| Abstract [eng] |
This thesis examines risk attribution methods (RAM) and their application to business loan portfolio analysis. The analysis is based on coherent risk measure principles and classical risk attribution methods like the Euler risk allocation principle and game-theory based attribution methods. That makes it possible to decompose total portfolio risk into contributions of individual components. Risk attribution is illustrated using a numerical example, demonstrating how portfolio tail risk can be allocated to individual portfolio components. The methodology is then applied to real business loan portfolio data. The obtained results make it possible to assess risk concentration and identify the segments that contribute most to extreme losses in a business loan portfolio. |
