Estimating the expected credit losses for
loans and advances to customers
In accordance with the provisions of IFRS 9 the
Management is required to determine expected
credit loss (“ECL”) that may occur over either a
12 month period or the remaining life of a
financial asset, depending on the classification
of individual assets into risk categories
("stages"), taking into account the impact of
future macroeconomic conditions on the level of
credit risk allowances.
The Bank’s loan portfolio consists of exposures
assessed for expected credit losses:
●
on an individual basis for individually
significant credit exposures; and
●
on a portfolio basis
with the use of statistical
models which estimate allowances for credit
losses for each of the homogenous portfolios
identified by the Bank.
Estimating the level of allowances for expected
credit losses requires the application of a
significant degree of
judgement
with regard to
the identification of impaired loans and
significant increase in credit risk (“SICR”), the
assessment of the customer's credit quality, the
value of collaterals and expected recoveries.
The management monitors the
correctness
of
performance of the models, by comparing the
estimated results of the models to actual credit
losses (‘back-testing procedures’) to ensure that
the level of allowances for the expected credit
losses for loans and advances to customers is
adequate.
In the models of expected credit losses the
Bank uses large volumes of data, therefore the
completeness and reliability of data can
significantly impact accuracy of the allowances
for credit losses.
Due to the dynamic changes taking place in the
macroeconomic environment, e.g. as a result of
We started our procedures with obtaining an
understanding of the internal control environment in
terms of recognizing and calculating expected credit
losses, and we also checked the effectiveness of
selected key control mechanisms implemented
by
the Bank, in particular:
●
procedures in the area of entering customer data
used for the calculation of expected credit losses;
●
data flow between the Bank's key IT systems and
the ECL calculation tool;
●
procedures for timely and complete identification of
significant increases in credit risk (stage 2) and
impairment (stage 3).
We also assessed whether the methodology used by
the Bank for estimating allowances for expected
credit losses complies with the requirements of IFRS
9. In particular, we assessed the Bank's approach to
the application of the SICR criteria, default definition,
probability of default ("PD"), loss given default
(“LGD”) parameters and
considering
forward-looking
information when calculating expected credit losses.
In the case of individually insignificant loans and
advances, which are assessed for impairment at the
portfolio level, we have performed the following
procedures:
●
assessment of the Bank's assumptions and expert
adjustments used in the model;
●
critical analysis of key judgments and assumptions,
including macroeconomic scenarios and assumed
probabilities of occurrence of individual scenarios;
●
analysis of the model stability and its adaptation to
the current conditions;
●
independent tests of credit risk parameters.
We have engaged our in-house credit risk
modelling
specialists to carry out the above-mentioned
procedures.