Log in / Register
Home arrow Management arrow Risk management in banking


The resiliency of a structure designates its ability to sustain variations of the risk factors without generating losses to the various classes of structured notes. Rating agencies focus on the risk of issued notes, while credit portfolio models make up the foundation for modeling the risk of the underlying credit portfolio. Both methodologies serve for assessing how resilient a securitization structure is and should be used in conjunction.

Rating Methodologies for Structured Notes

Critical parameters that determine the resiliency of structures depend on the type of structure. For consumers or mortgages loans, typical critical parameters include:

• the delinquency rate (delays in payment)

• the charge-off rate (losses due to default)

• the payment rate (both monthly payments of interest and principal)

• the recovery rate (both percentage amount and timing of recoveries)

• the average yield of the portfolio of loans.

The degree of over-collateralization of each note measures the risk, when such factors are stress-tested.

The stress tests serve for rating the notes, in addition to measuring the resiliency of the structure. The main purpose of such stress scenarios is finding out the stress level of each factor that each note can sustain without impairing its risk. For example, a common way of assessing the risk of an asset-backed note is to use multiples of expected charge-offs for finding out when the credit risk of notes is impaired. For a given note, the higher the multiple sustainable by the note without any loss, the higher the rating. For instance, the required minimum multiple for an "AAA" scenario is highest compared with a riskier subordinated note. An "AAA" scenario designates the minimum multiple required to preserve the AAA rating. Minimum required multiples of charge-off rates are, for example, 5 or 6 times the expected average charge-off rate for senior notes.

One easy way to assign ratings to notes is to map these minimum multiples with the note ratings. Evidently, the minimum multiple required to have a given rating also depends on the quality of the assets. Hence, the AAA scenario, or the "AAA multiple" of expected loss, is lower when the average risk of assets is low than when it is higher. This is a shortcut to the full modeling of the risk of the portfolio of assets. The full assessment of the structured note ratings requires plugging these scenarios to the "spreadsheet" model of the waterfall of cash flows. Various simulations show when losses hit each note.

Such methodologies make sense to the extent that the stress scenarios are in line with past experience. They are not inconsistent with modeling the portfolio loss distribution, since they stress risk factors that affect this distribution.

But such methodologies are no substitutes for credit portfolio models that generate the loss distribution of the underlying asset portfolio and allow modeling the risk of each note. Notably, stress testing the correlation of losses within the portfolio of assets is mostly important for at least two reasons. First, the portfolio loss distribution is highly sensitive to the uniform correlation; and, second, it has been shown that default correlations are sensitive to economic conditions. One of the lessons of the 2007-2008 crisis is that the risk of notes issued by securitization vehicles was massively under-estimated. Portfolio models did not help either since securitization arrangers use them, but a model is as good as the inputs are. Models should have allowed conducting more stringent stress tests, both on the quality of inputs and on loss correlations. In any case, they remain the unique tool for better addressing dependencies than simpler techniques such as those described above.

Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
Business & Finance
Computer Science
Language & Literature
Political science