Scribd
Upload
66 views14 pages

Introduction to Credit Risk Plus Model

The document provides an introduction to Credit Risk Plus (CR+), a powerful but complex credit risk modeling technique. CR+ models the probability distribution of total credit losses from a loan portfolio by combining models of the probability of default for each loan, the number of expected defaults across the portfolio, and the loss distribution for each defaulting loan. While the document only outlines a simple version of CR+ assuming independence between defaults, in practice the full model accounts for more realistic factors like correlation between defaults and variable default rates.

Uploaded by

Alice
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
66 views14 pages

Introduction to Credit Risk Plus Model

The document provides an introduction to Credit Risk Plus (CR+), a powerful but complex credit risk modeling technique. CR+ models the probability distribution of total credit losses from a loan portfolio by combining models of the probability of default for each loan, the number of expected defaults across the portfolio, and the loss distribution for each defaulting loan. While the document only outlines a simple version of CR+ assuming independence between defaults, in practice the full model accounts for more realistic factors like correlation between defaults and variable default rates.

Uploaded by

Alice
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

TW3421x - An Introduction to Credit Risk Management

Default Probabilities
Credit Risk Plus
!

Dr. Pasquale Cirillo

Week 6
Lesson 3
Credit Risk Plus (CR+)

✤ Introduced in 1997 by Credit Suisse 



Financial Products.!

✤ It is based on well-known tools 



of actuarial mathematics.!

✤ It is a powerful but complex model. 



Here we just sketch the very basic idea.

The Credit Suisse logo is the property of the Credit Suisse Group. Its use here is only for didactic purposes.
Basic Idea

✤ Suppose that a financial institution has n loans of a given type.!

✤ For simplicity we assume these loans to be homogeneous in terms of risk, so


that we can say that the 1-year PD of each loan is p.!

✤ p can be obtained from external or internal credit ratings, for example.


Number of Defaults

✤ Let µ be the expected number of defaults for the whole portfolio of loans.!

✤ Then we have that

µ = np
Number of Defaults

✤ Let µ be the expected number of defaults for the whole portfolio of loans.!

✤ Then we have that

µ = np
Number of Defaults

✤ If we assume defaults to be independent, the probability of observing m


defaults over the total of n loans will be like the probability of tossing a
(possibly biased) coin n times and observing m heads, when the probability of
getting a head is p.
Number of Defaults

✤ If you are familiar with basic probability, you know that such a probability is 





n!

 pm (1 p)n m
m!(n m)!


Number of Defaults

✤ If you are familiar with basic probability, you know that such a probability is 





n!

 pm (1 p)n m
m!(n m)!



Binomial distribution
Poisson Approximation

✤ If we assume p to be small and n large, the Binomial distribution is well


approximated by a Poisson distribution.!

✤ The probability of observing m defaults thus becomes






 e µ m
µ

 m!
Total distribution of losses

✤ The previous information about the probability of observing a certain number


of defaults can be combined with the probability distribution for the losses
experienced when a certain type of counterparty defaults.!

✤ This leads us to the computation of a probability distribution for the total


losses from defaults.!

✤ On that distribution we can compute quantities such as VaR and ES.


Losses from a counterparty

✤ The probability distribution for the


losses from a counterparty, when it
defaults, can be determined from
historical data.!

✤ For example, from historical data


Losses
about EADs and LGDs.
The “real” CR+

✤ The simple approach we have just seen is just a very special and unrealistic version of CR+.!

✤ The model which is actually used by banks is much more complex from a mathematical
point of view, because it introduces more realistic components, e.g. :!

• Correlation/dependence among defaults;!

• Variable default rates;!

• Macroeconomic factors;!

• etc.
Practical use of CR+

✤ An interesting characteristic of CR+ is the possibility of obtaining closed-form


results, once we make some technical assumptions about the parameters of the
model.!

✤ At the same time, CR+ is easy to simulate, and it can also be studied using
computational techniques.
Thank You

You might also like