Investigation of dynamic volatility relationships among global indices in the presence of COVID-19 impact: Evidence from TVP-VAR approach


AKKUŞ H. T., MEMİŞ H., ŞARKAYA İÇELLİOĞLU C.

JOURNAL OF COMPUTATIONAL AND APPLIED MATHEMATICS, vol.473, 2026 (SCI-Expanded, Scopus) identifier identifier

Abstract

This study aims to examine the dynamic connectedness relationship between global economic and financial indices. In this context, the volatility relationships among Bitcoin, oil, gold, S&P 500 index, US 10-year bond interest, and the US dollar index are examined over the period from April 24, 2015, to March 19, 2024. Studies using such different types and numbers of variables are quite limited in the literature. The TVP-VAR approach with a time-varying covariance structure developed by Antonakakis and Gabauer [1] is used as the analysis method in the study. As a result of the study, it was determined that the relevant variables, the US 10-year bond rate and S&P500 variables, emit net volatility to other variables; oil, gold, US dollar index and Bitcoin variables have also been found to be net volatility receivers. On the other hand, the fact that Bitcoin has less volatility relationship with other variables shows that Bitcoin can provide potential benefits in terms of portfolio diversification with relevant variables. In addition, since the total volatility value among the variables is low, the relevant variables can be used together in international portfolio diversification. This situation is important for decision-makers on issues such as asset pricing, portfolio management, and risk management.