A Semi-Markov Approach to Financial Modelling During the COVID-19 Pandemic

We assess the capabilities of the weighted-indexed semi-Markov chain (WISMC) model applied to high-frequency financial data during the COVID-19 pandemic which was characterised by periods of extreme volatility. In particular, we test the ability of the WISMC model to reproduce the typical stylised facts of the financial time series, such as the persistence of volatility. For a general analysis, we apply the model to three major indexes of the financial markets, i.e. the Standard & Poor 500 (SPX), the Dow Jones Industrial Average (DJI) and the Financial Times Stock Exchange 100 (FTSE) over a period that covers the first year of the COVID-19 pandemic, from January 2020 to December 2020. Moreover, we compare the results with the standard GARCH model. A Monte Carlo simulation shows that the WISMC model is able to reproduce the persistence of volatility and clearly outperforms the GARCH model in three different specifications.