An RNN Model for Exploring the Macroeconomic and Financial Indicators in the Context of the COVID-19 Pandemic

Abstract

A recurrent neural network (RNN) was developed to explore the impact of the COVID-19 pandemic on the stochastic macroeconomic and financial indicators of the yield spread (Spread), crude oil prices (Oil), recession (USrec), and two stock market indices of the volatility index (VIX) and Wilshire 5000 total market index (Wil5000). A time-series dataset was obtained from the Federal Reserve Bank of St. Louis (for Jan/2/1990 - Feb/2/2022). For each indicator, separately, the dataset was partitioned into 'before' and 'during' the pandemic using the dataset's breakpoint (transition block) established based on the indicator behavior. The results revealed: (a) VIX was explained by Wil5000, Spread, Oil, and USrec more accurately before the pandemic, indicating that other observed and unobserved factors arising from the COVID-19 pandemic would affect the VIX more than the macroeconomic and financial indicators, (b) USrec is predicted less accurately compared to other indicators during the pandemic, which shows the sensitivity of this indicator to health and geopolitical challenges, and (c) effects of the indicators on the bond market diminished during the pandemic. The sensitivity analysis suggests that the results remain highly robust to the changes in the number of records chosen for the different test sets.