Modelling and Measurement Methods of Operational Risk in Banking

The last decades have seen a fast paced expansion in the range and depth of professional financial services, including electronic banking, cash management and liquidity control, as well as the introduction of a wide range of complex derivative products. This development, which regained new strength on a global scale as new economic poles emerge in Eastern Europe, India and China, has led to an increased demand for operational risk management within financial institutions. Recurring financial crises that hit the banking industry in the early years of the twenty-first century, such as the burst of the new economy bubble or the more recent subprime mortgage crisis, have shown the need and rene-wed the demand for integrative risk management concepts, which inclu-de operational risk management within the global financial services industry.

Drawing on years of theoretical and practical experience in the area of banking risk management, the author bridges the gap between theory and implementation to introduce an innovative approach to identifying, measuring, analysing, modelling and managing operational risk within financial institutions.

Leseprobe:

Chapter Four (p. 115-116)
4 Interrelations between the Operational Risk Measurement and Consideration of Risk in Operational Decisions

4.1 Data Base Collection

After describing measurement methods and the role of internal structures of operational activity processes, the subject of operational decisions quantification is the next step for a further consideration. There are interrelations between measurement and quantification of operational risk, which have to be researched. The reason for it is that a judgement of the operational risk is not possible without connecting measurement and quantification.

In subchapter 2.3, models and measurements methods of operational risk are discussed and, referring to table 10, a collection of top-down and bottom-up risk model types and their conceptual features are presented. As a result, among other models, the data based loss potential model, a statistical model, was named and preferred. The advantage of a loss database is that the loss event database captures and accumulates banks individual loss events with respect to its businesses and risk types. This information is detailed, in structured manner and in logical sequences. Therefore, the data-collection is an especially detailed method because there are still no other tools for a data-capture and measurement. Based on missing historical data about operational risk events, as noticed in former chapters, the loss event database is the first step in measuring and quantifying operational risk. Therefore, the loss event database is the most important model in starting operational risk decisions in commercial banking. An essential foundation for any rigorous operational risk decision process is comprehensive, reasonable, verifiable and validated data covering the historical operational risk loss experience of the bank. The discipline of collecting loss data is not only needed to understand the dimensions of operational risk the bank faces. It can also be used to motivate staff to consider and more actively control key elements of operational risk.

The discipline of bank-wide data collection promotes a dialogue within the bank about determining the major operational risk exposures and drivers and reinforces more qualitative efforts to manage operational risk within each of the business lines. To the subject of business lines see subchapter 3.1, where identification of the main areas of operational activity are stated. Thus, it is a sound practice for financial institutions to have a framework for collecting data on their actual operational risk loss experience within material business lines. Events are described and stored in various locations for analysis. Operational risk loss data consists primarily of routine, generally high-frequency, low-impact events, as well as low-frequency, high-impact events. In future, banks have to implement reporting systems to track both types of loss events, including reference to external data on large loss events for an operational risk quantification. Average losses or expected losses in a bank are generally driven by the highfrequency and low-impact events. Currently, there is no active recording of these losses, because of a missing active decision quantification of operational risk in commercial banks.

These expected losses generally should be budgeted with a high degree of confidence and routinely flow through annual prediction of the income statement, also called prognosis calculation. In contrast, unexpected losses – which tend to reflect the impact of low-frequency, high-impact events – occur infrequently and are sometimes sufficiently large as to result in a periodic loss and reduction in the regulatory tier one capital.293 For a measurement, the data should be collected and registered in the database. Based on this collection, it is a quantitative method of operational risk events measurement. Creating a loss database it is important that the chosen scheme fit for all operational risk elements, described in subchapter 2.2 as contents of operational risks elements and features. Referring to this subchapter as a repetition, those are external and internal, such as people, systems and technology, processes, management failures including strategic management risk and reputational risk.