A1125
Title: Can uncertainty affect extreme events in the oil market? A MIDAS touch to dynamic POT models
Authors: Katarzyna Bien-Barkowska - Poznan University of Economics and Business (Poland) [presenting]
Abstract: A novel econometric framework is introduced for forecasting extreme events in oil markets by incorporating macro-financial uncertainty indicators into a dynamic peak-over-threshold (POT) model within a mixed-frequency setting. Based on three decades of daily returns from two major global oil benchmarks, WTI and Brent, robust empirical evidence that rising uncertainty is significantly associated with the frequency and magnitude of extreme returns is provided. The Economic Policy Uncertainty index, Equity Market Volatility Tracker: Commodity Markets, and Jurado-Ludvigson-Ng 1-month ahead macroeconomic uncertainty index are utilized to capture different dimensions of uncertainty. These insights are derived through the novel framework, the autoregressive conditional MIDAS-POT (AC-MIDAS-POT), which integrates dynamic specifications for inter-exceedance times and magnitudes of extreme oil returns with the MIDAS components to account for macroeconomic and financial information. The AC-MIDAS-POT model significantly outperforms the VaR forecasts from the GARCH-MIDAS models with Gaussian and Student's t-distributions. This novel modeling framework offers a robust tool for forecasting tail risk in financial markets and provides valuable insights for investors and policymakers to forecast and mitigate the impact of extreme returns.
