Implications of spatial modeling to issues in financial economics and real estate.

摘要

The importance of physical location to numerous issues in real estate finance is nearly common sense but has frequently been overlooked, modeled improperly, or underestimated. Location can serve as a proxy to numerous important and influential variables that we can never hope to measure or account for. In this dissertation I extend state-of-the art non-parametric statistical methods to explicitly model the spatial distribution of real estate properties and address two of the most important issues in real estate finance and economics---real estate valuation and mortgage prepayments.; The dissertation is organized in three chapters. The first one examines the relevant spatial statistics methods and suggests some necessary extensions and adaptations. The second chapter explores the issues of residential real estate valuation and provides an application for the methodology developed in Chapter 1. The third chapter investigates the determinants of mortgage prepayments and defaults decisions.