Skip to content

FRM Part II · FRM Exam Part II · Risk Measurement and Assessment

A bank's operational risk team builds a loss distribution approach (LDA) model using only internal loss data collected over the past five years. The data contain very few events above USD 5 million, yet the bank has business lines exposed to rare, severe events. Which limitation of the data most directly threatens the reliability of the capital estimate at a high percentile?

The main problem is that few extreme losses exist in the internal data, so the severity tail is estimated from very little information. Since high-percentile capital depends mostly on the tail, parameter uncertainty is large and the capital estimate is unreliable unless supplemented with external data or scenarios.

  1. AInternal data are too heavily weighted toward high-severity events
  2. BSparse tail observations make severity tail parameters highly uncertainCorrect
  3. CInternal data cannot capture loss frequency
  4. DInternal data always overstate risk because they are backward-looking

Explanation

At high percentiles (e.g., 99.9%) the capital figure is driven by the tail of the severity distribution. With few large losses, the tail parameters are estimated from very little information and are unstable. Option A is reversed, since internal data usually under-represent severe events. Internal data do capture frequency well.

Did you get it right without looking?

One question tells you little. A timed set on Risk Measurement and Assessment shows your real accuracy, how long you take and where you lose marks.

Practise Risk Measurement and AssessmentFree diagnostic mock

More Risk Measurement and Assessment questions