Methodological Guidelines for Applying IFRS 9 to Asset Classification and Loan Loss Provisioning

1
Approved by the decision of the Management Board of the Republic of Azerbaijan dated 23 December 2025 Protocol № 45/2 Methodological Guidelines for the application of International Financial Reporting Standard 9 for asset classification and loan loss provisioning

1. General provisions
   1.1. These Methodological Guidelines (hereinafter – the Guidelines) have been
   developed in accordance with the requirements of International Financial Reporting Standard 9 (hereinafter – IFRS 9) and are applied in conjunction with the ‘Regulation on asset classification and creation of specific reserves for loan loss provisioning’ approved by Decision No. 29/1-1 of the Management Board of the Central Bank of the Republic of Azerbaijan dated 22 July 2022.
   1.2. These Guidelines set out recommendations solely with respect to the classification
   of financial assets and the implementation of expected loss provisioning in banks and local branches of foreign banks operating in the Republic of Azerbaijan (hereinafter – banks).
   1.3. These Guidelines do not cover multiple topics addressed in IFRS 9, such as hedge
   accounting, classification of financial liabilities, approaches to fair valuation of FVOCI/FVTPL assets, and accounting of embedded derivative financial instruments (except where they affect classification).
2. Definitions
   2.1. The definitions used in these Guidelines bear the following meanings:
   2.1.1. significant increase in credit risk (hereinafter – SICR) – difficulties encountered by
   the borrower in fulfilling their obligations within a specified period.
   2.1.2. default – the inability of the borrower to meet their obligations within a specified period.
   2.1.3. probability of default (hereinafter – PD) – the likelihood of a default event occurring
   within a specified period.
   2.1.4. marginal PD (hereinafter – marginal PD) – the probability of default occurring
   within a specific time interval, particularly between any two future dates.
   2.1.5. cumulative probability of a default (hereinafter – cumulative PD) – the probability
   of default occurring between the reporting date and any future date.
   2.1.6. point-in-time probability of default (hereinafter – PIT PD) – the probability of
   default assessed in accordance with the current economic and financial conditions.

2
2.1.7. through-the-cycle probability of default (hereinafter – TTC PD) – the probability of
default assessed for an economic, financial, or product cycle.
2.1.8. loss given default (hereinafter – LGD) – the level of loss in the event of default.
2.1.9. exposure at default (hereinafter – EAD) – the amount of credit exposure owed by
the borrower at the time of default.
2.1.10. discount factor (hereinafter – DF) – used to discount shortfalls between expected
cash flows and contractual cash flows to the reporting date.
2.1.11. expected credit loss (hereinafter – ECL) – the average credit loss weighted by the
relevant probabilities of default.
2.1.12. the effective interest rate (hereinafter – EIR) – the rate that exactly discounts
estimated future cash payments or receipts through the expected life of the financial asset to the gross carrying amount of a financial asset.
2.1.13. solely payments of principal and interest (hereinafter – SPPI) – payments
consisting solely of principal and interest.
3. Classification
3.1. Under sub-tems 4.1.1–4.1.4 of IFRS 9, financial assets are classified at initial
recognition into one of the categories specified by IFRS 9:
3.1.1. Amortized cost (AC):
3.1.1.1. business model: assets held to collect contractual cash flows.
3.1.1.2. assets passed the SPPI test.
3.1.2. Fair value through other comprehensive income (FVOCI):
3.1.2.1. business model: assets held for both collecting and selling contractual cash flows.
3.1.3. Fair value through profit or loss (FVTPL):
3.1.3.1. all other financial assets that do not fall into the above two categories.
3.1.3.2. assets held for trading.

3
3.2. When classifying financial instruments, banks ensure a robust governance process.
The process primarily involves analyzing the business model and performing the SPPI test for financial instruments. The classification of financial instruments into various categories is reviewed by the competent structural unit and assigned to the appropriate category.
3.3. The classification categories can be summarized as follows:
3.4. The SPPI test:
3.4.1. Banks assess contractual terms of each financial asset (or each type of standardized
instrument) at initial recognition to identify any features that could cause variability in cash flows beyond principal and interest payments. If such variability is found, the asset fails the SPPI test and is classified as FVTPL.
3.4.2. Examples of contractual terms that cause variability in cash flows beyond principal
and interest and lead to unconditional fail of the SPPI test:
3.4.2.1. Non-basic indexation: if payments of interest or principal are linked to an equity
index, commodity prices, or non-interest variables (e.g., if bond’s payments are linked to the price of gold, it will fail the SPPI test—its cash flows will not be limited to principal and interest only).
3.4.2.2. Embedded derivatives that modify cash flows in a non-basic way: e.g., a
conversion option to equity for a debt instrument (convertible bond) introduces equity-linked cash flow characteristics.

