A Study On Measurement Of Operational Risk Of Commercial Banks

Rapid development in financial industry and significant changes in business environments have led a great deal of operational risks. Although financial institutions, especially banks, operates in such an environment full of increasing operational risks; its importance has not been fully recognized by bankers and regulators until a series of severe and fatal operational loss events in international financial industry since 1990s. At present, how to effectively measure operational risk is the focus and challenge for all practitioners and researchers.It is crucial to understand the causes, sources and consequences of operational risk, as a better understanding of operational risk helps us to measure it more accurately. In this paper, I review the research works of prior literatures and analyse the regulatory framework for operational risk proposed by the Basel committee of Banking Supervision. These important principles greatly impact the measurement of operational risk in practice.In current literatures, there are two methodologies in measuring operational risk, top-down and bottom-up. They are applied under different circumstances. A lot of bottom-up approaches have been developed, including internal measurement approaches, loss distribution approaches, scorecard approaches, extreme value models, Bayesian network and so on. This field is evolving rapidly.The scarce of operational loss data made us difficult to study the nature of operational risk, especially for those low frequency/high severity events, such as internal fraud. Internal fraud has brought many disastrous losses to the banking industry in China. For the first time, this dissertation evaluates the internal fraud risk and the corresponding economic capital for Chinese banks. The POT (peaks over threshold) model is employed to estimate the severity and the frequency of internal fraud loss, and the parameters required by POT are estimated by using a Bayesian MCMC method which overcomes the problems caused by insufficiency of loss data. My study reveals that the internal fraud risk, as one of the seven loss types of operational risk present in the new Basel Accord, is the major element of operational risk for banking industry in China.I also establish an operational risk evaluation model, using top-down approach, to measure comprehensive operational risk of listed banks in China. Using a quarterly panel data of equity returns over the period 2000-2006, the model returns a"residual"operational risk measure for (five) Chinese listed banks. The above method avoids the data problems with bottom-up approach, and it can be a supplement for advanced approach for estimating economic capital purposes. We find the ratio of operational risk to total equity for listed banks in Chinaâ‘ . The risk level, after transformed into the corresponding economic capital, is consistent with the outcome of Basic Indicator Approach in the new Basel Accord. This also suggests that the Basic Indicator Approach be a suitable operational risk capital estimation approach for banks in China.Furthermore, this paper shows how to establish and implement the scorecard system /approach as a mean to measure operational risk in a bank. The advantage of scorecard approach is that it measures operational risk by taking quantitative and qualitative data into account. Once an internal scorecard system is established, potential risk profiles can be captured more efficiently and effectively. This approach also provides a foreword-looking estimation for required operational risk capital.