Machine-learning Growth at Risk

2025-05-31

AI Summary

This paper addresses the challenge of identifying, attributing, and forecasting downside risks to U.S. economic growth. Conventional approaches struggle to disentangle sectoral contributions to aggregate risk; to overcome this, we propose a time-varying machine learning framework based on quantile partial correlation regression—enabling, for the first time, orthogonal decomposition and dynamic weighting of downside risks originating from financial, labor, and housing sectors. The method ensures both model selection consistency and temporal robustness, isolating sector-specific risk signals while yielding interpretable, sector-level downside risk indices. Empirically, these indices significantly improve out-of-sample forecasting accuracy for negative GDP shocks, outperforming standard macrofinancial indicators. By providing a structural, real-time monitoring tool for systemic vulnerability, the framework advances macroprudential policy design and implementation.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Game Theory and Economic Paradigms: Adversarial Learning
• Reasoning under Uncertainty: Stochastic Optimization

Application Category

• Economics, Online Markets and Human Computation: Sustainability of Web economics
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Web Mining and Content Analysis: Machine learning and data science for the Web

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.

Problem

• Analyze US growth vulnerabilities using machine-learning methods
• Identify financial, labor-market, and housing drivers of downside risk
• Construct sector-specific indices to predict isolated downside risks

Innovation

• Quantile partial correlation regression for model selection
• Sector-specific indices to predict downside risk
• Machine-learning method for time series consistency