Operational Risk economic capital model

  • Featured
  • Risk Management

Prescio developed an Operational Risk economic capital model for a major U.S bank following a Bayesian methodology to combine internal, external, and structured scenario data according to Basel II requirements. In this approach, posterior distributions were generated by combining the prior distributions and likelihood functions, and these distributions were sampled at each Monte Carlo step to obtain the frequency and severity parameters used in the calculation.

A benchmark model was developed following the Buhlmann-Straub approach which applies credibility theory to combine multiple data types. The distributions obtained in the credibility calculation were sampled by Monte Carlo to calculate economic capital.

Prescio also developed time series models to forecast Operational Risk losses in different economic scenarios for CCAR stress testing exercises.