A list of models utilized in the formulation of the ICAAP should be provided giving relevant and appropriate details as given below:
|
| ■ | The key assumptions and parameters within the capital modeling work and background information on the derivation of any key assumptions.
|
| ■ | How parameters have been chosen including the historical period used and the calibration process.
|
| ■ | The limitations of the model.
|
| ■ | The sensitivity of the model to changes in the key assumptions or parameters chosen.
|
| ■ | The validation work undertaken to ensure the continuing adequacy of the model.
|
| ■ | Whether the model is internally or externally developed. If externally acquired its generic name and details on the model developer.
|
| ■ | Details should also be provided as to the extent of its acceptance by other regulatory bodies, users in the international financial community, overall reputation and market acceptance.
|
| ■ | Specific details on the applications within the Bank, i.e. measurement of risks such as credit, liquidity, market, concentration, etc. or for the purpose of establishing internal credit risk classification ratings, risk estimates, PDs, LGDs, EADs, etc.
|
| ■ | Major merits and demerits of the chosen models.
|
| ■ | Results of the model validation obtained through
|
| | ■ | Back testing / Scenario testing
|
| | ■ | Analysis of the internal logic
|
| ■ | Major methodologies or statistical technique used, i.e. value at risk models employing methods such as variance/co-variance; historical simulation, Monte Carlo method, etc.
|
| ■ | Confidence levels embedded for regulatory capital, economic capital, or for external rating purposes.
|
Further, the explanation of the differences between results of the internal model for Pillar 1 would be set out at the level at which the ICAAP is applied. Therefore, if the firm's ICAAP document breaks downs the calculation by major legal regulated entities, an explanation for each of those individual entities would be appropriate.
|
SAMA would expect the explanation to be sufficiently granular to show the differences at the level of each of the Pillar 1 risks.
|
Data definition, i.e. whether the source is external or internal and if any data, manipulation of external data has been done for it to conform with internal data.
|
