Abstract
According to the last proposal by Basel Commitee, banks are allowed to use Advanced Measurement Approach(AMA) for operational risk. Since basic indicator and standard approach considers operational risk as a percentage of gross profit, these methedologies are not satisfactory as real lost or probability of lost do not taken into consideration. This article provides a theoretical and simulated empirical analysis of Loss Distribution Approach(LDA) in operational risk management as Advanced Measurement Approach(AMA). In this article, LDA is applied based on 20 distributions for univariate case and We adviced to mix internal and external data with LDA(Frachot and Roncalli,2002) based on bootsrapping method and introdıuced Switching Distribution approach in LDA. We also found that %99.9 confidence intervel is unapplicable in unexpected loss(UL) determination according to cumulative ascending plot for sofisticated distributions(same as BIS[2004] foresight) and adviced to choose confidence interval between %99.0 and %99.5.
| Translated title of the contribution | Advanced Measurement Approach With Loss Distribution In Operational Risk Management: ‘IX. National Finance Symposium’, Subject: Strategic Finance |
|---|---|
| Original language | Other |
| Title of host publication | Symposium was established in Nevsehir in Turkey, with the contributions of Finance Science Platform and Gazi University Faculty of Economics and Administrative Sciences |
| Publisher | Gazi University Publications |
| Pages | p. 206-227 |
| Publication status | Published - Sept 2005 |
Keywords
- Operasyonel risk, gelişmiş ölçüm yaklaşımı , zarar dağılımı, deşine dağılım, iç ve dış data birleştirilmesi, güven aralığı