Asymmetric correlation in S&P500 relationship with stock’s characteristics

Abstract

Abstract This research investigates the dynamics of asymmetric conditional correlation between individual equities and the broader market of S&P500 during periods of financial distress. Utilizing a bivariate ARMA(1,1)-aDCC-GARCH econometric framework over a comprehensive 4,000-day sample, we extract standardized innovations to model time-varying correlations and isolate the asymmetry parameter g triggered by joint negative market shocks. The time-series results confirm the presence of highly significant asymmetric correlation, demonstrating that equities like many stocks experience severe correlation spikes with the S&P 500 during market downturns, thereby deteriorating theoretical diversification benefits precisely when investors need them most. Furthermore, a cross-sectional Ordinary Least Squares (OLS) regression analysis is conducted across multiple sectors including Financials and a targeted high-asymmetry cohort to determine if firm-specific fundamentals (Market Capitalization, Book-to-Market ratio, and Debt-to-Equity leverage) explain the magnitude of the panic response g. The empirical cross-sectional findings reveal a lack of statistical significance and low explanatory power across these variables, indicating that internal balance sheet metrics do not reliably predict a firm's correlation breakdown. We conclude that asymmetric correlation is primarily driven by exogenous macroeconomic shocks and systemic investor behavior, highlighting the critical need for dynamic risk management strategies beyond traditional static portfolio diversification.