Model Inputs
From Open Risk Manual
Definition
Model Inputs denotes the list of numerical or categorical variables (Risk Data) that are used as mathematical inputs to be processed by a Risk Model
Examples
There is a wide variety of possible model inputs:
- Continuous Numerical values (valuations, hedge sizes, continuous risk scores)
- Discrete Numerical values or Categorical variables (ratings)
- Distributions (rankings of possible scenario losses)
- Discrete indicators (yes / no)
- Tabular Data (risk allocation to positions)
Issues and Challenges
- Model inputs are frequently the outcome of previous Model Outputs creating chains of dependencies and possible Model Risk propagation
- Careful documentation of all model inputs in use is part of good Model Governance
