Measuring and managing operational risk in the insurance and banking sectors.

Summary Our interest in this thesis is to combine the different techniques for measuring operational risk in the financial sectors, and we are particularly interested in the consequences of estimation risk in models, which is a particular operational risk. We will present the associated mathematical and actuarial concepts as well as a numerical application with respect to the advanced measurement approach as Loss Distribution to calculate the capital requirement. In addition, we focus on estimation risk illustrated with expert opinion scenario analysis in conjunction with internal loss data to assess our exposure to severity events. We conclude this first part by defining an OLS-based scaling technique that allows us to normalize our external data to a local Lebanese bank.In the second part, we give importance on the measurement of the error induced on the SCR by the estimation error of the parameters, we propose an alternative method to estimate a yield curve and we conclude by drawing attention on the reflections around the assumptions of calculation and what we agree to qualify as an assumption "consistent with market values" would be much more relevant and effective than the complexification of the model, source of additional instability, thus highlighting the estimation risk which is linked to the operational risk and must be given much more attention in our working models.