How do credit transition matrices work and how are they used in portfolio credit risk?

A credit transition matrix shows the probability that an obligor rated in a given category at the start of a period will migrate to any other rating category (or default) by the end of the period.

Reading the Matrix

Each row represents the starting rating; each column represents the ending rating. The row must sum to 100%.

How It's Used in Practice

Suppose Meridian Capital holds $50M in BBB-rated bonds from Hargrove Industries. Using the transition matrix above:

• 84.2% probability the bond stays BBB — no mark-to-market change
• 5.9% probability of downgrade to BB — the spread widens, causing a mark-to-market loss
• 0.7% probability of default — potential recovery-rate-adjusted loss

Multi-Period Matrices

To get a 2-year transition matrix, you multiply the 1-year matrix by itself: T₂ = T₁ × T₁. This assumes the Markov property — future transitions depend only on the current state, not the path taken.

Key Limitations:

1. Non-stationarity — real-world migration rates change with the economic cycle
2. Rating momentum — recently downgraded issuers have higher migration risk than stable issuers at the same rating
3. Through-the-cycle vs. point-in-time — agency ratings are through-the-cycle and may lag

For the FRM exam, be ready to calculate portfolio credit VaR by combining transition probabilities with spread changes at each rating level. Practice with our FRM question bank for worked examples.