Models

Models

Notes on building, testing, and understanding models used in financial applications.

Model risk

Model risk, from scratch

How models fail, why validation matters, and why no model is ever neutral.

Assumptions

Assumptions are the model

Every model contains beliefs about data, behavior, time, and uncertainty.

Validation

What validation actually checks

Validation is not proving a model is true. It is understanding where it works, where it fails, and why.

Testing

Overfitting and false confidence

How a model can look accurate on past data and fail in real use.

Governance

Model limits

A useful model should explain what it cannot know.