Data Scaling for Operational Risk Modelling

24 Pages Posted: 2 Mar 2006

See all articles by H. S. Na

H. S. Na

Erasmus University Rotterdam (EUR) - Department of Computer Science

Lourenco Miranda

Erasmus University Rotterdam (EUR) - Department of Computer Science

J. van den Berg

Delft University of Technology, Faculty of Technology, Policy and Management, Section of ICT

M. Leipoldt

Erasmus University Rotterdam (EUR) - Department of Computer Science

Date Written: January 12, 2006

Abstract

In 2004, the Basel Committee on Banking Supervision defined Operational Risk (OR) as the risk of loss resulting from inadequate or failed internal processes, people and systems or from external events. After publication of the new capital accord containing this dfinition, statistical properties of OR losses have attracted considerable attention in the financial industry since financial institutions have to quantify their exposures towards OR events. One of the major topics related to loss data is the non-availability of a suficient amount of data within the Financial Institutions. This paper describes a way to circumvent the problem of data availability by proposing a scaling mechanism that enables an organization to put together data originating from several business units, each one having its specific characteristics like size and exposure towards operational risk. The same scaling mechanism can also be used to enable an institution to include external data originating from other institutions into their own exposure calculations. Using both internal data from different business units and publicly available data from other (anonymous) institutions, we show that there is a strong relationship between losses incurred in one business unit respectively institution, and a specific size driver, in this case gross revenue. We study an appropriate scaling power law as a mechanism that explains this relationship. Having properly scaled the data from different business units, we also show how the resulting aggregated data set can be used to calculate the Value-at-OR for each business unit and present the principles of calculating the value of the OR capital charge according the minimal capital requirements of the Basel committee.

Keywords: Minimal Capital Requirements, Operational Risk, Power Law Scaling, Loss Distribution, Value at Operational Risk

Suggested Citation

Na, H. S. and Miranda, Lourenco and van den Berg, J. and Leipoldt, M., Data Scaling for Operational Risk Modelling (January 12, 2006). ERIM Report Series Reference No. ERS-2005-092-LIS, Available at SSRN: https://ssrn.com/abstract=875782

H. S. Na

Erasmus University Rotterdam (EUR) - Department of Computer Science ( email )

Faculty of Economics
3000 DR Rotterdam
Netherlands

Lourenco Miranda

Erasmus University Rotterdam (EUR) - Department of Computer Science ( email )

Faculty of Economics
3000 DR Rotterdam
Netherlands

J. Van den Berg (Contact Author)

Delft University of Technology, Faculty of Technology, Policy and Management, Section of ICT ( email )

P.O. Box 5015
2600 GB Delft
Netherlands

HOME PAGE: http://tbm.tudelft.nl/index.php?id=30084&L=1

M. Leipoldt

Erasmus University Rotterdam (EUR) - Department of Computer Science ( email )

Faculty of Economics
3000 DR Rotterdam
Netherlands

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
668
Abstract Views
2,836
rank
46,886
PlumX Metrics