Operational Risk Assessment: Qualitative Methodological Approaches Beyond Self-Assessment

Operational risk, defined as the risk of losses resulting from inadequate or failed internal processes, people and systems or external events, is a significant risk for insurers and banks, and carries a considerable capital charge under the Solvency II and Basel III regimes. Particularly for insurers, operational risk is perhaps the most challenging risk to quantify, due to lack and inconsistency of historical data (particularly extreme loss data), lack of adjustment due to changes in underlying drivers (management/culture/processes), inapplicability of external loss data, and strong overlap with other risk categories. The shortfalls of current quantification methods underline the fact that operational risk is by nature qualitative and therefore the qualitative part of its assessment must be robust.
So far, the qualitative aspect of operational risk modelling has almost exclusively focused on key risk indicators and self-assessment questionnaires based on which distributions are constructed. Yet at the same time, self-assessment – broadly defined as the involvement of practitioners in judging whether or not self-identified standards have been met – has been consistently shown to be inadequate and inaccurate, across a range of examined fields (medicine, counselling, law, engineering, behavioural science, psychology, education, etc). In complex systems such as insurance organizations, much operational risk comes from explicit and implicit human error (due to miscalculations, a range of social desirability, confirmation and other cognitive biases), which are difficult to estimate using questionnaires. In this paper, we outline the methodological inadequacy of relying on self-assessment questionnaires, and offer suggestions based on qualitative approaches from the social sciences, including qualitative and mixed methods approaches such as focus groups, vignettes, field experiments, and critical incidence narrative in-depth interviewing. Better methodologies would allow insurance companies to both better identify and mitigate the sources and factors of operational risk as well as better estimate the frequency and potential size of these errors.
 
Find the Q&A here: Q&A on 'Operational Risk'