Operational Risk under AMA - some fundamental issues

It is highly pertinent for banks which are going for Advanced Measurement Approaches (AMA) for assessment and management of operational risk to clearly understand and implement a framework for capturing loss events across business lines. However, reaching on a decision as what will be the exact approach and methodology to capture these events is easier said then done. Since the fundamental essence of AMA is based on the flexibility for the bank to decide on the exact approach and methodology, it brings in plenty of subjectivity and also discretion of national supervisors for allowing the AMA to be eligible for Capital calculation purpose.

A bank needs to face and resolve many micro-level issues once the overall framework is decided. Many a banks face issues related to availability of internal loss data and standardisation of the same. Another chicken-and-egg issue is that for starting the exercise of internal loss data collection, banks face many fundamental issues which ultimately go towards deciding on the overall framework as well. Below are few of such fundamental issues related to internal data and AMA

• How to take into account loss events in supporting departments like Information Technology, Human Resources and Administrative Services. These departments' loss events cut-across core business lines of a bank and hence loss amount allocation is not definitive. Also, its not very easy to allocate losses from each core-business line-wise for IT or HR related loss events as underlying all actual losses may not be captured. What is the best approach for this issue needs proper deliberation.
• Can qualitative measure be used to quantify probability of loss event in absence of sufficient historical loss data? How acceptable this method can be to national supervisors, if loss data is gradually used over few years to replace the qualitative model?
• Can gamma factor (used in Internal Measurement Approaches) be totally ignored if industry-wide initiative is not there to decide on gamma? How to factor in the effect of gamma in such a situation.
• How to standardise Potential loss events which are Low-frequency, High-impact? how low is low and how high is high?
• How to define and differentiate between near miss and potential loss events?

Expert and experienced professionals' ideas in these regards will be highly valuable for banks which are in the starting phase of Operational Risk assessment and management initiative.