Rating migration explorer
About this app
Explore a rating migration matrix, transition probabilities between rating grades and migration over multiple years.
Transition matrices
A rating migration (transition) matrix M holds, in row i and column j, the probability of moving from rating class i to class j within one year. Each row sums to 100 %. The diagonal shows how likely a rating is to stay put; entries to the right are downgrades and entries to the left are upgrades. Default is an absorbing state. Under the Markov assumption, the n-year transition matrix is the matrix power Mⁿ, and its default column gives cumulative default probabilities.
What the app shows
The app uses four rating classes, A to D, plus default. The “Heatmap” view colours the one-year matrix; clicking a row shows the distribution for that starting rating. The “Paths” view branches into every class each year: multiply the probabilities along a path, add up all paths that end in default, and you get the corresponding entry of Mⁿ. The metrics panel lists cumulative PD, survival probability and marginal and cumulative PD per year.
Things to try
Pick a starting rating and a horizon of 1 to 5 years and press “Play”: the routes into default light up one by one and add up to the cumulative PD. Switch between the “Standard”, “Stable” and “Stress” presets and compare how quickly default probability grows over several years. Notice that even a good rating can reach default by passing through intermediate classes.
Frequently asked questions
- What is a rating migration matrix?
- A migration matrix gives the probability that a borrower moves from its current rating class to every other class, or to default, within a period, usually one year. Each row sums to 100 %. Rating agencies and banks estimate these matrices from historical rating histories.
- How do you get a multi-year default probability from a transition matrix?
- If transitions are assumed independent over time (a Markov chain), you multiply the one-year matrix by itself n times. The default-column entry of Mⁿ is the cumulative default probability after n years. It equals the sum of the probabilities of all rating paths that end in default by then.
- Why is default called an absorbing state?
- An absorbing state is never left once entered. In a migration matrix, the default row has a 1 on the diagonal. As a result, cumulative default probability can only increase with the horizon.
- Which matrices can I compare in the Migration Explorer?
- The app offers three presets: “Standard”, “Stable” and “Stress”. For each, you choose the starting rating and horizon and switch between heatmap and path grid; the values are example matrices with classes A to D.
Subject: Credit risk
More from Credit risk
- Adverse selection
- Default correlation surface in 3D
- CreditPortfolioView: macro factors
- CreditRisk+: Poisson modelling
- Exposure profile
- Integrated risk measurement
- PD master scale
- Risk-bearing capacity cockpit
A VisuApp by heyprof: interactive, free in your browser, no sign-up. What is a VisuApp? · All apps