4
3.4.2.3. Inverse floating rates or exotic interest formulas: e.g., an interest rate formula that
inverts an index or is otherwise not a consideration for time value of money.
3.4.2.4. Terms that involve sharing in the profits of the borrower, which would introduce
elements of equity risk rather than simple interest and principal repayment
3.4.3. Existence of interest rate caps/floors; currency clause and repayment option does
not prevent instruments from passing SPPI test, unless these terms significantly deviate from market practice and reasonable estimates.
3.4.4. Banks should maintain a central SPPI library of instrument types and SPPI
outcomes. This library should list common contract features and whether they are deemed SPPI-compliant (referencing IFRS 9 guidance for basic (B4.1.10) and modified time value features (B4.1.13–B4.1.17)). New product proposals should be reviewed against this library.
3.4.5. If an instrument could fall into multiple categories, the bank should make a decision
considering both accounting policy and risk management.
3.4.6. The analysis should include simulation of cash flows under various scenarios to
confirm that they reflect principal and interest payments.
3.5. Eligibility of sales from amortized cost category:
3.5.1. Instruments classified as amortized cost can be sold in certain circumstances. In
such cases, the bank should be ready to prove that the sale was not planned or considered at initial recognition. For these reasons, sales should be infrequent and immaterial. Banks should quantify thresholds for frequency and significance in their methodologies. Threshold values cannot be more than:
3.5.1.1. Frequency: 1% of borrowers in corporate and retail portfolios, separately.
3.5.1.2. Significance: 1% of the amounts in corporate and retail portfolios, separately.
3.5.2. Banks should monitor actual sales vs defined thresholds on frequency and
significance on a quarterly basis and ensure appropriate reporting.
3.6. Reclassification:
3.6.1. Changes in the business model (and thus reclassifications) should occur only in
exceptional circumstances, and any change in the management of assets should result from management’s decision.
3.6.2. Taking into account expected market developments, reclassification is not allowed
to be made based on ‘artificial business model changes’ (e.g., reclassification of fixed rate instruments from FVOCI or FVTPL into AC in case of increase of interest rates) or expected changes on the level of the borrower.
3.6.3. As stated in IFRS 9, Para 5.6.1, if a reclassification is deemed eligible, the new
classification is applied prospectively from the reclassification date.
3.6.4. The bank may not restate any previously recognized gains, losses (including
impairment gains or losses) or interest.

5
4. Impairment
4.1. Scope of impairment under IFRS 9:
4.1.1. These guidelines require institutions to develop and implement ECL methodology
to recognize loss allowances for expected credit losses for all financial assets that are subject to impairment under IFRS 9 requirements [5.5.1]:
4.1.1.1. All financial assets measured at amortised cost (e.g., loans, receivables, lease
receivables, debt securities measured at amortised cost, etc.).
4.1.1.2. Debt instruments at FVOCI. Ror debt instruments measured at FVOCI the ECL
should be recognized in profit or loss statement. A corresponding adjustment should be made to Other Comprehensive Income (OCI) to ensure that the carrying amount of the instrument reflects its Fair Value (FV).
4.1.1.3. Loan commitments, financial guarantees, letters of credit and other credit-related
off-balance-sheet commitments not measured at FVTPL.
4.1.2. Financial instruments measured at FVTPL are not subject to IFRS 9 Impairment. All
changes in credit quality should be included in FV of the instrument directly and no credit￾related losses in profit or loss statement are needed.
4.1.3. ECL should be recognized in profit or loss statement at initial recognition of an in￾scope asset. At each reporting date an entity increases or decreases loss allowance for expected
credit losses to reflect changes in credit risk
4.2. Impairment provisions for financial instruments are calculated based on changes in
the credit quality of the instruments after initial recognition:
4.2.1. Stage 1 – assets with no significant increase in credit risk after initial recognition.
This category comprises financial instruments for which provisions are created based on 12- month expected credit losses, and interest income is calculated from the gross carrying amount. At initial recognition, financial instruments are classified either as purchased or originated credit-impaired assets (hereinafter – POCI) or as Stage 1 assets. At the reporting date, instruments that fail to meet the criteria for Stage 2 or Stage 3 are classified as Stage 1.
4.2.2. Stage 2 – assets with significant increase in credit risk after initial recognition. This
category comprises financial instruments for which provisions are created based on expected credit losses over the lifetime of the asset, and interest income is calculated from the gross carrying amount. If, at the reporting date, there is a significant increase in credit risk compared to the expected credit risk level at initial recognition, the financial instrument (on contract level) is transferred from Stage 1 to Stage 2.
4.2.3. Stage 3 – assets with credir impairment. This category comprises financial
instruments for which provisions are created based on expected credit losses over the lifetime of the asset, and interest income is calculated from the net carrying amount (after deducting the recognized impairment). If a default is identified at the borrower level, all contracts of the borrower are considered to be in default if any default (stage 3) criteria are present at the

