Journal of Economics and Management Volume 22, No. 1 March, 2026

Asymmetric Volatility, Investor Sentiment, and Market Co-movements: An Analysis of Cryptocurrencies and U.S. Equities Using GARCH and LSTM Models

Ai-Chi Hsu

Finance, National Yunlin University of Science and Technology, Taiwan

Chia Hung Chen

Finance, National Yunlin University of Science and Technology, Taiwan

Chiu-Ming Hsiao

Finance, National Yunlin University of Science and Technology, Taiwan

Abstract

This study examines the dynamic co-movement, return spillovers, and volatility transmission mechanisms between major cryptocurrencies and the U.S. equity market before, during, and after the COVID-19 pandemic. In this study, co-movement refers to the degree to which asset prices or returns exhibit synchronous or mutually dependent fluctuations over time, reflecting the extent of information integration, investor behavior linkage, and systemic risk propagation across markets. Focusing on Bitcoin, Ethereum, Litecoin, and Ripple, we analyze how their return and volatility dynamics interact with the S&P 500 Index and investor sentiment indicators across different pandemic phases. The empirical findings reveal that, during the COVID-19 outbreak, return-based co-movement and spillovers among the four cryptocurrencies remain relatively weak, suggesting short-term market segmentation in return behavior. In contrast, volatility spillovers become substantially stronger, indicating heightened interdependence in risk transmission. These volatility spillovers originate from both own-market shocks and lagged cross-market influences, producing heterogeneous positive and negative effects on volatility across different assets. Moreover, the cryptocurrencies exhibit pronounced asymmetric behavior: negative shocks exert significantly greater effects on volatility than positive shocks. In the post-pandemic period, negative news continues to dominate the volatility co-movement patterns of cryptocurrencies, reflecting their increased sensitivity to U.S. stock market fluctuations. Further spillover analysis shows that both the S&P 500 Index and the VIX influence cryptocurrency returns through volatility transmission channels. Specifically, the S&P 500 exerts a negative unidirectional spillover effect, whereas the VIX demonstrates either unidirectional or bidirectional transmission dynamics, highlighting the role of investor fear and sentiment in shaping cross-market linkages.

Keywords：COVID-19 Pandemic, Cryptocurrencies, Co-movement, VIX, Volatility Spillover

JEL Classifications：G17

BACK