Return and Volatility Transmission in U.S. Housing Markets

Abstract

This article uses the Case‐Shiller U.S. Home Price Indices to analyze spatial dependencies across 16 metropolitan markets for the period January 1989 to June 2006. Return transmission patterns establish New York, San Francisco and Miami as among the most influential markets. In terms of volatility linkages, there is a considerable amount of transmission in the East between New York, Boston and Washington, DC, and innovations in the housing markets of Miami, Los Angeles and San Francisco play an influential role within their respective regions. In comparison, markets in the Central and Mountain regions appear to be relatively independent from external influences. Overall, the linkages appear to be more intensive during the active phase of the real estate market (1999–2006) than during the calm phase (1989–1998).

The spatiality of housing market volatility and selective migration

“…In more recent work, Gray (2011) explored district-level house price movements in England and Wales in relation to a 'regional ripple effect'. In the USA, Miao et al (2011) find that intermetropolitan linkages transmit price effects more readily during market upswings.…”

“…Yet, where they are strong, local linkages can lead to intra-regional price transmissions (Jones and Leishman, 2006). Observations such as Meen's (1999) of the ripple effect in the UK and, more recently, Miao et al's (2011) work in the USA suggest that, at a more macroscopic level, transmission of housing market 'shocks' occurs spatially in response to a number of factors. Such spatial links arise owing to a number of dynamic factors on both the demand and the supply sides of the housing market.…”

“…Hamnett (1989) describes a set of countercyclical conditions, led by London and the south-east, that help to explain variations in house price change between regions at any point in time. The idea of spatiotemporal transmission processes such as lags and 'ripple effects' has been developed to reconcile cycles with cross-sectional price asymmetries (Meen, 1999;Miao et al, 2011). While agreeing that regional differentials have become more important, Drake (1995) importantly finds that the idea of a ripple effect spreading out from the south-east of England is tempered by the development of structural breaks in outlying regions such as Wales and the north of England.…”

Ripple effect and contagious effect in the US regional housing markets

“…Previous works have investigated ripple effect using various econometric methodologies, such as unit root test (Cook 2005;Holmes 2007), Vector Autoregressive (VAR) model Miller 2009, 2010;Pollakowski and Ray 1997), Generalized AutoRegressive Conditional Heteroskedasticity (GARCH) model (Miao et al 2011;Zhu et al 2012) and Granger causality . Recently, Copula methodology has received attentions and been widely applied in finance research.…”

“…The results showed that comovements existed only in geographical proximity and affected only nearby regions, whereas the spillover effect was not limited by proximity and extended to any market characterized by economic similarities and comparable mortgage market conditions. Miao et al (2011) suggested that the ripple effect was an economic phenomenon supporting the presence of spatial dependencies. They posited that the transmission of return and volatilities across metropolitan housing markets originated from common information or information spillover.…”