Tail Parameters of Stable Distributions Using One Million Observations of Real Estate Returns from Five Continents

Abstract

This study focuses on global real estate return distributions. For our analysis, we employ the class of stable distributions that has become prominent in the real estate literature. We add to the literature by undertaking a global-scale analysis for the first time. By using data since the early 1990s, we show that there is considerable variation in the tail weights of return distributions, both between countries as well as among sectors within the countries. It is important to note that the tail parameters vary over time as well. Our results strengthen the notion of non-constant tail parameters in stable distributions that followed earlier findings of constant tail parameters. In addition, our results provide evidence that it is merely the time horizon that causes variation in parameters, than purely methodological differences.