Machine-learning Growth at Risk
Preprint
Abstract
We analyse growth vulnerabilities in the US using quantile partial correlation regression, a selection-based machine-learning method that achieves model selection consistency under time series. We find that downside risk is primarily driven by financial, labour-market, and housing variables, with their importance changing over time. Decomposing downside risk into its individual components, we construct sector-specific indices that predict it, while controlling for information from other sectors, thereby isolating the downside risks emanating from each sector.Keywords
All Related Versions
This publication has 26 references indexed in Scilit:
- A sorted penalty estimator: Inference for a correlation-robust shrinkage methodJournal of Econometrics, 2026
- Density forecasts of inflation: A quantile regression forest approachEuropean Economic Review, 2025
- Specification Choices in Quantile Regression for Empirical MacroeconomicsJournal of Applied Econometrics, 2024
- Nonlinearities in macroeconomic tail risk through the lens of big data quantile regressionsJournal of Applied Econometrics, 2023
- On the real‐time predictive content of financial condition indices for growthJournal of Applied Econometrics, 2022
- Nowcasting tail risk to economic activity at a weekly frequencyJournal of Applied Econometrics, 2022
- Backtesting global Growth-at-RiskJournal of Monetary Economics, 2021
- Penalized Time Series Regression2019
- Macroeconomic forecast accuracy in a data‐rich environmentJournal of Applied Econometrics, 2019
- Large Bayesian vector auto regressionsJournal of Applied Econometrics, 2010