Extreme Events : Robust Portfolio Construction in the Presence of Fat Tails

by Kemp, Malcolm

• ISBN13: 9780470750131
• ISBN10: 0470750138
• Format: Hardcover
• Copyright: 2010-12-28
• Publisher: Wiley

Summary

With slight exaggeration, a case can be made that modern finance has been built, in practice, if not in theory, on implicit tolerance and widespread ignorance of extreme events.Jean_Pierre Landau, Deputy Governor, Banque du France Markets are fat-tailed; extreme outcomes occur more often than many might hope, or indeed the statistics or normal distributions might indicate. In this book, the author provides readers with the latest tools and techniques on how best to adapt portfolio construction techniques to cope with extreme events. Beginning with an overview of portfolio construction and market drivers, the book will analyze fat tails, what they are, their behavior, how they can differ and what their underlying causes are. The book will then move on to look at portfolio construction techniques which take into account fat tailed behavior, and how to stress test your portfolio against extreme events. Finally, the book will analyze really extreme events in the context of portfolio choice and problems. The book will offer readers: Ways of understanding and analyzing sources of extreme events Tools for analyzing the key drivers of risk and return, their potential magnitude and how they might interact Methodologies for achieving efficient portfolio construction and risk budgeting Approaches for catering for the time-varying nature of the world in which we live Back-stop approaches for coping with really extreme events Illustrations and real life examples of extreme events across asset classes This will be an indispensible guide for portfolio and risk managers who will need to better protect their portfolios against extreme events which, within the financial markets, occur more frequently than we might expect. Table of contents IntroductionTraditional Portfolio Construction TechniquesFactors that Drive the MarketsFat Tails what causes them?Heteroscedasticity: the time-varying nature of our worldUnivariate fat-tailed behaviorMultivariate fat-tailed behaviorPortfolio construction in the presence of fat-tailsRobust portfolio constructionStress testingReally extreme events

Author Biography

Malcolm Kemp (London, UK) is Founder and Managing director of Nematrian Ltd, a consulting firm delivering services to the quantitative finance and actuarial communities. Previously, he was Director and Head of the Quantitative Research Team at Threadneedle Asset Management, responsible for the derivative desk and its portfolio risk measurement and management activities. He is a leading expert on derivatives, performance measurement, risk measurement, liability driven investment and other quantitative investment techniques. Prior to this, Malcolm was a partner at Bacon & Woodrow in their investment consultancy practice. He holds a first class degree in Mathematics from Cambridge University and is also a Fellow of the Institute of Actuaries. He is a regular on the conference circuit, including Risk Europe and GARP events where he speaks on a range of portfolio management and derivatives topics.

Table of Contents

Preface

Acknowledgements

Abbreviations

Notation

Introduction

Extreme events

The portfolio construction problem

Coping with really extreme events

Risk budgeting

Elements designed to maximise benefit to readers

Book structure

Fat Tails - In Single (i.e. Univariate) Return Series

Introduction

A fat tail relative to what?

Empirical examples of fat-tailed behaviour in return series

Characterising fat-tailed distributions by their moments

What causes fat tails?

Lack of diversification

A time-varying world

Stable distributions

Extreme value theory (EVT)

Parsimony

Combining different possible source mechanisms

The practitioner perspective

Implementation challenges

Fat Tails - In Joint (i.e. Multivariate) Return Series

Introduction

Visualisation of fat tails in multiple return series

Copulas and marginals - Sklar's theorem

Example analytical copulas

Empirical estimation of fat tails in joint return series

Causal dependency models

The practitioner perspective

Implementation challenges

Identifying Factors That Significantly Influence Markets

Introduction

Portfolio risk models

Signal extraction and principal components analysis

Independent Components Analysis

Blending together PCA and ICA

The potential importance of selection effects

Market dynamics

Distributional mixtures

The practitioner perspective

Implementation challenges

Traditional Portfolio Construction Techniques

Introduction

Quantitative versus qualitative approaches?

Risk-return optimisation

More general features of mean-variance optimisation

Manager Selection

Dynamic optimisation

Portfolio construction in the presence of transaction costs

Risk budgeting

Backtesting portfolio construction techniques

Reverse optimisation and implied view analysis

Portfolio optimisation with options

The practitioner perspective

Implementation challenges

Robust Mean-Variance Portfolio Construction

Introduction

Sensitivity to the input assumptions

Certainty equivalence, credibility weighting and Bayesian statistics

Traditional robust portfolio construction approaches

Shrinkage

Bayesian approaches applied to position sizes

The 'universality' of Bayesian approaches

Market consistent portfolio construction

Re-sampled mean-variance portfolio optimisation

The practitioner perspective

Implementation challenges

Regime Switching and Time-Varying Risk and Return Parameters

Introduction

Regime switching

Investor utilities

Optimal portfolio allocations for regime switching models

Links with derivative pricing theory

Transaction costs

Incorporating more complex autoregressive behaviour

Incorporating more intrinsically fat-tailed behaviour

More heuristic ways of handling fat tails

The practitioner perspective

Implementation challenges

Stress Testing

Introduction

Limitations of current stress testing methodologies

Traditional stress testing approaches

Reverse stress testing

Taking due account of stress tests in portfolio construction

Designing stress tests statistically

The practitioner perspective

Implementation challenges

Really Extreme Events

Introduction

Thinking outside the box

Portfolio purpose

Uncertainty as a fact of life

Market implied data

The importance of good governance and operational management

The practitioner perspective

Implementation challenges

The Final Word

Conclusions

Portfolio construction principles in the presence of fat tails

Appendix: Exercises

Introduction

Fat Tails - In Single (i.e. Univariate) Return Series

Fat Tails - In Joint (i.e. Multivariate) Return Series

Identifying Factors That Significantly Influence Markets

Traditional Portfolio Construction Techniques

Robust Mean-Variance Portfolio Construction

Regime Switching and Time-Varying Risk and Return Parameters

Stress Testing

Really Extreme Events

Table of Contents provided by Publisher. All Rights Reserved.