---
title: Shared Causal Manifolds for Risk Management
url: https://www.ml-quant.com/papers/ssrn/4637990/
site: ML-Quant (https://www.ml-quant.com)
updated: 2026-09-26
license: Summaries CC BY 4.0; links go to the original sources
index: https://www.ml-quant.com/llms.txt
identifier: SSRN 4637990
source_url: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4637990
featured: 2023-11-29
citations: unknown
topic: Risk, Credit & Banking
---


# Shared Causal Manifolds for Risk Management

A finance webinar presented a machine learning-based framework for optimizing portfolio sensitivities and predicting future positions.

- Source: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4637990
- Identifier: SSRN 4637990
- Released: 2023-11-18
- First featured: Quant Letter No. 27 (2023-11-29): https://www.ml-quant.com/issues/2023-11-29/
- Citations (Semantic Scholar): not tracked
- Published in: not yet
- Topic: Risk, Credit & Banking

## Related

- [Defining and comparing SICR-events for classifying impaired loans under IFRS 9](https://www.ml-quant.com/papers/arxiv/2303.03080/): The paper presents a new framework for predicting credit deterioration using three parameters, validated using South African mortgage data.
- [Joint model for longitudinal and spatio-temporal survival data](https://www.ml-quant.com/papers/arxiv/2311.04008/): Longitudinal and Spatio-Temporal Survival Model: The Spatio-Temporal Joint Model (STJM) is a new method for credit risk analysis that uses spatial and temporal data to predict a borrower's risk, showing better results when spatial data is included.
- [Law-invariant return and star-shaped risk measures](https://www.ml-quant.com/papers/arxiv/2310.19552/): The paper introduces new characterizations for law-invariant star-shaped functionals, demonstrating their wide use in finance, insurance, and probability scenarios.
- [Model Aggregation for Risk Evaluation and Robust Optimization](https://www.ml-quant.com/papers/arxiv/2201.06370/): The model aggregation (MA) approach is a new method for risk evaluation that provides a robust value and distributional model, refining Value-at-Risk and Expected Shortfall characterizations.
- [Neural Networks for Insurance Pricing with Frequency and Severity Data: A Benchmark Study from Data Preprocessing to Technical Tariff](https://www.ml-quant.com/papers/arxiv/2310.12671/): The research evaluates different models for insurance claim frequency and severity data, presenting combined actuarial neural networks (CANNs) that merge baseline predictions with a neural network correction, and recommends using global surrogate models for practical deployment.
- [Comparative Evaluation of Anomaly Detection Methods for Fraud Detection in Online Credit Card Payments](https://www.ml-quant.com/papers/arxiv/2312.13896/): A study found that LightGBM was the best for fraud detection when comparing anomaly detection and standard supervised learning methods, but it was more susceptible to distribution shifts, questioning the advantage of combining these two methods.