Difference between revisions of "Expert Scorecards"
Wiki admin (talk | contribs) |
Wiki admin (talk | contribs) |
||
| Line 1: | Line 1: | ||
== Definition == | == Definition == | ||
| − | '''Expert Scorecards''' are a class of simple ''linear'' mathematical models that are used in [[Credit Risk]] and [[Operational Risk]] context, in particular [[Risk Acceptance]] decisions for new clients or in [[Risk Analysis]] for existing clients | + | '''Expert Scorecards''' are a class of simple ''linear'' mathematical models that are used in [[Credit Risk]] and [[Operational Risk]] context, in particular [[Risk Acceptance]] decisions for new clients or in [[Risk Analysis]] for existing clients. |
| − | While any quantitative | + | == Context == |
| + | Many scorecard are developed by experts rather than using a quantitative (or statistical) process. | ||
| + | |||
| + | While any linear quantitative model shares some commonalities with a scorecard it is best to keep separate terminologies for statistically versus expert based scorecards | ||
== Methodology == | == Methodology == | ||
| Line 10: | Line 13: | ||
== Advantages == | == Advantages == | ||
| − | The main reason scorecards see widespread use is because the technique allows non-quantitative staff to develop a risk analysis tool (e.g., | + | The main reason scorecards see widespread use is because the technique allows non-quantitative staff to develop a risk analysis tool (e.g., using spreadsheet tools). In the absence of adequate data a scorecard approach may be the only means of articulating an expert's best estimate. |
== Issues and Challenges == | == Issues and Challenges == | ||
| − | Formally a scorecard is a | + | Formally a scorecard is a generalized linear model. In practice the development of scorecards may be missing many elements of proper [[Model Development]] practice (e.g. in the unbiased selection of variables, the proper weighting etc. |
== See Also == | == See Also == | ||
* [[Credit Scorecard]] | * [[Credit Scorecard]] | ||
| + | * [[Energy Risk Scorecard]] | ||
== References == | == References == | ||
| Line 23: | Line 27: | ||
[[Category:Operational Risk]] | [[Category:Operational Risk]] | ||
[[Category:Credit Scoring]] | [[Category:Credit Scoring]] | ||
| + | [[Category:Energy Risk]] | ||
Latest revision as of 19:55, 11 March 2024
Contents
Definition
Expert Scorecards are a class of simple linear mathematical models that are used in Credit Risk and Operational Risk context, in particular Risk Acceptance decisions for new clients or in Risk Analysis for existing clients.
Context
Many scorecard are developed by experts rather than using a quantitative (or statistical) process.
While any linear quantitative model shares some commonalities with a scorecard it is best to keep separate terminologies for statistically versus expert based scorecards
Methodology
The structure of the scorecard is a list of factors / characteristics (that may be transformed /scaled) and are given statistical weights towards a total score. The sum of the weights must equal unity.
A score does not imply in itself any concrete probability or probability range for the occurrence of an event. Sometimes this step is accomplished via a Default Probability Table
Advantages
The main reason scorecards see widespread use is because the technique allows non-quantitative staff to develop a risk analysis tool (e.g., using spreadsheet tools). In the absence of adequate data a scorecard approach may be the only means of articulating an expert's best estimate.
Issues and Challenges
Formally a scorecard is a generalized linear model. In practice the development of scorecards may be missing many elements of proper Model Development practice (e.g. in the unbiased selection of variables, the proper weighting etc.
