---
title: Explainable Machine Learning for Basel and IFRS 9 Credit Risk: A Placebo-Controlled Test of Narrative Features
url: https://www.ml-quant.com/papers/ssrn/7622180/
site: ML-Quant (https://www.ml-quant.com)
updated: 2026-10-23
license: Summaries CC BY 4.0; links go to the original sources
index: https://www.ml-quant.com/llms.txt
identifier: SSRN 7622180
source_url: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7622180
featured: 2026-10-23
citations: unknown
topic: ML & AI Methods
---


# Explainable Machine Learning for Basel and IFRS 9 Credit Risk: A Placebo-Controlled Test of Narrative Features

A placebo test reveals that features like central bank tone can raise credit model accuracy merely by indicating when loans were made, not by predicting defaults.

- Source: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7622180
- Identifier: SSRN 7622180
- Released: 2026-10-17
- First featured: Quant Letter No. 136 (2026-10-23): https://www.ml-quant.com/issues/2026-10-23/
- Citations (Semantic Scholar): not tracked
- Published in: not yet
- Topic: ML & AI Methods
- Authors: Nawaz Muhammad

## Related

- [Dynamic Graph Neural Networks for Real Time Systemic Risk Surveillance: An Explainable AI Framework for Financial Stability](https://www.ml-quant.com/papers/ssrn/7522878/): The research proposes a temporal graph neural network with explainability tools for real-time systemic risk surveillance, achieving early warning signals 3-4 quarters ahead of financial distress on bank data.
- [Interpretable Machine Learning Recovery Rates](https://www.ml-quant.com/papers/repec/eee-jbfina-v-164-y-2024-i-c-s0378426624001043/): Machine learning methods offer better performance and insights in modeling corporate bond recovery rates than traditional methods.
- [Investment Grade Rating with Explainable AI](https://www.ml-quant.com/papers/ssrn/4535478/): The article explores the use of explainable AI in formulating rules on financial ratios to help companies enhance their credit ratings.
- [Jumping Ahead: Improving Reconstruction Fidelity with JumpReLU Sparse Autoencoders](https://www.ml-quant.com/papers/arxiv/2407.14435/): The paper presents JumpReLU Sparse Autoencoders (SAEs), which provide high-quality reconstruction of language model activations at a specific sparsity level, while maintaining interpretability.
- [Neuro-Symbolic AI in 2024: A Systematic Review](https://www.ml-quant.com/papers/arxiv/2501.05435/): Neuro-Symbolic AI has grown since 2020, focusing on learning and inference, but still lacks in areas like explainability, trustworthiness, and Meta-Cognition.
- [Year-over-Year Developments in Financial Fraud Detection via Deep Learning: A Systematic Literature Review](https://www.ml-quant.com/papers/arxiv/2502.00201/): The study examines progress in deep learning methods for detecting financial fraud, reviewing 57 studies from 2019 to 2024, and discusses challenges and opportunities such as data privacy, feature engineering, and model interpretability.