FRM Part II · FRM Exam Part II · Credit Scoring and Rating
A bank's risk team studies a one-year rating transition matrix built from agency data. Which statement about the default row (D) of a standard one-year transition matrix is correct?
Default is modeled as an absorbing state, so the default row shows 100% probability of remaining in default and 0% of moving to any other rating. This lets multi-year cumulative default probabilities be computed by repeated matrix multiplication, and every row still sums to 100%.
- ADefault is an absorbing state: the probability of staying in default is 100% and the probability of moving to any other rating is 0%Correct
- BDefault is a transient state: defaulted issuers migrate back to investment grade with a small positive probability
- CThe default row contains the average default probabilities of all rating classes
- DThe default row sums to less than 100% because recoveries are excluded
Explanation
In a standard transition matrix the default state is treated as absorbing, so its row has a 1 on the diagonal and zeros elsewhere. Treating it as transient would contradict the way multi-period default probabilities are derived. Every row, including default, sums to 100%.
Did you get it right without looking?
One question tells you little. A timed set on Credit Scoring and Rating shows your real accuracy, how long you take and where you lose marks.
More Credit Scoring and Rating questions
- A risk manager monitors a deployed scorecard and finds a population stability index (PSI) of 0.32 between the development and current applic…
- An analyst uses a one-year transition matrix from a rating agency. A bond is rated BBB. The matrix gives one-year probabilities from BBB: to…
- A bank validates a new application scorecard on a holdout sample. At a given cutoff, the cumulative share of defaulters scoring below the cu…
- A portfolio analyst uses a one-year rating transition matrix from an agency. The one-year probabilities for a BBB-rated bond are: upgrade to…
- A retail lender builds a logistic regression scorecard to predict 12-month default on personal loans. Each predictor is first grouped into b…
- A lender's logistic regression scorecard has a log-odds (good:bad) score. Points are scaled so that 600 points correspond to odds of 50:1 an…