Problem of the fat tails of yield distribution for globally diversified portfolios

Abstract

In this article we discuss the non-normality of the distribution of portfolio returns in global investments and presence of the “fat tails” problem. Various global investment strategies in 2006-2010 are considered. We analyze the global diversification strategies on separate markets of developed countries (components of MSCI EAFE index), emerging countries (components of MSCI EEM index), as well as geographic diversification use of those indices, MSCI ACWI all-country index, MSCI BRIC index. Additionally, we created an actively managed portfolio investing in the U.S. and BRIC markets and in precious metals. We conduct fit tests between the actual returns distribution and several theoretical distributions using the Pearson’s chi-squared test and the Kolmogorov-Smirnov test, and both criteria indicate non-normality of returns for all the portfolios we have analyzed. Thus, it is not possible to solve the “fat tails” problems by using only the geographic diversification. Using the statistical tests we show that this problem can be addressed using the Laplace distribution and the Student’s t-distribution.