The Heston model and a well-calibrated local volatility model produce identical prices for all European options on a given date — they are both perfectly calibrated to the vanilla surface. Yet they can differ by 5–15% on exotic products (cliquets, Napoleons, barriers). This is not numerical error: it is model risk, the price of choosing between two equally plausible models.
This note develops the formal framework for measuring and managing model risk in an exotic derivatives book, identifies the forward smile as the key discriminant between models, and describes the practical risk management tools that exotic desks use to control model risk exposure.
Learning objectives. By the end of this note you should be able to:
- define model risk for a derivative and distinguish it from market risk and parameter risk;
- compute the model risk P&L for a cliquet using two calibrated models;
- explain why the forward smile is the discriminant between local vol and stochastic vol for exotic products;
- describe the standard industry practice for model risk reserves and model validation.
In this lesson
- 4.1 What is model risk?
- 4.2 The forward smile as the discriminant
- 4.3 The model risk spectrum for structured products
- 4.4 Industry practice: model risk reserves
Before you start
The full lesson, with worked examples and exercise solutions, is for members.
Read the full lessonNew to SelfMadeQuant.com? See membership options or browse the Advanced Derivatives and Exotic Options syllabus.