Dynamic Dependence Among Economic Sectors in Equity Markets

Abstract

This study investigates the dynamic dependence structure between equity market sectors using a Markov regime-switching copula framework. The analysis focuses on return distributions across economic sectors and their lower-tail dependence, particularly across different market regimes. The proposed model builds on empirical evidence that market drawdowns can be contagious and that pairwise dependence between assets tends to increase during periods of market stress. Modeling dynamic dependence can provide a more robust risk framework for portfolio evaluation and asset allocation. In this setting, dependence is allowed to vary over time through regime shifts and is linked to market sentiment. This thesis extends Bubbles and dependence between international equity markets by Wuyi Ye, Lingbo Gao and Xiaoquan Liu (2024) by applying a similar framework to a different data sample. Specifically, the model is applied to equity market sectors (S&P 500 sub-indices: Industrials, Materials, Energy, Healthcare, Financials, Information Technology, and Consumer Non-Cyclical) rather than a geographic cross section, using daily returns over the period 1989-2026. Empirically, the conditional regime-switching copula provides a better fit for many index pairs than an unconditional regime-switching copula and a static copula benchmark. However, in an out-of-sample asset allocation exercise, we are unable to replicate the economic gains reported in Wuyi Ye, Lingbo Gao and Xiaoquan Liu (2024): portfolios based on the conditional model do not consistently achieve higher riskadjusted returns, evaluated by their Sharpe ratio, than equally weighted portfolios or portfolios constructed using a static copula model. Nonetheless, the model-based portfolios often exhibit lower maximum drawdowns, indicating that the framework can capture and mitigate some tail risk.