Authors
Christiane Baumeister, Florian Huber, Massimiliano Marcellino
- Acknowledgements & Disclosure
- We thank Jim Hamilton for useful suggestions, and participants at the IWEEE 2024 Workshop and the 2024 IIASA Workshop on Agricultural Commodity Prices for helpful comments. Marcellino thanks MUR- Prin 2022- Prot. 20227YZ9JK, financed by the European Union - Next Generation EU, for partial financial support. Huber gratefully acknowledges funding from the Austrian Science Fund: ZK-35. The views expressed herein are those of the authors and do not necessarily reflect the views of the National Bureau of Economic Research.
- DOI
- https://doi.org/10.3386/w32524
- Published in
- United States of America
Table of Contents
- Introduction 3
- Alternative Approaches to Modeling Tail Risks in Oil Price Dynamics 5
- Bayesian Vector Autoregressions 5
- Bayesian Quantile Regressions 6
- Bayesian Additive Regression Trees 7
- Capturing Heteroskedasticity 8
- The Role of Nonlinearities in Oil Markets 9
- A Comprehensive Set of Economically-Motivated Predictors 9
- An In-Sample Primer on Nonlinear Relationships 10
- (Tail) Forecasting the Real Price of Oil 12
- Forecasting Environment 13
- Forecast Evaluation 14
- Point Forecasts 14
- Tail Forecasts 14
- Tracking the Forecasting Performance over Time 15
- Monitoring Oil Price Risks in Real Time 18
- Real-Time Assessment of Tail Risks 18
- The COVID-19 Pandemic and the Russia-Saudi Oil Price War 21
- The Russian Invasion of Ukraine and the Energy Crisis 21
- The Israel-Hamas War and the Red Sea Crisis 22
- Tail-Risk Indicators for Producers and Consumers 22
- Risk Scenarios for 2024 25
- Conclusions 27
- Illustration of a Tree-Based Model 29
- Prior Specifications and Bayesian Inference 30
- Priors for regression coefficients 31
- Priors for factor loadings 32
- Priors for stochastic volatilities 32
- Priors for trees 32
- Prior for scale parameter at quantile 33
- Bayesian inference 33
- Data Sources and Backcasting 34
- Results for the Full Hold-Out Period 34