A1490
Title: Forecasting jointly value at risk and expected shortfall for energy commodities using quantile regression
Authors: Sjur Westgaard - Norwegian University of Science and Technology (Norway) [presenting]
Abstract: Expected shortfall (ES) offers advantages over value at risk (VaR) by addressing VaRs limitations in capturing extreme tail risks. While VaR estimates the maximum loss at a given confidence level, ES provides the average loss beyond this threshold, offering a more comprehensive measure of extreme risks. Consequently, ES is increasingly adopted in risk assessment frameworks, with the Basel Committee on Banking Supervision, recommending a shift from VaR to ES. Scenario analysis and stress testing are also crucial for assessing the impact of extreme events on financial institutions. Quantile regression models estimating VaR and ES jointly has been developed by past studies. The framework also facilitates scenario analysis and stress testing by investigating how changes in the risk factors directly influence VaR and ES at different quantiles. The R package "esreg" implements these methods, and "esback" offers joint VaR and ES prediction tests. The purpose is to explore these methods in estimating VaR and ES for energy commodities like oil, natural gas, electricity, and coal, and examine the influence of risk factors on tail risks.
