Real Estate Price Modeling and Empirical Analysis

Abstract

Real estate is an important part of asset markets, particularly in Japan, where the total monetary value of real estate amounts to two-thirds of national wealth. It comprises a similarly large fraction of the asset markets of most developed countries, indicating that the appraisal of real estate values is essential in ensuring steady and sustainable economic development. Despite the importance of real estate appraisal, real-world practices in the real estate and financial sectors are not consistent with academic theories, and even these theories fail to bridge the gap between the concepts presented in the hedonic pricing model and the concept of discounted cash flow valuation. From this point of view, our aim is to present a framework to resolve the gap between these concepts to help understand real estate pricing both in theory and practice.

We construct a theory of real estate pricing that is directly applicable to empirical analysis. Using dynamic portfolio optimization, we first consider as a norm a model of theoretical equilibrium prices of pieces of real estate in respect to attribute prices common to all pieces of real estate. Then, we investigate how we can extend the norm to more realistic pricing models. A logical consideration suggests the utility of introducing a mixed effect model developed with the application of the Box-Cox transformation. By using our model to analyze data obtained from Japanese Real Estate Investment Trust (J-REIT) records, we demonstrate our model’s ability to some extent.