Modeling expert opinions on operational risk in pension funds
In this thesis we model expert opinions on operational risk in pension funds. The approach used in this research allows the usage of a very simple questionnaire which, in theory, reduces psychological biases in eliciting expert opinions. However, we can not test its success in doing so. The judgements of 11 experts are aggregated to find an estimate of a 97.5% VaR. The resulting VaR can be used to find a formula for operational risk in the solvability assessment of the FTK. We find that the point estimate of this VaR is inaccurate and its reliability can not be tested. Nevertheless, it is the first estimate ever made for operational risk in the FTK and lack of real world data makes it difficult to find a better estimate. A case study shows that the effect of operational risk on the required capital (VEV) of a pension fund is small.