Skip to content

FRM Part II · FRM Exam Part II · Risk Identification

An analyst reviews an internal loss event database and finds that it contains only events above a USD 50,000 collection threshold. Which is the most direct consequence if the analyst fits a severity distribution to these data without adjustment?

Fitting without adjustment can bias severity estimates and understate event frequency, because losses below USD 50,000 are missing and the data are left-truncated. The model must be conditioned on the threshold to correct for this.

  1. ASeverity parameters may be biased because the data are left-truncated, and frequency of total events will be understatedCorrect
  2. BThe data will overstate the tail because large losses are double counted
  3. CFrequency will be overstated because small events are recorded multiple times
  4. DNo bias arises because small losses do not contribute to capital

Explanation

A collection threshold truncates the data from the left. Fitting a distribution as if it were complete biases severity estimates, and the count of events below the threshold is missing, so frequency is understated unless adjusted for the truncation. The other options describe effects that do not arise from a threshold.

Did you get it right without looking?

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

Practise Risk IdentificationFree diagnostic mock

More Risk Identification questions