This study aims to develop a financial stress index for Pakistan and figure out the stressful and non-stressful regimes for pre- and post-COVID-19 era by making use of available annual time-series data from January 1996 to December 2020.
Two state-of-the-art approaches are used to carry out empirical analysis. First, the generalized dynamic principal components analysis for constructing financial stress index, and second, Markov regime-switching model.
The empirical results support the hypothesis that economic and political risks play a key role in Pakistan’s greater degree of financial stress in addition to financial market stress. The banking sectors and stock market are found to be among the major contributors to financial stress in both pre- and post-COVID-19 eras. Furthermore, the estimation of the transition probabilities indicates that the model has a high probability of keeping its current state, resulting in only a few state shifts.
In contrast to the existing studies, which typically transform the indicators into a stationary state before analysis and the resultant index is likely to lose of data originality, this study makes use of state-of-the-art recently developed generalized dynamic principal components analysis to develop financial stress index. In addition, Markov regime-switching model is used in this study to distinguish between stressful and unstressful periods of the economy.
