Model Risk and Uncertainty in the Financial World
Model Risk and Uncertainty in the Financial World
Couldn't load pickup availability
Model Risk and Uncertainty in the Financial World is a comprehensive work that addresses the art and science of building and managing financial models with an appreciation of the risks and uncertainties that lie within. Soulellis and Ghose highlight how financial models can be laden with uncertainty, prone to failure, and capable of sparking crises when their risks are ignored. They distil foundational concepts in economics, statistics and machine learning into an intuitive accessible read for model builders and users, emphasising a useful distinction between risk and uncertainty, and the importance of managing them differently.
The authors’ narrative blends history, theory, and practice to show how model risk has affected markets — from Black Monday to the global financial crisis, from COVID-19 to Silicon Valley Bank. Drawing on their deep professional experience, they guide readers through key themes: the limits of probability, the psychology of risk, specification and operations failures, and how the past informs lessons for the future. They argue that uncertainty can never be eliminated — but it can be managed through the application of various measurement and estimation methods as well as a willingness to acknowledge the unknown.
This book equips bankers, regulators, quants, and policymakers with a sharper awareness of model risk and a toolkit for living with uncertainty. Model Risk and Uncertainty in the Financial World is essential reading for anyone who wants to understand where financial modelling has been — and where it’s going next.
Share
More information
About the Author
Devajyoti Ghose is co-founder and managing
director at Kah Capital Management, a mortgage investment firm. In 2018, he retired
from Freddie Mac, where he had been Treasurer, a member of the management
committee and Senior Vice President of models and portfolio analytics in its
capital markets division. He has a PhD in economics from the University of
California at San Diego, where he specialised in econometrics, with Nobel
Laureates Sir Clive Granger and Robert Engle as his advisors. He has taught economics
at the University of Arizona, Tucson and The University of New South Wales,
Sydney.
George Soulellis formerly served as the enterprise
model risk officer for Freddie Mac, with overall responsibility for model risk management
in the firm. Prior to that, he served as managing director, risk analytics for
Barclays Bank in the UK, overseeing risk model development and analytics. He
has also held leadership positions in the risk management/modelling/analytics
space at Citigroup, General Electric and JPMorgan Chase. He holds a BSc in
statistics from Concordia University and an MBA in artificial intelligence from
the University of Cumbria. He has also undertaken studies in statistics at
Columbia University and holds a MicroMasters in data science and specialisation
in mathematics for machine learning from the University of California at San
Diego and Imperial College London, respectively. His interests primarily lie in
model uncertainty measurement, model risk under conditions of extrapolation,
machine learning methods and modelling for capital requirements.
Table of contents
PART I FOUNDATIONS
1 The evolution of models
2 The foundations of risk and
uncertainty
3 Uncertainty: a taxonomy
4 Model risk and uncertainty: a
survey of the institutional landscape
5 Model specification risk and
uncertainty
6 Model operation risk and
uncertainty
PART II THE MODEL BUILDER’S
CANVAS
7 Data, models and their purpose
8 Artificial intelligence in
finance: a synthesis of human and machine
9 A deeper dive into machine
learning methods: their opportunities, limitations, risks and uncertainties
PART III MEASUREMENT,
APPLICATIONS AND USES
10 Measurement of risk and
estimation of uncertainty in prediction models
11 Using models under risk and
uncertainty
12 When models fail
Epilogue: models and the future
Index