COVID-19 and index returns of sub-continent: An exploration via Markov-Switching model

Aamir Azeem and Muhammad Zubair Akhtar

Abstract

This paper empirically investigates the relationship between stock market returns and COVID-19 cases. The dataset consists of the daily frequency of hallmark indices of Pakistan, India, and Bangladesh from January 2019 to November 2020. The methodologies applied in the paper are ARCH-GARCH and two-state Markov-Switching Model (MSM), a better approach for structural breaks, in all indices. The findings reveal that internal and external factors jointly contribute to the volatility of index returns in Pakistan and Bangladesh, but internal factors affect India's volatility transmission. MSM finds that the negativity of returns increases, while the positivity decreases in both states during the COVID-19 period with higher variance and low persistence in transition probability than the pre-COVID-19 phase. In the final MS model, COVID-19 cases depict significant relation with index returns in states1 while insignificant in state2. The probability of
staying state1 is not persistent, whereas state2 is persistent during COVID-19 crises with higher duration.
Keywords: Markov-Switching Model, Index Returns, Probability, COVID19 Cases, Duration.