[en] While the Basel II accord has now gone live in most parts of the world, many discrepancies still remain on advanced modeling techniques for operational risk among large international banks. The two major families of models include the loss distribution approaches (LDAs) that focus on observed past internal and external loss events and the scenario-based techniques that use subjective opinions from experts as the starting point to determine the regulatory capital charge to cover operational risk. A major methodological challenge is the combination of both techniques so as to fulfill Basel II requirements. In this chapter we discuss and investigate the use of various alternatives to model expert opinion in a sound statistical way so as to allow for subsequent integration with loss distributions fitted on internal and/or external data. A numerical example supports the analysis and shows that solutions exist to merge information arising from both sources.
