---
title: When Stress Tests Miss the Risk: Statistical Scenario Analysis for Financial Portfolios
url: https://www.ml-quant.com/papers/arxiv/2609.37273/
site: ML-Quant (https://www.ml-quant.com)
updated: 2026-10-02
license: Summaries CC BY 4.0; links go to the original sources
index: https://www.ml-quant.com/llms.txt
identifier: arXiv:2609.37273
source_url: https://arxiv.org/abs/2609.37273
featured: 2026-10-02
citations: unknown
topic: Portfolio & Allocation
---


# When Stress Tests Miss the Risk: Statistical Scenario Analysis for Financial Portfolios

Kernel scenario analysis estimates quantile predictions for portfolio stress-test gains, with online calibration to capture dependence between stressed and unstressed risk factors.

- Source: https://arxiv.org/abs/2609.37273
- Identifier: arXiv:2609.37273
- Released: 2026-09-30
- First featured: Quant Letter No. 133 (2026-10-02): https://www.ml-quant.com/issues/2026-10-02/
- Citations (Semantic Scholar): not tracked
- Published in: not yet
- Topic: Portfolio & Allocation
- Authors: Zhongze Cai, Martin B. Haugh, Xiaocheng Li

## Abstract (arXiv, CC0)

Scenario analysis is widely used to stress test financial portfolios, yet conventional approaches often summarize scenario gains using point estimates that overlook dependence between stressed and unstressed risk factors. We develop a machine learning framework for quantifying uncertainty in realized next-day scenario gains through prediction intervals whose widths can be calibrated online. Adaptive conformal scenario analysis (ACSA) calibrates scenario-specific quantile predictions and provides a long-run empirical coverage guarantee over realized scenarios. Our main method, kernel scenario analysis (KSA), estimates scenario-conditional quantiles directly from the specified stress and current market information. KSA can also be combined with ACSA to obtain online-calibrated prediction intervals. In experiments designed to reflect real-world markets, conventional scenario analysis can substantially understate risk, including for portfolios that appear safe under standard stress tests. Our proposed methods achieve better calibration and sharper intervals than empirical-quantile baselines. Overall, the framework moves scenario analysis beyond point estimates by quantifying predictive uncertainty and providing tools for statistically validating scenario gains.

## Related

- [The Critical Line Algorithm and the Constrained LASSO: One Curve, Two Literatures](https://www.ml-quant.com/papers/arxiv/2609.25704/): Shows that mean-variance portfolio selection and the constrained LASSO trace identical piecewise-linear solution paths, mapping their parametrizations exactly.
- [DeePM: Regime-Robust Deep Learning for Systematic Macro Portfolio Management](https://www.ml-quant.com/papers/arxiv/2601.05975/): Deep Learning for Portfolio Management: DeePM uses deep learning to improve macro portfolio management, delivering better risk-adjusted returns than traditional methods across various economic conditions.
- [Regulating Cash Holdings: Assessing Lost Returns in Mutual Funds](https://www.ml-quant.com/papers/ssrn/4478272/): Israeli mutual funds hold excessive cash, indicating a need for better liquidity management to reduce redemption risks.
- [Effective and Scalable Programs to Facilitate Labor Market Transitions for Women in Technology](https://www.ml-quant.com/papers/arxiv/2211.09968/): In Poland, cheap online portfolio challenges and one‑on‑one mentoring sharply increased women’s tech employment, and data-driven targeting improved admissions.
- [A mathematical study of the excess growth rate](https://www.ml-quant.com/papers/arxiv/2510.25740/): - Excess Growth - Excess Rate - Growth Excess - Surplus Growth - Overgrowth - Growth Surplus Recommended: Excess Growth (keeps meaning but is more concise).: The paper proves that a central portfolio metric—the excess growth rate—can be exactly described using basic information‑theory ideas and a few natural axioms. In short, it shows that the extra growth a portfolio achieves is essentially an information quantity, so portfolio performance can be understood like information gain.
- [Signed network models for portfolio optimization](https://www.ml-quant.com/papers/arxiv/2510.05377/): The study shows that using negative edges in weighted signed network representations of financial markets can help reduce portfolio risk, performing on par with traditional models.