Expertise
Expertise
Quantitative methods, built for application. The details of our work are in our publications and notes.
Methods
Statistics and econometrics
Measuring, testing and monitoring risk models with clear statistical guarantees: calibration, extreme values, continuous monitoring.
Machine learning and artificial intelligence
More flexible models, provided they stay explainable, robust and validatable — from interpretable approaches to foundation models.
Stress testing and scenarios
Propagating scenarios through risk models, or starting from the breaking threshold to work back to the most plausible scenario.
Stress testing, in motion
Model validation and monitoring
Checking that a model stays well calibrated, and detecting drift as early as possible.
Applications
Credit risk
Developing, validating and monitoring credit models, and measuring how they behave under stress.
Climate risks
Bringing physical and transition risks into the assessment of financial risk.
Geopolitical and macro-financial risk
Understanding how shocks spread to portfolios, and which ones could put them in difficulty.