Inaugural-Dissertation zur Erlangung des Grades Doctor oeconomiae publicae an der Ludwig-Maximilians-Universität München (D19).
Operational risk and its quantification have become an area of growing concern in financial institutions in the past years. The Basel Committee on Banking Supervision requires internationally active banks to set a capital cushion against operational risk losses. Therefore finding the appropriate ways of operational risk modelling has become an important factor that influences on banks' competitiveness. The peculiarities of operational risk processes and events limit the usage of conventional statistical and actuarial methods. It is proposed in this dissertation to overcome these limitations by means of structural models, which would allow to describe operational risk taxonomy and to combine data from different sources. In particular, it investigates the applicability of graphical models to structural operational risk modelling. The application of structural methods is illustrated on the example on information systems and infrastructure risks.