Page 83 - BKT Annual Report 2024 EN

Notes to the Consolidated Financial Statements for the year ended 31 December 2024 Notes to the Consolidated Financial Statements for the year ended 31 December 2024
(amounts in USD, unless otherwise stated) (amounts in USD, unless otherwise stated)

ii. Expected credit loss measurement Significant Deterioration through relative threshold
Under IFRS 9’s impairment framework, banks are required to recognize ECLs at all times, considering past events, current conditions The bank computes a relative threshold matrix that gives for each rating a prediction of what the expected rating for each time horizon
and forward-looking information, and to update the amount of ECLs recognized at each reporting date to reflect changes in an asset’s should be.
credit risk. This is a more forward-looking approach and results in a timelier recognition of credit losses. Through-the-cycle (TTC) transition matrices give the percentage of counterparties which moved from one rating to another over a
The estimation of credit exposure for risk management purposes is complex and requires the use of models, as the exposure varies specific time interval. TTC matrices over a 10 year horizon are taken since this gives a comfortable horizon for a relative threshold.
with changes in market conditions, expected cash flows and the passage of time. The assessment of credit risk of a portfolio of assets
entails further estimations as to the likelihood of defaults occurring, of the associated loss ratios and of default correlations between Computing relative thresholds for SICR
counterparties. The Group measures credit risk using Probability of Default (PD), Exposure at Default (EAD) and Loss Given Default Given that the Bank does not have sufficient internal historical data to compile transition matrices, the bank relied on proxies provided
(LGD). by external credit rating agencies. External TTC transition matrices for all European/World entities provided by international credit rating
agencies are used. The following steps are then carried out:
The guiding principle of the expected credit loss model is to reflect the general pattern of deterioration or improvement in the credit First the mapping between the internal and external rating systems is performed;
quality of financial instruments. Based on the change of credit risk since initial recognition, assets are classified into 3 stages: From there, a TTC transition matrix with the number of observations (number of entities that changed from one specific rating to another)
“Stage 1” comprises of assets that have not suffered any significant deterioration of credit quality since initial recognition; is used and a weighted average per mapped rating is computed in order to compute the internal ratings TTC matrix.
“Stage 2” comprises of assets that have suffered significant deterioration since initial recognition;
“Stage 3” concerns all assets where a default has occurred. Once TTC matrices are computed, the following steps are taken:
Under this general approach, the ECL for an asset is calculated over different time horizons according to the stage it was assigned to: • The weighted average 1-year probability of default for each rating and for each time horizon is computed by multiplying each TTC
ECL over one year for assets in stage 1; matrix (for the different time horizons) by the PD vector over one year;
ECL over remaining lifetime for assets in stage 2 and stage 3. • Then the weighted average 1-year PDs are mapped to the closest corresponding ratings, which yields the average degradation
matrix in terms of ratings;
The stage assignment is done according to the following rules: • A downgrade is considered significant, with subsequent classification into Stage 2, if the current rating is worse than the relative
Impairment: if the counterparty for the considered asset has defaulted, the asset is assigned to “Stage 3”. An asset is considered as threshold for the respective initial rating and time passed since initial recognition.
having defaulted if any repayment (principal or interest) is overdue for more than 90 days or if the counterparty is in a proven situation
of default (bankruptcy). Forward-looking information incorporated in the ECL models
Rating D (lower than C): Assets with this rating are currently considered to be in “Stage 3”. The TTC PDs are transformed into PIT PDs by taking in account the macroeconomic environment through a set of macroeconomic
Qualitative factors: IFRS 9 has advised to take in account qualitative factors such as watch lists or financial analysis by experts. Similar variables, such as real GDP growth rate, inflation rate, unemployment rate, investments, import change , current account balance,
to the previous case, there is also a second time threshold. In case the repayment of an asset is overdue for more than 30 days and debt, export change etc. This data is sourced from the IMF’s World Economic Outlook, including historical data starting from 1990 up
less than 90 days, it is assigned to “Stage 2”. to today and projections for the upcoming 4 years under three scenarios: baseline, best and adverse. The PIT PD model consists of a
Relative Threshold: if the counterparty has suffered significant deterioration in credit risk, that is if its credit quality since initial recognition simplified form of the Merton model. In this framework, a systemic variable “Xt” which represents the macroeconomic environment is
has dropped more than a specific pre-defined relative threshold, then it is assigned to “Stage 2”. introduced. The sensitivity of each rating’s PD to this variable is obtained via the calibration of the “p” parameter. The model takes in
All assets that are not in the previous cases are assigned to “Stage 1”. account the default rate of each year and calculates “Xt” for each year, which is then regressed with different combinations of macro-
economic variables. Using projections of the macroeconomic variables, the regression formula is used to deduce projections of “Xt”,
Grouping of instruments for losses measured on a collective basis and based on the one factor Merton model the PIT PDs are obtained. The second PD model considers the default rate per rating in
In order to model expected losses on a collective basis, a grouping of exposures is performed on the basis of shared risk characteristics, each year, which enables the calibration of the sensitivity pi for all ratings.
such as credit rating, product type, remaining maturity, etc.
Measuring ECL – Explanation of inputs, assumptions and estimation techniques
The subsequent ECL calculation is based on historical performance data for the relevant group. Where sufficient internal historical data
The ECL calculation is based on the following key parameters:
is not available, the Group has used benchmarking internal/external supplementary data for modelling purposes..
1. Probability of default (PD);
2. Loss given default (LGD); and
The Bank has three main portfolios, which are: 3. Exposure at default (EAD).
- Loan portfolio
This category includes wholesale and individual/retail accounts loans.
- Treasury portfolio ECL is estimated under Baseline (typical), Best (favourable) and Adverse (unfavourable) conditions.
This category includes bonds, treasury bills and equity accounts. The only Point-in-Time estimates are for Probability of Default. LGD assumes Basel estimates and EAD uses amortisation type payment
- Project and Structured Finance schedules. Once all components (PD, EAD, LGD) have been computed, the ECL is estimated under three different scenarios: Baseline
This category includes letters of credit and bank guarantees. (typical), Best (favourable) and Adverse (unfavourable) condition, with weights 52%, 18% and 30% respectively. The final ECL is the
probability-weighted ECL under those three scenarios.

ANNUAL REPORT 2024 28