A regional decomposition of US housing prices and volume: market dynamics and portfolio diversification

Nikos Antonakakis, Ioannis Chatziantoniou, David Gabauer

Research output: Contribution to journalArticlepeer-review

Abstract

In this study, we investigate the lead-lag relationship between housing prices and sales volume across four US regional housing markets; namely Midwest, Northeast, South, and West. To achieve this, we employ a time-varying parameter vector autoregressive framework of analysis that focuses on dynamic connectedness. We not only investigate how either prices or volumes independently co-move across regions but also, we provide evidence on how prices and volumes combined interact with each other across regions over time. In addition, considering the fact that the relevant connectedness index that emerges from our analysis can be used as a measure of risk, we proceed with the development of portfolios aiming to identify opportunities for reducing investment risk in the housing market. Main findings indicate that (i) all four regions can either transmit or receive shocks in the housing market with regard to prices and volume, (ii) during turbulent economic periods, it is sales volume shocks that drive developments in the US housing market, (iii) there is potential for effective portfolio diversification. Results have policy implications particularly considering the negative outcomes of overheated housing markets and are also relevant to investors and finance professionals for formulating effective portfolio diversification strategies.
Original languageEnglish
JournalThe Annals of Regional Science
Publication statusAccepted for publication - 1 Sep 2020

Keywords

  • US Real Estate Market
  • Transaction Volume
  • TVP-VAR
  • Dynamic Connected-ness
  • Regional Connectedness Decomposition
  • Portfolio Management

Fingerprint

Dive into the research topics of 'A regional decomposition of US housing prices and volume: market dynamics and portfolio diversification'. Together they form a unique fingerprint.

Cite this