Favara and Imbs, 2015; Cerutti et al., 2017; Mian & Sufi, 2019) and heterogeneity in housing supply elasticity (Glaeser et al., 2008; Oikarinen et al., 2018; Paciorek, 2013) are also important determinants of house prices.
The cities in our sample are based on Case and Shiller’s 20 cities index except Dallas. The data for Dallas were not available prior to 2001.
In Fig. 12. we also report the contribution of each factor to real housing price growth of each city in the U.S. sample. Note that common factor’s contribution is big for all cities but exhibits substantial variation over time.
Note that a relative lack of easily accessible data on economic activity at city-level makes it difficult to rigorously examine this issue. We believe a case-study approach with focused analysis of a few cities in a cluster is perhaps a way forward to examine factors that help explain why housing price cycles of certain cities in a cluster are more integrated than with those outside their cluster.
Similar finding was supported by Funke et al. (2018) who suggest that regionally differentiated macroprudential policy such as regional loan-to-value ratios perform best in reducing variance of housing prices in an environment with regionally segmented housing markets.
References
Alter, A., Dokko, J., & Seneviratne, D. (2018). House price synchronicity, banking integration, and global financial conditions. IMF Working Paper No. 18/250.
B., D., Dewachter, H., Ferrari, S., Pirovano, M., & Van Nieuwenhuyze, C. (2016). Credit gaps in Belgium: Identification, characteristics and lessons for macro-prudential policy. Financial Stability Report. National Bank of Belgium.
Bayer, P., Mangum, K., & Roberts, J.W. (2021). Speculative fever: investor contagion in the housing bubble. American Economic Review, 111(2), 609–51.
Cerutti, E., Dagher, J., & Dell’Ariccia, G. (2017). Housing finance and real-estate booms: A cross-country perspective. Journal of Housing Economics, 38, 1–13.
Claessens, S., Ayhan Kose, M., & Terrones, M.E (2010). The global financial crisis: How similar? how different? how costly? Journal of Asian Economics, 21(3), 247–264.
Claessens, S., Kose, M.A., & Terrones, M.E. (2012). How do business and financial cycles interact? Journal of International Economics, 87(1), 178–190.
Favara, G., & Imbs, J. (2015). Credit supply and the price of housing. American Economic Review, 105(3), 958–992.
Forbes, K.J., & Rigobon, R. (2002). No contagion, only interdependence: measuring stock market comovements. The Journal of Finance, 57(5), 2223–2261.
Francis, N., Owyang, M.T., & Savascin, O. (2017). An endogenously clustered factor approach to international business cycles. Journal of Applied Econometrics, 32(7), 1261–1276.
Funke, M., Mihaylovski, P., & Wende, A. (2018). Out of sync subnational housing markets and macroprudential policies. CESifo Working Paper No. 6887.
Glaeser, E., Gyourko, J., & Saiz, A. (2008). Housing supply and housing bubbles. Journal of Urban Economics, 64(2), 198–217.
Gyourko, J., Mayer, C., & Sinai, T. (2013). Superstar cities. American Economic Journal: Economic Policy, 5(4), 167–99.
Hale, G. (2012). Bank relationships, business cycles, and financial crises. Journal of International Economics, 88(2), 312–325.
Harding, D., & Pagan, A. (2006). Synchronization of cycles. Journal of Econometrics, 132(1), 59–79.
Hernández-Murillo, R., Owyang, M.T., & Rubio, M. (2017). Clustered housing cycles. Regional Science and Urban Economics, 66, 185–197.
Hiebert, P., & Roma, M. (2010). Relative house price dynamics across euro area and us cities: Convergence or divergence? Working Paper No 1206, European Central Bank.
Hirata, H., Kose, M.A., Otrok, C., & Terrones, M.E. (2013). Global house price fluctuations: Synchronization and determinants. NBER International Seminar on Macroeconomics, 9(1), 119–166.
Hoesli, M. (2020). An investigation of the synchronization in global house prices. Journal of European Real Estate Research, 13(1), 17–27.
Hoesli, M., & Reka, K. (2013). Volatility spillovers, comovements and contagion in securitized real estate markets. Journal of Real Estate Finance and Economics, 47, 1–35.
Hosszejni, D., & Kastner, G. (2020). Modeling univariate and multivariate stochastic volatility in r with stochvol and factorstochvol. arXiv:1906.12123.