6 reporting date. If none of the borrower's contracts meet the default criteria, the impairment stage is determined at the level of the financial instrument
4.2.4. POCI – credit-impaired assets on initial recognition. Lifetime ECL is built into the
credit-adjusted effective interest rate; ECL is recognized only to the extent of subsequent changes in lifetime ECL (no Stage 1/2/3 allocation since they start impaired).
4.2.4.1. ECL for the POCI assets are recommended to be calculated based on individual
approach described in Para 4.8.2 of these guidelines.
4.2.4.2. POCI financial assets include the following:
4.2.4.2.1. Assets acquired at a significant discount due to deterioration in credit quality
(e.g., assets acquired through auctions).
4.2.4.2.2. Assets that meet impairment criteria at the time of issuance or acquisition.
4.2.4.2.3. Reclassification of Stage 3 credit exposures, following modification, as newly
originated loans.
4.2.4.2.4. Newly originated loans granted to credit-impaired borrowers. Credit
impairment is understood as a significant deterioration in creditworthiness, indicating the borrower’s inability to meet its contractual obligations.
4.2.4.3. An asset acquired at a discount not related to the borrower’s unsatisfactory credit
quality is not deemed a POCI asset.
4.2.4.4. No impairment provisions are created at the time of initial recognition of POCI
assets; initial recognition is based on fair value.
4.2.4.5. POCI assets are evaluated based on forecasted expected cash flows the life of
the financial instrument from the time of recognition until the time of derecognition. The approach described in Para 4.8.2 can be used for such calculations.
4.2.4.6. The initial recognition amount of POCI assets (PV(t0)) is equal to:
4.2.4.6.1. the purchase price, if acquired.
4.2.4.6.2. the fair value of the asset at the recognition date (if originated).
4.2.4.7. No provision is created at the initial recognition date. For each subsequent
reporting date, the bank recognizes or derecognizes ECL for POCI assets in the amount of the accumulated expected losses over the asset's lifetime from the initial recognition date.
4.3. Allocation of impairment stages:
4.3.1. A symmetrical transition principle is followed during allocation of impairment
stages for assets at each reporting date: when transition criteria are no longer relevant, financial instruments are transferred from Stage 2 to Stage 1 or from Stage 3 to Stage 1 (or Stage 2).
4.4. Materiality thresholds for DPD calculation:
4.4.1. Institutions may establish materiality thresholds to start DPD counting, but it
should be limited by:
4.4.1.1. Monetary threshold: AZN 200 for all borrowers.
4.4.1.2. Relative threshold: 1% of the on-balance exposure.

7
4.5. Criteria for rehabilitating financial instruments and transitioning to Stage 1:
4.5.1. For both corporate and individual customers, to be reclassified to Stage 1, any
indicators of Stage 2 or 3 should have been absent for at least 90 days before the reporting date.
4.6. Criteria for Stage 2 – SICR:
4.6.1. Banks should consider a SICR to have occurred where the PD at the financial
instrument level has increased by two times or more since initial recognition. The comparison should be based on cumulative PD over the remaining contractual life of the credit exposure.
4.6.2. Methodology for calculating delinquent days under FIFO and LIFO:
4.6.2.1. In the context of counting DPD for Stage 2 and Stage 3, the FIFO approach
assumes that the oldest unpaid amount (either full tranches or material part of it according to materiality thresholds stated in Stage 3 triggers description) is settled first, which results in underestimation of delinquency.
4.6.2.2. The LIFO approach assumes that newest unpaid amounts are paid first, which
reflects prudent approach to delinquency assessment. For instance, if client missed 3 monthly payments and then repaid 1 of these payments - DPD by FIFO approach would be 60 days and DPD by LIFO approach would be 90 days.
4.6.2.3. The LIFO method is used for the stages and modelling of ECL risk metrics.
4.6.3. Criteria that may lead to a borrower being classified in Stage 2 subsequent to initial
recognition:
Asset type No Stage 2 criteria
Business loans
1
The failure to provide periodic reports as specified in the contract (e.g., balance sheet, income statement) 2 The necessity of securing funds from additional sources to ensure debt repayment (e.g., asset sales, additional funds raised by the borrower) or providing additional collateral for the loan 3 Continuous losses by the borrower for more than one year (excluding scenarios where the borrower is expected to operate at a loss for a specified period according to the business or strategic plan) 4 Bullet schedule of repayment: term of more than three years, and for which repayment of more than 50% of the principal debt is made in the last year of the loan 5 More than 20% drop in relevant external market indicators. (e.g., the borrower is oil producer and oil prices dropped). This trigger is a rebuttable presumption 6 10% decrease in capital or negative capital during the 12- month period. This trigger is a rebuttable presumption

