Quantifying operational loss data

Dear Sir/Madam,

Organisations that are in the process of building infrastructure to satisfy the Basel Accord may be driven to spend resources most likely to mitigate the compliance and not the root causes of operational risk (op risk) that result in operational losses.

For example, op risk loss data collection includes the gross losses that are posted to the general ledger in a specific time frame; there can be multitude of near-miss losses called potential losses that may be filtered from entering the loss history database due to benchmarking and reporting reasons. These near misses can be a catalyst for future operational losses. A potential loss can also be recognised as a key risk indicator for a particular business line if the frequency of its occurrence is on the rise in a given time interval.

Cash loss control and specialised units generally include potential losses in their database for a variety of reasons. These losses associated with potential risk events, if included in the loss history database, can provide Op risk analysts with data to forecast and assess major risks associated with business lines.

Another key issue is that the op risk loss data for benchmarking (available to financial institutions from outside sources) is not specific in size and income exposure. In other words, there should be a way where a bank can benchmark its data to the data of its peer group which is similar in size and exposure. At the same time regional benchmarking can be beneficial when understanding the overall risk exposure for a financial institution.

Most companies have their own standard op risk categories, which are derived from their corporate risk policy, and applied across the board to diverse business lines. The ‘one shoe fits all’ strategy is likely to fail in the absence of a clear-cut risk assessment methodology. Financial companies that are similar in scope in their risk exposure have from 40 to 2,200 risk categories. There should be a manageable number of risk categories established only after consultation with relevant business units. These risk categories can be linked to a business line according to its risk profile. Linking or quantifying these risk categories to specific business lines and op risk loss events data, can help op risk managers make sound decisions.

Haji Amin

Manager operational risk
Bank of Nova Scotia
ORX site Administrator