Título:
|
Quantitative analysis of commercial and residential real estate markets (an approach from cointegration and spatial econometrics)
|
Autores:
|
Rodríguez Ramírez, Ramiro J.
|
Tipo de documento:
|
texto impreso
|
Editorial:
|
Universidad Complutense de Madrid, 2016-06-28
|
Dimensiones:
|
application/pdf
|
Nota general:
|
info:eu-repo/semantics/openAccess
|
Idiomas:
|
|
Palabras clave:
|
Estado = No publicado
,
Materia = Ciencias Sociales: Derecho: Propiedad inmobiliaria
,
Materia = Ciencias Sociales: Economía: Econometría
,
Tipo = Tesis
|
Resumen:
|
The first chapter of this thesis examines the formation process of residential prices in Spain (1995Q1 – 2012Q4). We propose two models to compare their performance in the context of comparative dynamics and predictive capacity. A structural model is derived from an eclectic theoretical framework in which we review published literature on the housing market and select a set of variables representative of this literature. We used GDP pre-capita, interest rates, the supply of new residential buildings and the gross residential-capital formation as explanatory variables for the average house price per square meter in Spain. The other model is generated by an algorithm known as GASIC2. Using our review of the literature we select a set of 46 variables, we form the respective database and let the algorithm to select the best model out the 2 (70 trillion) nested models. The condition imposed on the algorithm is to be parsimonious, i.e. having only 4 regressors. Annual theoretical effort of families to pay for their residence, the apparent concrete consumption, the mortgage interest rate and the real GDP are selected by GASIC to explain the average residential price in Spain; a similar model to the structural one. Our analytical framework is cointegration. Therefore, we assessed the integration order of both models’ variables. We identified all variables have order of integration of first degree (some with a structural break in the recent economic crisis). This leads us to test the hypothesis of cointegration. Proving such an existence, two error correction models (ECM) were estimated (one for the structural approach and one for the algorithmic) to calculate price and income elasticities, and produce dynamic forecasts. The long-term equations in both models behave similarly and give a good idea of the long-term equilibrium relationship between housing prices and their fundamentals. It is in the short term specification where the structural model and the algorithmic model differ. The model generated with GASIC has got a non-significant error correction mechanism, implying that the gap between the change in housing prices and longterm path is not traced. The consequence of such failure generates less accurate house price forecasts. However, the analysis of elasticities remains valid in both long and short term price equations...
|
En línea:
|
https://eprints.ucm.es/id/eprint/44236/1/T39100.pdf
|