8
7
10% reduction in estimated future operating cash flows (OCF=Net Income + Non-cash Expenses + Changes in Working Capital).This trigger is a rebuttable presumption. 8 Current debt service coverage ratio (the ratio of discounted cash flows available for debt servicing to discounted principal and interest payments) is lower than 100%. This trigger is a rebuttable presumption 9 A 20% decrease in revenue or 20% increase in expenses. This trigger is a rebuttable presumption 10 The presence of force majeure circumstances, as well as other circumstances that caused material damage to the borrower (co-borrower), but did not lead to the termination of his activities Business/consumer loans 11 The issuance of loans to an unhedged borrower in foreign currency (borrower has no revenue in the currency of the loan and absence of the FX hedge for this loan purchased by institution) Business/Consumer/Mortgag e loans 12 31-90 days past due for all segments, except transactions with financial institutions and issuers of securities 13 The existence of risks that could affect the borrower's ability to pay 14 Negative events affecting the borrower’s ability to pay due to a deteriorating economic environment (e.g., two instances of being more than 30 days overdue in the last six months) 15 Absolute 1 year PD of the borrower on the reporting date exceeds 20% 16 Restructured assets. This trigger is a rebuttable presumption. The bank may determine that the trigger is not applicable; however, compelling evidence must be obtained, and documented. 17 Collateral value decreases more than 30% since initial recognition 18 Suspicious loan reduction (e.g. through sale of collateral). This trigger is a rebuttable presumption 19 Breach of contractual obligation (covenant) that the bank has not waived. This trigger is a rebuttable presumption 20 Legal action taken against the borrower with significant probability “to lose” (expert judgment) affecting its financial condition

9
Other 21
The presence of debt that is 1-7 business days past due for transactions with financial institutions and issuers of securities
4.6.4. Exemption for low credit risk assets:
4.6.4.1. Where a borrower has a credit rating higher than the minimum investment-grade
rating assigned by Fitch Ratings, Standard & Poor’s, or Moody’s (hereinafter – international rating agencies), or where the 12-month PD is below 1% as at the reporting date, the criterion of ‘PD has increased by 2 times since initial recognition at the facility level’ is not applicable and borrower or facility stays in Stage 1. For financial institutions and issuers of debt securities, credit ratings are determined based on assessments from international rating agencies.
4.7. Criteria for Stage 3 – Default:
4.7.1. When a default occurs, the financial instrument is transferred from Stage 1 or Stage
2 to Stage 3. For the purposes of these guidelines, a default event is considered equivalent to an impairment event.
4.7.2. The impairment event for a financial instrument is determined at the level of the
financial instrument across all portfolios.
4.7.3. The criteria for impairment include the following:
Asset type No Stage 3 criteria
Business loans
1
The counterparty/issuer has been declared bankrupt by a court or is undergoing bankruptcy proceedings related to the borrower. This criterion applies to all portfolios except individual customer portfolios 2 Revocation of the license and application of temporary administration (applies to financial institutions and issuers of securities) 3 If the borrower represents 20% of assets or 10% of revenues of “the group of connected clients” and meets the Stage 3 criteria of the institution, all exposures of “the group of connected clients” are classified in stage 3 4 External factors causing material damage and preventing further commercial activity 5 Concentration risk: Loss of a major client which generates more than 30% of revenues of the borrower. This trigger is a rebuttable presumption 6 Delays in payments to the government (taxes and other mandatory payments) or to employees (wages, salaries, etc.)

