Glossary term
Loss Event Database
Subject: Risk Management, Operational and Model Risk
Also known as: Internal Loss Data, Operational Loss Database
A loss event database is a central record of operational loss events, with details such as date, amount, cause, business line and recovery. Firms use it to understand patterns and to feed capital and risk models.
Explanation
Each entry captures what happened, where, why and what it cost, including amounts later recovered. Banks under the Basel rules record losses above a chosen threshold and keep several years of history, and supervisors expect the data to be complete and reconciled to the accounts. Entries are classified using a standard list, commonly the Basel event types. The data supports root cause analysis, trend reporting to management and the design of key risk indicators. Under the older advanced measurement approach it also fed statistical models of operational loss. Firms often join industry consortia to pool anonymised losses, since a single firm has too few large events to see the tail. Its weak point is under-reporting: staff may stay quiet about their own slips, so small, embarrassing events are the ones most likely to be missing.
Example
Every operational loss above a set threshold goes into a bank's database. One quarter, thirty-one entries show duplicate customer payments made during manual overrides, costing a total of 400,000. The risk team groups them by cause, finds that most arose in one branch cluster, and recommends a system block on repeat payments. Following the fix, the entries drop sharply.
Not to be confused with
- Risk register
- A risk register lists risks that might happen. A loss event database records losses that actually occurred.