Recalibrating Tail Risk Forecasts under Temporal Dependence
-
- 0 Rating
- 0 Reviews
- 4 Students Enrolled
Recalibrating Tail Risk Forecasts under Temporal Dependence
Model-Free Conformal VaR Correction
-
- 0 Rating
- 0 Reviews
- 4 Students Enrolled
Requirements
- Stat Fin Markets
General Overview
Description
We audit six zero-shot Time Series Foundation Models and four classical benchmarks across 24 assets using a one-parameter conformal shift, whose magnitude separates forecasters that retain tail information from those whose Basel coverage is restored by the calibration layer alone. Four of the six foundation models fall into the latter regime; a within-family Moirai contrast attributes the gap to the predictive interface. Rolling recalibration delivers near-uniform Basel Green Zone compliance, but fitted GARCH retains a per-asset Quantile Score advantage that zero-shot deployment cannot match.
Recommended for you
Meet the instructors !
.jpg)
Daniel Traian Pele is a Prof. dr. Department of Statistics and Econometrics Faculty of Cybernetics, Statistics and Economic Informatics, The Bucharest University of Economic Studies. https://scholar.google.com/citations?user=tN32HYcAAAAJ&hl=en