10
7
Collateral value decreases more than 30% since initial recognition and the main source of repayment of the loan is the sale of the collateral. Relevant to asset-based lending such as acquisition finance, trade finance, in some cases project finance and etc 8 Bond trading (temporarily) suspended due to rumors or facts related to financial difficulties Consumer loans 9 Permanent and significant deterioration in the level of income or solvency of the borrower (co-borrower) which expressed in PTI higher than 60%. Business/Consumer loans 10 Misuse of loan purpose. E.g., borrower took a loan for financing working capital and instead buying OTC debt notes Business/consumer/Mortgage loans 11 Lack of employment or commercial activity. This trigger is a rebuttable presumption 12 The presence of factors that have caused material damage to the borrower (co-borrower) like injury, illness or death 13 For all segments except transactions with financial institutions and issuers of securities, the debt is more than 90 calendar days overdue as of the reporting date 14 Default or forced restructuring, as further detailed in Para
4.7.4 of these Guidelines
15
Sale of part or full amount of the loan with discount of more than 5% 16 The borrower requested emergency financing. 17 There are no possible refinancing options for the borrower. May indicate negative risk-adjusted return and should be a stage 3 criteria 18 More than one restructuring in the last 12 months. This trigger is a rebuttable presumption 19 There is a high probability that the borrower (co-borrower) will not repay his obligations to the bank Other 20 For transactions with financial institutions and issuers of securities, the debt is more than 7 business days overdue as of the reporting date
4.7.4. Distressed restructuring refers to a restructuring with either a significant decrease
in the loan’s net present value (NPV) (by more than 2.5%) or restructuring occurring when client is facing a significant deterioration of credit quality:
4.7.4.1. The NPV of the loan after restructuring is discounted using current market
interest rates. This prevents a reduction in NPV caused solely by changes in market interest rates from being treated as distressed restructuring.

11
4.7.4.2. The general causes of a reduction in the NPV of a loan include:
4.7.4.2.1. reduction in the interest rate caused by changes in the refinancing rate or
yields of government bonds.
4.7.4.2.2. forgiveness of principal or interest.
4.7.4.2.3. rescheduling of payments without accrual of interest.
4.7.4.2.4. other changes, (flipping into different FX, etc).
4.7.4.3. Client is considered to face a significant deterioration of credit quality at the
time of restructuring if any of the following conditions were present in 12 months before restructuring:
4.7.4.3.1. 10+ days of overdue.
4.7.4.3.2. expectation that the borrower will miss future payments.
4.7.4.3.3. projected cash flows will not be sufficient to cover the debt in full.
4.7.4.3.4. more than 1 notch decrease of internal/external ratings.
4.7.4.3.5. delisting of bonds.
4.7.4.3.6. any SICR criteria.
4.7.5. Concealing default / distressed restructurings.
4.7.5.1. Concealment of distressed restructuring through artificial loan re-origination is
strictly prohibited. This practice typically occurs when loans representing at least 20% of the borrower’s current debt are closed or repaid, and a new loan is granted to the same borrower, with a new identification number and an amount within ±30% of the closed loan, within 30 calendar days.
4.7.5.2. Any such restructuring should be flagged, subject to enhanced review, and—
unless robust justification is documented—classified as stage 3.
4.8. Provisioning.
4.8.1. The following methodological principles of ECL modeling are fundamental to any
models, approaches and techniques used in ECL calculations:
4.8.1.1. Credit loss is the difference between all contractual cash flows that are due to a
bank in accordance with the contract and all the cash flows that the bank expects to receive, discounted at the original effective interest rate.
4.8.1.2. The bank should measure ECL on a financial instrument as follows expected
credit losses of a financial instrument in a way that reflects:
4.8.1.2.1. an unbiased and probability-weighted amount that is determined by
evaluating a range of possible outcomes, even if the possibility of a credit loss occurring is very low.
4.8.1.2.2. the time value of money.
4.8.1.2.3. reasonable and supportable information that is available without undue cost
or effort at the reporting date about past events, current conditions and forecasts of future economic conditions.

