Reconstructing Emerging and Developed Markets Using Hierarchical Clustering

Abstract

The distinction between emerging and developed markets is of first-order importance for investors. In this article, the authors use hierarchical clustering to objectively identify the countries or regions that cluster from an investment viewpoint. They go beyond classifications based on economic fundamentals and group countries based on returns in equity and bond markets. The authors find an important geographical footprint that differs significantly from the groupings that are used by most practitioners. This analysis has practical implications for both active and index investors.

TOPICS: Statistical methods, simulations, big data/machine learning, emerging markets, developed markets

Key Findings

• • Unsupervised learning can help index investors and allocators develop more precise exposures to emerging stock and bond markets.

• • There are also alpha applications from studying clusters of countries that exhibit distance from one another.

• • We find a surprisingly strong geographic footprint, together with a grouping into resource rich and resource poor countries.