Presented by Tatevik Zohrabyan

1
ANALYSIS OF U.S. REGIONAL HOUSE PRICE DYNAMICS

• Presented by Tatevik Zohrabyan
• Co-authors David Leatham and David Bessler
• Department of Agricultural Economics
• Texas AM University
• NC-1014 Conference
• October 4, 2007

2
INTRODUCTION

• Real estate is an important component of the
  personal consumption which comprises about 70 of
  the U.S. GDP.
• In addition, over the short period of time
  residential investments often account for larger
  share of GDP changes mainly due to their high
  volatility.
• Analysis of housing market would be of great
  interest to households from both rural and urban
  communities, government, and everyone in general
  by providing clues about the short-run
  performance of the economy and possible changes
  in financial conditions.

3
OBJECTIVE

• The primary objective is to study the dynamic
  interrelationships and causal structure of the
  U.S. regional house prices.
• The second objective is to provide a new approach
  for identification of the long-run structure.
• The third objective is to provide inferences
  about price discovery in the US housing market by
  identifying possible reasons for the obtained
  results.

4
CONTRIBUTION

• Many studies that closely relate to this study
  are conducted on the UK dataset. The most
  relevant articles are by MacDonald and Taylor
  (1993) and Alexander and Barrow (1994).
• Only Pollakowski and Ray (1997) have studied the
  dynamics of US house prices using census division
  data.
• The contribution of this paper to the existing
  housing market literature includes
• Updated dataset including various important
  dates.
• New and more sophisticated methodology for
  analyzes.
• Proposal for a new method for identification of
  the long-run structure.

5
DATA

• Quarterly house price indexes for nine U.S.
  census divisions are used. Time period ranges
  from 19751 to 20061
• The nine U.S. census divisions are
• Pacific (PC)
• Mountain (MT)
• West North Central (WNC)
• West South Central (WSC)
• East North Central (ENC)
• East South Central (ESC)
• South Atlantic (SA)
• Middle Atlantic (MA)
• New England (NE)

6
HPI TIME SERIES PLOT
7
METHODOLOGY

• VAR and later the VECM are used to investigate
  the dynamic structure of linkages among the nine
  series.
• The Directed Acyclic Graphs (DAG) are employed
  which provide the contemporaneous causal
  structure among the nine series and information
  for the identification of the long-run structure.
  
• Impulse Response Functions and the Forecast Error
  Variance Decomposition are utilized as well.

8
RESULTS

• In contemporaneous time, South and Middle
  Atlantics as well as the New England markets
  appear to be highly exogenous.
• On the contrary, only East South Central is
  information sink in the real estate market.

9
RESULTS

• New England, Pacific, West North Central, and
  Middle Atlantic are highly exogenous explaining
  most of the uncertainty in house prices of other
  regions.
• About 8.5 year ahead, less than 10 of variations
  in Mountains, South Atlantic, New England, West
  South Central, and East North Central are
  explained by the information arising in
  themselves.

10
RESULTS

• Pacific, New England, and West North Central
  generate significantly large positive responses
  from almost all the regions.
• Middle Atlantic and the West South Central
  generate significantly large negative responses
  from almost all other regions.
• Mountain and East North Central have nearly
  negligible influence on other regions.

11
CONCLUSION

• Dynamic and contemporaneous linkages among the
  house prices of nine US regions is uncovered in
  this study.
• Although in short-run the house prices of U.S.
  regions wander from each other, they move in a
  very similar fashion in the long-run.
• The results suggest that Pacific, Middle
  Atlantic, New England, and West South Central
  appear to be the highly exogenous leading mainly
  the house prices of East North Central and
  Mountain.

12
IMPLICATIONS

• Real estate market in US appears to mainly be led
  by regions that are influential in many other
  ways, such as financial, economic, etc.
• In addition, these regions tend to play a
  prominent role in terms of policy analysis in the
  housing market, such as creation of the safety
  net for the large banks.
• The relation between the social capital and the
  determinants of house prices might be blamed for
  such causality and price discovery results.

13
ANALYSIS OF U.S. REGIONAL HOUSE PRICE DYNAMICS
Presented by Tatevik Zohrabyan Co-authors David
Leatham and David Bessler Department of
Agricultural Economics Texas AM University

• THANK YOU !!!