Statistical Analysis of Operational Risk Data (SpringerBriefs in Statistics)

✍ Scribed by Giovanni De Luca, Danilo Carità, Francesco Martinelli

• Publisher: Springer
• Year: 2020
• Tongue: English
• Leaves: 92
• Category: Library

✦ Synopsis

This concise book for practitioners presents the statistical analysis of operational risk, which is considered the most relevant source of bank risk, after market and credit risk. The book shows that a careful statistical analysis can improve the results of the popular loss distribution approach. The authors identify the risk classes by applying a pooling rule based on statistical tests of goodness-of-fit, use the theory of the mixture of distributions to analyze the loss severities, and apply copula functions for risk class aggregation. Lastly, they assess operational risk data in order to estimate the so-called capital-at-risk that represents the minimum capital requirement that a bank has to hold. The book is primarily intended for quantitative analysts and risk managers, but also appeals to graduate students and researchers interested in bank risks.

✦ Table of Contents

Contents
List of Figures
List of Tables
1 The Operational Risk
1.1 Introduction
1.2 Models for Operational Risk
1.2.1 Basic Indicator Approach
1.2.2 Standardized Approach
1.2.3 Advanced Measurement Approach
1.3 Loss Distribution Approach
1.4 DIPO Consortium
References
2 Identification of the Risk Classes
2.1 Introduction
2.2 Distributional Tests
2.3 Application to DIPO Data
References
3 Severity Analysis
3.1 Introduction
3.2 Mixture of Three-Parameter Log-Normal Distributions
3.3 Extreme Value Theory
3.4 Application to DIPO Data
3.4.1 Mixture of k Log-Normal Distributions
3.4.2 Log-Normal–GPD Distribution
3.4.3 Comparison
References
4 Frequency Analysis
4.1 Introduction
4.2 Mixture of Poisson Distributions
4.2.1 The Poisson Distribution
4.2.2 Finite Poisson Mixture
4.3 Mixture of Negative Binomial Distributions
4.3.1 The Negative Binomial Distribution
4.3.2 Relationship with Poisson Distribution
4.3.3 Maximum Likelihood Estimation
4.3.4 Finite Negative Binomial Mixture
4.4 Application to DIPO Data
References
5 Convolution and Risk Class Aggregation
5.1 Introduction
5.2 Overall Loss Distribution
5.3 Risk Class Aggregation and Copula Functions
5.3.1 Tail Dependence
5.3.2 Elliptical Copulae
5.3.3 Archimedean Copulae
5.4 Value-at-Risk Estimates Considering t-Copula
References
6 Conclusions