The rapid decline in housing prices of the United States (US), following a prolonged boom, is generally associated with the global economic and financial crisis of 2008-2009. Naturally, from a policy perspective, understanding what shocks drive the housing market performance is now of paramount importance in order to avoid the repeat of the catastrophic effects observed under the “Great Recession”. This research is motivated by the important effect changes in the housing market has on both households and the overall economy
The housing market plays an important role in the economy of the US, since it constitutes a significant share of many households’ asset holding and net worth. Various hypothesis and theories have been considered in literature to investigate the impact of different determinants that affect the housing market. We apply a variety of quantitative modeling methods to investigate the impact of various economic determinants such as inflation, monetary policy and macroeconomic shocks and housing sentiment on the US housing market. The thesis consists of five independent papers which are compiled into five chapters.
The first paper analyses the long-run relationship between U.S house prices and non-housing Consumer Price Index (CPI) over the monthly period 1953 to 2016 using a quantile cointegration analysis. The possibility of instability in standard cointegration models, suggesting the possible existence of structural breaks and nonlinearity in the relationship between house prices and non-housing CPI motivates the use of a time-varying approach, namely, a quantile cointegration analysis, which allows the cointegrating coefficient to vary over the conditional distribution of house prices and simultaneously test for the existence of cointegration at each quantile. Our results suggest that the U.S
non-housing CPI and house price index series are cointegrated at lower quantiles only, with house prices over-hedging inflation at these quantiles. In addition, we also show that this result holds for higher price levels only. Using these two sets of results, we conclude that house prices act as an inflation hedge when the latter is relatively higher and the former is lower.
The second paper explores the impact of monetary policy and macroeconomic surprises on the U.S market returns and volatility at the Metropolitan Statistical Area (MSA) and aggregate level using a GJR (Glosten–Jagannathan-Runkle) generalized autoregressive conditional heteroscedasticity (GARCH) model. Using daily data and sampling periods which cover both the conventional and unconventional monetary policy periods, empirical results show that monetary policy surprises have a greater impact on the volatility of housing market returns across time with particularly pronounced effect during the conventional monetary policy period. We also show that macroeconomic surprises do not have a significant impact on housing returns for most MSAs for the full sample, conventional and unconventional monetary policy periods.
The third paper examines the predictive ability of housing-related sentiment on housing market volatility for 50 states, District of Columbia, and the aggregate US economy, based on quarterly data covering 1975:3 and 2014:3. Given that existing studies have already shown housing sentiment to predict movements in aggregate and state-level housing returns, we will use a k-th order causality-in-quantiles test for our purpose, since this methodology allows us to test for predictability for both housing returns and volatility simultaneously. In addition, this test being a data-driven approach accommodates the existing nonlinearity (as detected by formal tests) between volatility and sentiment, besides providing causality over the entire conditional distribution of (returns and) volatility. Our results show that barring 5 states (Connecticut, Georgia, Indiana, Iowa, and Nebraska), housing sentiment is observed to predict volatility barring the extreme ends of the conditional distribution. As far as returns are concerned, except for California, predictability is observed for all of the remaining 51 cases.
In the fourth paper we investigate the impact of uncertainty shocks on the United States housing market using the time-varying parameter vector autoregression (TVP-VAR) following Mumtaz and Theodoris (2018). We will use quarterly time-series data on real economic activity, price, financial and housing market variables, covering the period 1975:Q3 to 2014:Q3. Besides housing prices, we also consider variables related to home sales, permits, starts, as well as housing market sentiment. In general, the results of the cumulative response of housing variables to a 1 standard deviation positive uncertainty shock at the one-, four- and eight quarter horizon tends to change over time, both in terms of sign and magnitude, with the uncertainty shock primarily affecting home sales, permits and starts over short-, medium and long-runs, and housing sentiment in the medium-term. Interestingly, the impact on housing prices is statistically insignificant.
Our final paper applies Bayesian Additive Regression Trees (BART) to study the comovement of REIT returns with expected and unexpected inflation using U.S. monthly data covering the sample period 1979 2016 and survey data to decompose inflation into an expected and unexpected component. Our findings show that the two inflation components are not among the leading predictors of REIT returns in terms of their relative importance, but also that the marginal effects of the two inflation components for REIT returns changed over time. REIT returns exhibit an asymmetric response to unexpected inflation, a phenomenon mainly concentrated in the Greenspan era.