Weekly hedonic house price indices: an imputation approach from a spatio-temporal model

Hill, Robert J., Rambaldi, Alicia N. and Scholz, Michael (2017). Weekly hedonic house price indices: an imputation approach from a spatio-temporal model. In: The 4th Conference of the International Association for Applied Econometrics. Conference of the International Association for Applied Econometrics, Sapporo, Hokkaido, Japan, (). 26-28 June 2017.

Author Hill, Robert J.
Rambaldi, Alicia N.
Scholz, Michael
Title of paper Weekly hedonic house price indices: an imputation approach from a spatio-temporal model
Conference name Conference of the International Association for Applied Econometrics
Conference location Sapporo, Hokkaido, Japan
Conference dates 26-28 June 2017
Convener Siem Jan Koopman
Proceedings title The 4th Conference of the International Association for Applied Econometrics
Publisher Editorial Express
Publication Year 2017
Year available 2017
Sub-type Fully published paper
Open Access Status Not yet assessed
Total pages 23
Total chapters 372
Language eng
Formatted Abstract/Summary
Since the global financial crisis there is an increased demand for timely house price indices. The aim of this paper is to develop a method for computing house price indices at a weekly frequency using the hedonic imputation method. The hedonic imputation method provides a flexible way of constructing quality-adjusted house price indices using a matching sample approach. At annual frequencies the implementation of the hedonic imputation approach typically entails estimating the hedonic model period-by-period and then using the parameter estimates (i.e., characteristics shadow prices) to obtain the required imputed house prices. Once these imputed prices are available for a matched sample, standard price index formulas (e.g., Laspeyres, Fisher or T¨ornqvist) can be used to compute the overall price index. A common approach to control for location in hedonic models has been to include postcode dummies. This may not be feasible at higher frequencies as there may not be enough observations for each postcode and small ∗This project has benefited from funding from the samples might cause large variability in the shadow price parameters when estimated period-by-period. We develop a spatio-temporal model to obtain the imputed prices. A geospatial spline surface controls for location and is embedded in a state-space formulation that controls for trends and property quality. The advantage is that the model is parsimonious and shadow price parameters are connected over time while retaining the property that values are not revised as new time periods are added to the data set. We show the spatio-temporal specification leads to a modified form of the Kalman filter and a Goldberger’s adjusted form of the predictor to obtain the imputations. Using a recently developed measure of index performance and applying this hedonic geospatial spline/Kalman filter approach to data for Sydney (Australia) we show that it outperforms competing alternatives for computing house price indices at a weekly frequency. Furthermore, we show that weekly house price indices are much more sensitive than annual or quarterly indices to the choice of hedonic method. Hence the choice of hedonic method is of greater practical significance for weekly indices.
Keyword Housing market
House Price index
Hedonic imputation
Geospatial data
Spline
Quality adjustment
State Space Models
Q-Index Code E1
Q-Index Status Provisional Code
Institutional Status UQ
Additional Notes https://editorialexpress.com/conference/IAAE2017/program/IAAE2017.html; https://editorialexpress.com/cgi-bin/conference/download.cgi?db_name=IAAE2017&paper_id=136

Document type: Conference Paper
Sub-type: Fully published paper
Collections: HERDC Pre-Audit
School of Economics Publications
 
Versions
Version Filter Type
Citation counts: Google Scholar Search Google Scholar
Created: Wed, 29 Nov 2017, 16:13:10 EST by Alicia N. Rambaldi on behalf of School of Economics