Francesco
Castelli
Machine learning and banking regulation, theory and practice.
📍 Zürich
Experienced manager with 12 years building and auditing credit risk models.
I am passionate about banking regulations, machine learning techniques, and the gap between what models say and what banks actually decide.
The website is organized as follows:
Latest Banking Regulation Posts
Banking Regulation
15. The IRB Shortfall: where IFRS 9 provisions meet CET1
Banking Regulation
14. High-Quality Liquid Assets: the three levels, the haircuts and the caps
Banking Regulation
13. The Countercyclical Buffer: the capital requirement that follows the credit cycle
Banking Regulation
12. Maximum Distributable Amount
Latest Machine Learning Posts
Machine Learning
Discriminatory Power: ROC, AUC and the Gini coefficient
Machine Learning
Population Stability Index: the test that tells you when your data has moved
Machine Learning
Weight of Evidence: turning each variable into risk prediction
Machine Learning
Correlation: one relationship, four different numbers
Notes
Breakdowns of banking regulation, and machine learning topics. One concept per entry, from first principles to the practical implication.
Two streams:
Banking regulation: Basel, CRR3, IRB, IFRS 9, stress testing, liquidity concepts.
Machine learning: models, regularization and the techniques that are relevant in a real risk management work.
Tools
Interactive utilities for working through credit risk and machine learning concepts.
All tools are free, with reasonable usage limits to manage hosting costs.