A Markov-Based Economic Recession Modeling Through Financial Outcomes: Before and During the COVID-19 Pandemic

We analyze, design, implement, and validate a stochastic model for the economic recession using the financial outcomes of yield spread, crude oil prices, and stock market indices (volatility index and Wilshire 5000 total market index). The proposed model uses Markov Chain Model (MCM) and Hidden Markov Chain Model (HMCM) to (a) identify the impact of the COVID-19 pandemic on the recession by modeling the recession for two eras of “Before” and “During” the pandemic, (b) implement the “trend” analysis by forecasting the value of the financial outcomes for a date in the future and predicts the recession for that date using the forecasted values, and (c) entertain “what-if’ scenarios about the recession status for sudden drastic increases in the financial outcomes, one at a time. Results reveal that for case (a) the pandemic's onset disrupted the predictability of financial indices, for case (b) the correct future prediction of indices and the recession probability index for the “Before” dataset has a much higher margin over the “During” dataset, and for case (c) the yield spread has the highest total percentage of drastic and moderate changes among prediction of its values and prediction of the recession.