On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

168,695 papers · 148 categories

Trend · papers per month

count % of corpus

5 results for “CCAR”

SHARC: SHAP-Based Interpretability in Machine Learning Risk Models for Regulatory Capital under ICAAP and CCARq-fin.RM

SHARC explains machine learning risk models for regulatory capital, linking outputs to scenarios.

problem Inability to explain machine learning model outputs to regulatory bodies.

method SHAP-based explainability framework for Hybrid GPR-HS architecture and SVaR stress-testing.

result SHARC links SVaR outputs to scenario inputs, providing auditable traceability.

SHAP Machine Learning Regulatory Capital ICAAP CCAR

2026-07-06abs ↗pdf ↗

Forward-Looking Stress Testing Under Macro Scenarios: Stable SVaR Estimation Using a Hybrid GPR-HS Framework with SACSq-fin.RM

Paper improves SVaR estimation for stress testing under macro scenarios using a hybrid GPR-HS framework.

problem Numerical instability in traditional SVaR estimation under extreme shocks.

method Extends GPR-HS framework to forward-looking stress scenarios with SACS for stable covariance.

result Stable SVaR ranges from -2.1020% to -2.2231%, preserving coherence property.

Stress Testing SVaR Estimation Hybrid GPR-HS SACS Extreme Shocks

2026-05-26abs ↗pdf ↗

Using CPI in Loss Given Default Forecasting Models for Commercial Real Estate Portfolioq-fin.RM

The paper uses CPI growth rates to improve LGD predictions for CRE loans.

problem Challenges in forecasting LGD for CRE loans due to extended resolution times and restricted data.

method Combines internal and public data, including CPI growth rates, to forecast CRE LGD.

result Incorporating CPI at the time of default improves LGD prediction accuracy.

CPI growth CRE LGD LGD forecasting Commercial Real Estate Post-default

2024-02-23abs ↗pdf ↗

Time Series Simulation by Conditional Generative Adversarial Netstat.ML

Generative Adversarial Net (GAN) has been proven to be a powerful machine learning tool in image data analysis and generation. In this paper, we propose to use Conditional Generative Adversarial Net (CGAN) to learn and simulate time series data. The conditions can be both categorical and continuous variables containing…

2019-04-25abs ↗pdf ↗

A Hybrid Gaussian Process Regression Framework for Stable Volatility-Covariance Estimation: Evidence from Global Equity Indicesq-fin.RM

Paper proposes a new GPR-HS framework for accurate VCV estimation in global equity indices.

problem Accurate forecasting of Volatility-Covariance Matrix (VCV) for regulatory processes.

method Hybrid Gaussian Process Regression-Historical Simulation (GPR-HS) framework.

result GPR-HS framework achieves regulatory compliance and outperforms static VaR benchmarks.

Gaussian Process Regression Volatility-Covariance Matrix Value-at-Risk Expected Shortfall Historical Simulation

2026-05-17abs ↗pdf ↗