Examining the Impact of Covid-19 and Economic Indicators on US GDP using Midas- Simulation and Empirical Evidence

Abstract

In this study, we demonstrate the impact of the COVID-19 pandemic on the US GDP growth rate for the period 2000–2020. Other economic variables of different frequencies such as consumer price index, crude oil prices, exchange rate, unemployment rate, interest rate, export, import, government expenditure, and consumer confidence are considered in this study. In carrying out the analysis, we compare three approaches—the autoregressive distributed lag-unrestricted mixed data sampling (ADL-UMIDAS) model and autoregressive distributed lag-restricted mixed data sampling (ADL-RMIDAS) model with both exponential Almon and beta functions to the autoregressive distributed lag (ADL) model. Using a comprehensive simulation study, we examine the sensitivity of forecasting approaches to model misspecification. The efficiency of the ADL-UMIDAS model and ADL-RMIDAS model with both exponential Almon and Beta functions to the ADL model is computed throughout the simulation processes using the root mean square error (RMSE). Out-of-sample empirical and simulation analysis showed that the ADL-RMIDAS (Exponential Almon) model outperforms other competing models. The main finding shows that the COVID-19 pandemic has had a statistically significant positive effect on the US GDP growth rate.