A0669
Title: A clustering approach for analysing the impact of COVID-19 on stock market volatility
Authors: Jorge Caiado - University of Lisbon (Portugal) [presenting]
Franscisco Santos - ISEG University of Lisbon (Portugal)
Abstract: The COVID-19 impact on U.S. stock market volatility with a focus on 11 S\&P 500 sectors is investigated. For this purpose, we introduced a model feature-based method for clustering financial time series that accounts the useful information about the dependence structure of their conditional volatilities. This clustering approach consists in fitting parsimonious threshold GARCH models to the sector stock returns and then computing the distance matrix between the autocorrelations of their estimated conditional variances. By using hierarchical and non-hierarchical clustering methods, we conclude that there is a clear change in the composition of each cluster from the period before the first U.S. COVID-19 case to the period during the pandemic, leading to the conclusion that the similarities or distances between sectors have undergone a significant change and the industries most affected by the pandemic were the Hotels, Automobile and Airline.