In accordance with IFRS 9, for
- Deals assigned to stage 1 and stage 2 as well as
- Non-significant financial assets assigned to stage 3,
provisions for losses on loans and advances are applied collectively to various segments (groups of individual financial assets) with similar risk characteristics.
The solution supports the configuration of segments. Each segment will contain customers' individual deals with similar risk characteristics. It ensures that different models can be applied for deals with different credit risk characteristics during Expected Credit Loss calculation.
This component is necessary if
- segmentation has not already been performed in the source and has not been delivered from a source
- a heterogenous portfolio requires the application of various statistical methods for Expected Credit Loss calculation
In FlexFinance segments can be configured taking criteria such as product type, cost center, type of customer, customer rating etc. into account.
The configuration is organised in sets of rules. A set helps to determine, at a later point in time, which parameter and which characteristics were relevant for segmentation at the time when Expected Credit Loss calculation was performed.
A set contains the parameters and values for segmentation and can be used in a master scenario. Several sets can be maintained next to each other in parallel. For each master scenario, however, only one set can be assigned and is therefore valid. Different parallel master scenarios can take different sets into account.
As the segmentation groups individual deals that carry similar credit risk characteristics, it is necessary to consider this segmentation while
- selecting and applying the appropriate Expected Credit Loss model including PD, LGD and EAD as well as
- considering the historical default information that is linked to the segment.
For example: if a segment "retail loans" and a segment "corporate loans" are considered during Expected Credit Loss calculation, segment-specific historical default information should be used for the calculation of PD and LGD.
On the basis of the combination of segment and stage, the appropriate parameters for Expected Credit Loss calculation can be identified, taking the availability and relevance of performance data into account. For details about stage assignment, please refer to Stage Assignment.