Difference between revisions of "Loss Given Loss"

From Open Risk Manual
 
Line 5: Line 5:
  
 
== Loss Given Loss Models ==
 
== Loss Given Loss Models ==
Loss Given Loss models are similar in structure and estimation to [[Loss Given Default Models]], the main difference being the increased emphasis on factors that affect recovery values (degree of collateralisation, nature of collateral etc.)
+
Loss Given Loss models are similar in structure and estimation to [[Loss Given Default Models]], the main difference being the increased emphasis on factors that affect recovery values (degree of collateralisation, nature of collateral etc.) and the fact the modelling of the [[Cure Rate]] is no longer required.
  
 
== Issues and Challenges ==
 
== Issues and Challenges ==

Latest revision as of 12:37, 8 November 2019

Definition

Loss Given Loss (LGL) is a Risk Parameter that captures the uncertainty about the actual loss that will be realized when it is given (conditioned on) that such an event produces a material non-zero loss.

Loss Given Loss refines the broader Loss Given Default or Loss Given Impairment risk parameters. It is a useful concept in circumstances where there is a non-negligible probability that problematic credit exposures will return to performing status. In other words in circumstances where there is a measurable Cure Rate.

Loss Given Loss Models

Loss Given Loss models are similar in structure and estimation to Loss Given Default Models, the main difference being the increased emphasis on factors that affect recovery values (degree of collateralisation, nature of collateral etc.) and the fact the modelling of the Cure Rate is no longer required.

Issues and Challenges

  • The estimation of an LGL model is narrowing the sample of credit events (effectively implying a different definition of default). Hence care must exercised to ensure the overall system of risk parameters is consistent. [1]

References

  1. Proposed Supervisory Guidance for Internal Ratings-Based Systems for Credit Risk, Federal Register / Vol. 72, No. 39 / Wednesday, February 28, 2007

Contributors to this article

» Wiki admin