In IFRS 9 the transfer from stage 1 to stage 2 influences the way how in which ECL (Expected Credit Losses) are calculated. For stage 1, 12-month expected credit losses need to be considered for risk provisioning. For stage 2 lifetime expected credit losses need to be considered. Hence for Customercustomers, the stage assignment does not influence the calculation of the ECL because the current portfolio only contains deals with a term that does not exceed 12 months.
Nonetheless, the stage assignment for Customer customers is relevant, as the assigned stage impacts the disclosures required by the supervisor: depending on the assigned stage, expected credit losses of losses for individual deals must be assigned to different reporting positions.
To cope with fulfil these requirements, the solution covers the stage assignment.
We assume that, for stage assignment, the consideration of the number of days past due is sufficient for Customer sufficienta customer. Hence in this case, the auditor in line with GPPC requests that beside , besides this quantitative parameter, a qualitative parameter needs to be considered in line with GPPC.
Model "Staging DPD"
The functionality offered functionality assigns , assigns individual deals to a stage using the information about the days past due. For this purpose, the number of days past due shall be delivered on at individual deal level from the source.
In case If the number of days past due exceeds 30 days, a significant deterioration of credit quality is assumed and the individual deal will be assigned to stage 2. In case If the number of days past due exceed exceeds 90 days, the individual deal will be assigned to stage 3. The underlying configuration is done performed in a separate table.
Model "Staging DPD & PD"
The model “Staging DPD & PD” considers both, number of days past due and PD for stage assignment. The PD needs to be considered as “PD changes since deal origination”. Both, the percentage of PD change since deal origination and number of days past due, need to be linked with stage 1, 2 or 3. This model will assign the stage with minor credit quality in case if a minimum of the configured criteria is fulfilled.
The model “Staging DPD & Scoring” considers both, number of days past due and the change in scoring. Similar to the model “Staging DPD & PD”, this model links a quantitative and qualitative parameter for stage assignment. Instead of “relative PD - changes”, this model works with relative changes in score points.
1.2 ECL calculation
For different stages, IFRS 9 requires calls for different stages different ways of how to calculate expected calculating expected credit losses.
As the term of the individual deals (positions) does not exceed 12 months, the formula which that needs to be applied for stage 1 and stage 2 is equal.
IFRS 9 requires the consideration of probability-weighted scenarios for the calculation of ECL.
The processing of the ECL calculation is split and covered by the following models. Depending if Customer Depending on whether a customer decides (Variant A) to use the transfer todays of today's PD from capital requirements to a PD (PIT) or if it if he decides (Variant B) to use the available historic scoring information in combination with macroeconomic parameters to derive a PD (PIT) directly, different models need to be selected:
- Variant A: Model “Macroeconomic factor”
- Variant A: Model “Transfer PD (TTC) to PD (PIT)”
- Variant B: Model “Derive PD (PIT) based on historical the basis of historical scoring information and macroeconomic parameters”
The model “ECL calculation” calculates the expected credit loss while taking maximum a maximum of 3 different macroeconomic scenarios into account.
It is assumed that the ECL of a scenario will always be calculated based on the basis of
ECL = ∑PD * LGD * EAD * DCF(EIR)
This model calculates EAD for different time periods over the life time lifetime of a financial instrument, e.g. EAD in 3 months, 6 months, 1 year, 2 years etc. for a financial instrument with regular a regular repayment plan, e.g. annuity, EAD for a future period will be different to current book value due to repayments.
To reflect possible repayments, the estimated cash flow plan over the life time lifetime of a financial instrument will be generated based on the basis of the contractual agreement.
Based on the cash flow plan, the EAD for different periods in the future will be calculated based on using amortised cost and adjustment of 3 months overdue which may occur before default.
This model derives macroeconomic factors based on the basis of macroeconomic parameters.
This model is solely required in case Customer a customer decides to transfer the PD (TTC) into a PD (PIT) applying macroeconomic factors.
The model studies correlation between default pattern patterns in the past periods with relevant macroeconomic parameters available in periods. By applying regression the regression method, correlation can be set up between derivation of a default rate for a specific period comparing compared with the average default rate over periods and underlying macroeconomic parameters. For each scenario, the bank needs to feed enter a maximum of 5 parameters for a timeline.
PD (TTC) is commonly used in risk management with the concept of “through the cycle (TTC)”. It is conceptually different to requirements in IFRS 9 which requires forward- looking over forware over life time of a financial instrument with consideration of economic scenarios.
The target of this model is to utilise existing PDs by applying macroeconomic factors. By adjustment via macroeconomic factors, the economic scenario in a specific period will be considered. PD (PIT) for each period in the future is calculated by applying PD (TTC) with correlation between derivation of the default rate and projected macroeconomic parameters of for that period.
1.2.2 Variant B - PD (PIT)
This model derives PDs (PIT) for the life time lifetime of a deal directly from available historical default data and macroeconomic parameters. The PDs (PIT) will be used directly in ECL calculation in IFRS 9.
For expected credit losses in stage 1 and stage 2, it is sufficient to book post the expected credit loss into risk provisioning. The entire portfolio for the ECL Workbench needs to be categorised in AC. It is assumed that all AC related valuation elements such as residual capital, open amortisation and interest accrual are booked posted and delivered by the core system.
A common solution for the consideration of expected credit losses in the balance sheet is the booking posting “D Risk provision expense C risk provision”. These bookings entries will be reversed at on the following booking posting date. Instead of booking posting incremental changes of in the ECL, this approach books posts the full amount of ECL.
Model "Journal Entries ECL"
This model works with a fixed accounting logic. It generates, for each individual deal, a record which that contains exactly one debit entry on an expense account and one credit entry on a risk provision account.
Which account for the individual deal will be booked posted is subject to a configuration in a specific table. For each group of accounts (expense of risk provision), multiple general ledger accounts can be defined. For each individual deal, the appropriate general ledger account will be identified based on descriptive the basis of descriptive parameters.
The output of this model are is debit/credit entries on at individual deal level, provided in a table or file.
1.4 Report layer
The solution offered solution supports to fulfil reporting , supports compliance with reporting requirements according to FINREP standardstandards.
- Financial assets subject to impairment that are past due