Impact of COVID-19 on Stock Market Performance Using Efficient and Predictive LBL-LSTM Based Mathematical Model

Abstract

In this research work, Efficient Stock forecasting model using Log Bilinear and Long Short term memory (LBL-LSTM) is designed, considering external fluctuating factors to analyze impact of the pandemic COVID 19, on stock market performance using similar kind of historical records like past outbreaks of severe acute respiratory syndrome (SARS) virus occurred in 2002-2004. Earlier Machine learning (ML), based stock Market predictive models based on pre-epidemic reaction can become obsolete as theses models takes relatively longer historical training data as input parameters. These predictive ML models should be more robust to current happenings. Taking into account this problem, there is a need for model to be agile and flexible enough to sense external fluctuating factors. Model should be able to capture, adapt to rapid changes happening in economic dimension. Objective of this research work is to design and implement Log -Bilinear and Long Short term Memory based (LBL-LSTM) based model considering both long term and volatile short term external conditions. Experiments are conducted to analyze the performance of Proposed LBL -LSTM model, which shows significant performance improvement in terms of Root Mean Square Error (RMSE), Accuracy Score over existing ML based stock predictive models. Finally, after analyzing the effects conclusion can be made that markets will often react negatively to these incidents, in short term, but in the long term, the markets will eventually correct and improves.

Keywords: COVID-19, Sentiment Analysis, Stock Market Prediction, Long-Short Term Memory, Machine Learning Algorithms, Stock Market Indices, Log Bilinear Model

JEL Classification: C12