Technical paper/Loss data
Determination of the fraction of losses and their probabilities by type of risk and business line from aggregate loss data
This paper proposes a novel means to derive the individual loss severities and the frequency of these losses per business line and risk type.
A text analysis of operational risk loss descriptions
The authors put forward a workflow for using text analysis to identify underlying risks in operational risk event descriptions.
Central counterparty capital and nondefault losses
This paper analyses the components of central counterparty (CCP) capital requirements and makes several observations on the potential for loss absorption.
An investigation of cyber loss data and its links to operational risk
This paper investigates cyber loss data and focuses on quantifying the direct financial and compensatory losses emanating from cyber risks.
A simulation comparison of aggregation periods for estimating correlations within operational loss data
This paper investigates the differences in the values of correlations based on different aggregation periods of time series loss data.
A maximum entropy approach to the loss data aggregation problem
This paper examines and compares alternative ways of solving the problem of determining the density of aggregate losses.
Constructing an operational event database
Michael Haubenstock of US bank Capital One outlines a framework for an event database, formulated with current US regulatory guidance on the subject in mind. The text is an abstract from The Basel Handbook, which has just been published by Risk Books.
Exceptional operational risks: Three myths debunked
Are there common features among exceptional operational risks beyond their defining characteristics of rarity and severe consequences?
How to avoid overestimating capital charges for op risk
Pooling internal and external data is a central issue to estimating capital charges for operational risk. Here, Nicolas Baud, Antoine Frachot and Thierry Roncalli of Crédit Lyonnais discuss the methodology they have developed.
Op risk modelling for extremes
Part 2: Statistical methods In this second of two articles, Rodney Coleman, of Imperial College London, continues his demonstration of the uncertainty in measuring operational risk from small samples of loss data.
Using Bayesian networks to predict op risk
By combining qualitative and quantitative data, Bayesian networks offer the perfect solution to the compelling need for an integrated approach to operational risk management, say Martin Neil and Ed Tranham.