Huidrom, R., Kose, M.A., & Ohnsorge, F.L. (2016). A ride in rough waters: Emerging markets buoyed the world after the global financial crisis, but are now in a major slowdown. Finance and Development, 53(003).
International Monetary Fund. (2018). House Price Synchronization: What Role for Financial Factors?, chapter 3 International Monetary Fund, Monetary and Capital Markets Department, USA.
Jorda, O., Knoll, K., Kuvshinov, D., Schularick, M., & Taylor, A. (2019). The rate of return on everything, 1870-2015. Quarterly Journal of Economics, 134(3), 1225–1298.
Kallberg, J.G., Liu, C.H., & Pasquariello, P. (2014). On the price comovement of u.s. residential real estate markets. Real Estate Economics, 42(1), 71–108.
Kastner, G., Frühwirth-Schnatter, S., & Lopes, H.F. (2017). Efficient bayesian inference for multivariate factor stochastic volatility models. Journal of Computational and Graphical Statistics, 26(4), 905–917.
Katagiri, M. (2018). House price synchronization and financial openness: A dynamic factor model approach. IMF Working Paper No. 18/209.
Kim, S., Shephard, N., & Chib, S. (1998). Stochastic volatility: likelihood inference and comparison with ARCH models. The Review of Economic Studies, 65(3), 361–393.
Klarl, T. (2018). Housing is local: Applying a dynamic unobserved factor model for the dutch housing market. Economics Letters, 170, 79–84.
Kose, M.A., Otrok, C., & Whiteman, C.H. (2003). International business cycles: World, region, and country-specific factors. American Economic Review, 93(4), 1216–1239.
Meller, B., & Metiu, N. (2017). The synchronization of credit cycles. Journal of Banking and Finance, 82, 98–111.
Mian, A., & Sufi, A. (2019). Credit supply and housing speculation. NBER Working Paper No. 24823.
Miles, W. (2017). Has there actually been a sustained increase in the synchronization of house price (and business) cycles across countries? Journal of Housing Economics, 36, 25–43.
Miles, W. (2019). Regional convergence and structural change in us housing markets. Regional Studies Regional Science, 6(1), 520–538.
Miles, W. (2020). Regional UK house price co-movement. Applied Economics, 52, 4976–4991.
Miles, W.H. (2015). Regional house price segmentation and convergence in the us: A new approach. Journal of Real Estate Finance and Economics, 23, 57–72.
Mink, M., Jacobs, J.P., & de Haan, J. (2012). Measuring coherence of output gaps with an application to the euro area. Oxford Economic Papers, 64 (2), 217–236.
Oikarinen, E., Bourassa, S., Hoesli, M., & Engblom, J. (2018). US metropolitan house price dynamics. Journal of Urban Economics, 105, 54–69.
Otrok, C., & Terrones M. (2005). House prices, interest rates and macroeconomic fluctuations: International evidence. mimeo.
Paciorek, A. (2013). Supply constraints and housing market dynamics. Journal of Urban Economics, 77, 11–26.
Primiceri, G.E. (2005). Time varying structural vector autoregressions and monetary policy. The Review of Economic Studies, 72(3), 821–852.
Stock, J.H., & Watson, M.W. (2007). Why has U.S. inflation become harder to forecast? Journal of Money Credit and Banking, 39(s1), 3–33.
Stock, J.H., & Watson, M.W. (2016). Dynamic factor models, factor-augmented vector autoregressions, and structural vector autoregressions in macroeconomics. In J.B. Taylor H. Uhlig (Eds.) Handbook of Macroeconomics, (Vol. 2 pp. 415–525). Elsevier.
Tsai, I.-C. (2018). House price convergence in euro zone and non-euro zone countries. Economic Systems, 42(2), 269–281.
Zhou, J., & Gao, Y. (2012). Tail dependence in international real estate securities markets. Journal Real Estate Finance and Economics, 45, 128–151.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix: A
Appendix: A
Country-clusters based on similarity of housing cycles
Cluster analysis for U.S. cities sample
Variance contribution by factor for U.S. cities sample
Cluster analysis based on Y-O-Y real housing price growth
Cluster analysis based on HP filter cycle
Rights and permissions
About this article
Cite this article
Bhatt, V., Kishor, N.K. (A)Synchronous Housing Markets of Global Cities. J Real Estate Finan Econ (2022). https://doi.org/10.1007/s11146-022-09903-2
Accepted:
Published:
DOI: https://doi.org/10.1007/s11146-022-09903-2