Model Management System

Abstract

A model management system receives data for an inventory of models including data regarding issues for each of the models. The model management system determines a model risk score for each of the models based on the issues for each of the models. As the issues are opened and closed, the model management system receives updates and in real-time, updates the model risk score. The model risk score can be determined using a severity assigned to each issue for each of the open issues for a given model as well as the number of issues at each severity. Risk scores can be calculated across a line of business and across an enterprise, for example. The model management system also provides one or more of: model development, model validation, model integration, model use, model maintenance, and model retirement.

1 . A computer-implemented method, comprising:
 determining a model risk score for a model based on issues for the model, the issues having issue scores, the model being a quantitative mathematical or statistical construct for simulating events;   determining a highest issue score of the issue scores;   setting a model risk base to a value equal to the highest issue score;   calculating a model risk increment based on an amount of the issues and a severity of the issues;   updating the model risk score based on the model risk base, the model risk increment, and a model risk adjustment to provide an updated model risk score;   while the model is in a suspended state, making modifications to the model as simultaneous modifications to the model are being made by multiple stakeholders to provide a modified model; and   releasing the modified model, causing the modified model to change from the suspended state to a live state and causing the updated model risk score or a modified version of the updated model risk score to be accessible via a user interface.   
     
     
         2 . The computer-implemented method of  claim 1 , wherein while the model is in the suspended state, the model is unavailable for use in determination of another risk score for another model that is in the live state. 
     
     
         3 . The computer-implemented method of  claim 1 , further comprising:
 determining a performance rating for the model by:
 using one or more key performance indicators; and 
 testing the one or more key performance indicators against the model. 
   
     
     
         4 . The computer-implemented method of  claim 1 , further comprising:
 providing a notification to a stakeholder associated with the model when there is a change in one of the issues for the model.   
     
     
         5 . The computer-implemented method of  claim 1 , wherein the model is for predicting credit losses to estimate exposure to credit risk based on existing or prospective extensions of credit. 
     
     
         6 . The computer-implemented method of  claim 1 , wherein the model is for simulating economic scenarios to determine sufficiency of capital reserves. 
     
     
         7 . The computer-implemented method of  claim 1 , wherein the issues include an absence of consideration of certain data relevant to the model. 
     
     
         8 . The computer-implemented method of  claim 1 , wherein the issues include a misapplication of an equation or formula. 
     
     
         9 . The computer-implemented method of  claim 1 , wherein the issues include a change to an assumption on which the model is based. 
     
     
         10 . The computer-implemented method of  claim 1 , wherein the issues include an inaccuracy in an assumption on which the model is based. 
     
     
         11 . The computer-implemented method of  claim 1 , wherein the issues include a validation issue. 
     
     
         12 . The computer-implemented method of  claim 1 , wherein the issues include a line of business evaluation issue. 
     
     
         13 . An electronic computing device, comprising:
 a processing unit; and   system memory, the system memory including instructions that, when executed by the processing unit, cause the processing unit to:
 determine a model risk score for a model based on issues for the model, the issues having issue scores, the model being a quantitative mathematical or statistical construct for simulating events; 
 determine a highest issue score of the issue scores; 
 set a model risk base to a value equal to the highest issue score; 
 calculate a model risk increment based on an amount of the issues and a severity of the issues; 
 update the model risk score based on the model risk base, the model risk increment, and a model risk adjustment to provide an updated model risk score; 
 while the model is in a suspended state, make modifications to the model as simultaneous modifications to the model are being made by multiple stakeholders to provide a modified model; and 
 release the modified model, causing the modified model to change from the suspended state to a live state and causing the updated model risk score or a modified version of the updated model risk score to be accessible via a user interface. 
   
     
     
         14 . The electronic computing device of  claim 13 , wherein while the model is in the suspended state, the model is unavailable for use in determination of another risk score for another model that is in the live state. 
     
     
         15 . The electronic computing device of  claim 13 , wherein the system memory further includes instructions that, when executed by the processing unit, cause the electronic computing device to:
 determining a performance rating for the model by:
 using one or more key performance indicators; and 
 testing the one or more key performance indicators against the model. 
   
     
     
         16 . The electronic computing device of  claim 13 , wherein the system memory further includes instructions that, when executed by the processing unit, cause the electronic computing device to:
 provide a notification to a stakeholder associated with the model when there is a change in one of the issues for the model.   
     
     
         17 . The electronic computing device of  claim 13 , wherein the issues include an absence of consideration of certain data relevant to the model. 
     
     
         18 . The electronic computing device of  claim 13 , wherein the issues include a misapplication of an equation or formula. 
     
     
         19 . The electronic computing device of  claim 13 , wherein the issues include a change to an assumption on which the model is based. 
     
     
         20 . The electronic computing device of  claim 13 , wherein the issues include an inaccuracy in an assumption on which the model is based.