Aggregate Calibration Can Mask Stress-State Undercoverage: Bounded-Memory Forecast Intervals under Recurrent Financial Stress
Abstract
Sequential forecast intervals are often evaluated by aggregate coverage even when calibration memory is bounded. This can be misleading when economically important states recur after long absences: a recent-only buffer may satisfy its overall coverage target while no longer containing errors from earlier episodes of the returning state. We study this problem in next-day financial range forecasting, holding the forecasting model, conformity score, adaptive update, and memory budget constant while varying the historical residual information available at each query. Recent-only first-in, first-out (FIFO) calibration is near nominal in aggregate but undercovers substantially during high-volatility states in both an exchange-traded fund development panel and a fixed stock panel. An unbounded state-history reference yields a lower observed stress-period interval score; Rare replay yields lower high-VIX interval scores than the matched Recent-n and Hybrid-n controls in both panels, and the state-structured versus recent-only contrast is largest near stress re-entry. Bounded calibration memory is therefore part of the forecasting information design, and aggregate calibration alone can conceal state-specific reliability failures.
Related
- FRESH: Information-Geometric Calibration of Patient-Level Models to Aggregate Evidence
- Beyond Aggregate VARs: A Bayesian Benchmark for HANK Models
- Aggregate Disambiguation Systems
- Learning Behavioral Parameters of Network Equilibrium Models from Aggregate Mobility Data
- A primer on expert elicitation with IDEA (Investigate Discuss Estimate Aggregate) for rare disease trialists
BibTeX
@misc{han2026aggregate,
title = {{Aggregate Calibration Can Mask Stress-State Undercoverage: Bounded-Memory Forecast Intervals under Recurrent Financial Stress}},
author = {Ruihang Han and Li-Hsiang Lin},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.04223},
eprint = {2611.04223},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.04223},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Aggregate Calibration Can Mask Stress-State Undercoverage: Bounded-Memory Forecast Intervals under Recurrent Financial Stress
AU - Ruihang Han
AU - Li-Hsiang Lin
PY - 2026
DA - 2026/11/03
PB - arXiv
DO - 10.48550/arxiv.2611.04223
UR - https://arxiv.org/abs/2611.04223
AB - Sequential forecast intervals are often evaluated by aggregate coverage even when calibration memory is bounded. This can be misleading when economically important states recur after long absences: a recent-only buffer may satisfy its overall coverage target while no longer containing errors from earlier episodes of the returning state. We study this problem in next-day financial range forecasting, holding the forecasting model, conformity score, adaptive update, and memory budget constant while varying the historical residual information available at each query. Recent-only first-in, first-out (FIFO) calibration is near nominal in aggregate but undercovers substantially during high-volatility states in both an exchange-traded fund development panel and a fixed stock panel. An unbounded state-history reference yields a lower observed stress-period interval score; Rare replay yields lower high-VIX interval scores than the matched Recent-n and Hybrid-n controls in both panels, and the state-structured versus recent-only contrast is largest near stress re-entry. Bounded calibration memory is therefore part of the forecasting information design, and aggregate calibration alone can conceal state-specific reliability failures.
ER -