Estimación de precios hedónicos de arrendamiento para apartamentos administrados por la sociedad de activos especiales (SAE) en estratos socioeconómicos 4, 5 y 6 ubicados en la localidad de Usaquén, Bogotá

Abstract

This study was conducted as part of an internship at the Special Assets Society (SAE), an entity responsible for the administration and management of real estate with economic and social purposes. The main objective of this research is to estimate the rental value of apartments located in the Usaquén neighborhood (Bogotá D.C.) by applying a hedonic pricing model, incorporating spatial econometric techniques to identify the influence of physical and location attributes on rental value. To build the database, web scraping techniques were implemented to collect information on the real estate market, which was then georeferenced and supplemented with spatial layers (shapefiles) that allowed for the integration of variables related to the urban environment, accessibility, and service provision. Initially, a multiple linear regression model (OLS) was estimated to establish a baseline, and then different spatial regression models (SAR, SEM, SDM, SARAR, and GNS) were evaluated, finally selecting the general spatial nesting model (GNS) for its better fit and explanatory power in the face of the spatial autocorrelation detected in the residuals. Finally, a geographically weighted regression (GWR) model was applied to identify the spatial variability of local coefficients and understand the spatial heterogeneity of the rental market in the study area.