This paper proposes the use of a semiparametric model based on a locally weighted approach that controls for dynamic agglomeration and diffusion effects in constructing localized housing price indices. Based on residential transaction records in Singapore, we create three-dimensional interactive heat maps that allow for better measurement and visualization of spatial variations and heterogeneity in price appreciation. The heat map captures regional variations and temporal dispersions in price appreciation rates in the business cycle. Our methodology provides high-resolution information about price dynamics that could aid individual buyers, investors, and policy makers in making objective and informed decisions.
- heat map
- locally weighted estimation
- real estate price index
ASJC Scopus subject areas
- Environmental Science (miscellaneous)