🤖 AI Summary
This study addresses the causal inference challenges in credit loss forecasting for regulatory stress testing by proposing a causal panel prediction framework that explicitly separates components learnable from data from those reliant on untestable confounding assumptions. The approach integrates iterative regression, path-conditioned mean identification, causal set identification under bounded confounding, and recursive error analysis with importance-weighted conformal calibration to decompose predictive uncertainty into three interpretable layers. It innovatively disentangles estimation uncertainty from confounding uncertainty, yielding actionable outputs including robustness metrics, extrapolation cost diagnostics, an automatic abstention mechanism, and time-domain reliability bounds. The framework’s effectiveness and practical utility are validated through simulations and semi-synthetic experiments based on real unemployment data, including retrospective analyses of extreme scenarios such as the COVID-19 pandemic.
📝 Abstract
Regulatory stress testing requires projecting credit losses under hypothetical macroeconomic scenarios -- a fundamentally causal question typically treated as a prediction problem. We propose a framework for policy-path counterfactual inference in panels that transparently separates what can be learned from data from what requires assumptions about confounding. Our approach has four components: (i) observational identification of path-conditional means via iterated regression, enabling continuous macro-path contrasts without requiring a control group; (ii) causal set identification under bounded confounding, yielding sharp identified sets with interpretable breakdown values that communicate robustness in a single number; (iii) an oracle inequality showing that recursive rollout error is governed by a horizon-dependent amplification factor, providing a concrete answer to how far ahead one can reliably predict under stress; and (iv) importance-weighted conformal calibration bands with diagnostics that quantify extrapolation cost and trigger abstention when coverage guarantees degrade. The final output is a three-layer uncertainty decomposition that cleanly separates estimation uncertainty from confounding uncertainty. We validate all results through simulation and semi-synthetic experiments with real unemployment data, including a Covid retrospective demonstrating the framework's diagnostic value under extreme scenarios.