Model Risk Management: Risk Bounds under Uncertainty [Hardcover]

Develop the tools to quantify model risk, to study its effects in finance, insurance, and engineering, and to reduce it.The first systematic treatment of model risk, this book provides the tools needed to quantify and assess the impact of model uncertainty. It will be essential for all those working in portfolio theory and the theory of financial and engineering risk, for practitioners in these areas, and for graduate courses on risk bounds and model uncertainty.The first systematic treatment of model risk, this book provides the tools needed to quantify and assess the impact of model uncertainty. It will be essential for all those working in portfolio theory and the theory of financial and engineering risk, for practitioners in these areas, and for graduate courses on risk bounds and model uncertainty.This book provides the first systematic treatment of model risk, outlining the tools needed to quantify model uncertainty, to study its effects, and, in particular, to determine the best upper and lower risk bounds for various risk aggregation functionals of interest. Drawing on both numerical and analytical examples, this is a thorough reference work for actuaries, risk managers, and regulators. Supervisory authorities can use the methods discussed to challenge the models used by banks and insurers, and banks and insurers can use them to prioritize the activities on model development, identifying which ones require more attention than others. In sum, it is essential reading for all those working in portfolio theory and the theory of financial and engineering risk, as well as for practitioners in these areas. It can also be used as a textbook for graduate courses on risk bounds and model uncertainty.Introduction; Part I. Risk Bounds for Portfolios Based on Marginal Information: 1. Risk bounds with known marginal distributions; 2. Rearrangement algorithm; 3. Dual bounds; 4. Asymptotic equivalence results; Part II. Additional Dependence Constraints: 5. Improved standardlƒ#