Effectiveness of Different Value at Risk Models in Predicting Stock Prices during Covid-19: The Case of Colombo Stock Exchange

Abstract

This paper assesses the predictive capabilities (the performance and the reliability) of three widely used Value-at-Risk (VaR) models in forecasting stock values before and during a major economic disruption. Utilizing trading data from companies listed on the Colombo Stock Exchange in Sri Lanka, the research compares the performance of Parametric VaR, Historical Simulation VaR, and Monte Carlo VaR models across two distinct periods - the pre-Covid-19 era and the volatile pandemic market conditions. The analysis delves into how effectively these VaR methodologies capture and anticipate market volatilities and potential losses, offering valuable insights for financial risk management practices. The dataset is divided into three periods: pre-Covid-19 (01 February 2019 to 11 March 2020), during Covid-19 (11 March 2020 to 15 June 2021), with 270 observations for each period, and finally considering both periods. This split enables a balanced analysis of VaR model performance across different market conditions. By evaluating the model outputs across eight financial assets, including individual stocks, indices, and a diversified portfolio, the paper provides a comprehensive assessment of their strengths and limitations in turbulent market environments. The study's findings reveal that, based on backtesting, the Historical Simulation VaR model outperforms other models in predicting market risk. The Parametric VaR model shows moderate effectiveness, while the Monte Carlo VaR model proves to be the weakest in capturing market volatility. The findings shed light on enhancing the usability and applicability of VaR analysis during periods of extreme uncertainty, such as the Covid19 pandemic. The paper aims to empower risk managers, investors, and financial professionals with data-driven strategies to mitigate risks and make more informed investment decisions, even in the face of unprecedented market conditions.