12
4.8.2. Indivudal ECL calculation:
4.8.2.1. ECL is calculated on an individual basis under Para 4.8.1.1 of these Guidelines.
4.8.2.2. The individual ECL calculation is based on the discounted cash flow method. In
this approach, forecasted recoverable cash flows (from borrower’s operations, collateral value, or other sources) are estimated through an analysis of the borrower’s financial position, business plan, and market conditions under various scenarios (including at least one default scenario). These projected cash flows are then compared with the contractual cash flows of the instrument, and the difference is discounted at the original effective interest rate, and determined using the following formula. ECL = NPV of contractual cash flows − average weighed NPV of expected cash flows
4.8.2.3. Individual ECL calculation should be performed for significant clients
(principal and interest ≥ 5% of total IFRS Equity) with deteriorated credit quality (stage 2 and stage 3).
4.8.3. Collective ECL calculation:
4.8.3.1. Collective ECL calculation for Stage 1 and Stage 2 exposures is performed with
the following formula:
ECL = ∑mPDt
T t=1
× LGD × EAD𝑡 × DF𝑡 = ∑cPDt
T t=1
× LGD × CF𝑡 × DF𝑡 ≈ PD ∗ LGD ∗ EAD
4.8.3.1.1. The simplified ECL formula is calculated under Para 4.11.2.4 herein.
4.8.3.1.2. The underlying formula used, which incorporates the structure of the asset’s
cash flows, is as follows:
∑mPDt
T t=1
× LGD × EADt × DF = ∑cPDt
T t=1
× LGD × CF × DF
Where, mPDt– the marginal probability of default in period t, assuming that no default occurred in period t–1.
4.8.3.1.3. The marginal probability of default is determined with the below formula:
mPDt = cPDt − cPDt−1
Where, cPDt – the probability of default accumulated over [0; t] period.
4.8.3.2. Stage 2 ECL calculation:
4.8.3.2.1. For stage 2 ECL calculations either mPD or cPD should be determined based
on the time horizon corresponding to the maturity of each cash flow within the loan. Alternatively, a simplified calculation can be based on the maturity of the entire loan, in other

13 words, if the loan is expected to be repaid 10 years after the reporting date, the institution should have a robust estimate of the 10-year PD. Possible approaches are elaborated below:
4.8.3.2.2. A simplified approach for calculating cPD for the period longer than 1 year
is determined using the following formula:
cPDt = 1 −∏(1 − PD) t i=1
Where, PD – 12-month PD.
4.8.3.2.3. The proposed approach for calculating the cPD for maturities exceeding one
year involves the combined use of the simplified formula and cumulative default rate (hereinafter – DR) curves, as determined by the following formula:
cPDt = 1 −∏(1 − PDi
) t i=1
Where, PD_t – The PD over a 12-month period for each subsequent year in which the financial instrument is outstanding, measured from the reporting date. This probability is determined as follows:
PDi = {
(PIT)PD = 1st year
(PIT) PD = 2nd year
(TTC)PD > 2 years
Where, PDi - derived from existing DR curves based on the risk segment for which the ECL calculation is performed.
4.8.3.2.4. For contracts with a partial utilisation year, the cumulative probability of
default is determined using the following formula:
cPDT = 1 −∏(1 − PDi
)
T′ t=1
∙ (1 − PDT
)
Where, T - the remaining utilisation years (for example, if the remaining term of the contract is 30 months, T = 2.5); T′ – the integer part of T (e.g., if the remaining term of the contract is 30 months, T′ = 2).
4.8.3.2.4.1. If the final utilisation year of the financial instrument comprises
partial months, T shall be rounded to the nearest whole number.
4.8.3.2.4.2. The probability of default for the final utilisation year (which may be
partial) is determined using the following formula:

14
PDT = 1 − (1 − PDt
)
T−T′
4.8.3.2.5. An alternative proposed approach for the calculation of cPDt
is the Markov chain or Markov process. This approach may be used to calculate migration matrices based on the migration of rating grades or days past due within the relevant risk segment for the purpose of ECL estimation. Multiplying the migration matrices allows the institution to obtain forecasts for the required number of years. Additionally, simple linear interpolation may be used to derive cPDₜ at a specific point in time.
4.8.3.2.6. For all remaining assets not subject to individual assessment, ECL is estimated
on a collective basis.
4.8.3.3. The calculation of ECL on a collective basis for Stage 3 exposures is
determined using the following formula:
ECL = EADfact ∗ LGDID
Where, EADfact – The carrying amount of the credit exposure as at the date of allowance calculation. This amount includes the principal and past-due principal, accrued interest, past￾due interest, and penalty charges. LGDID –the LGD applied in default represents a separate model and is not the same as the LGD used for Stage 1 and Stage 2 in ECL calculations. LGD in default is primarily driven by the time spent in default and also incorporates other risk factors, including the LTV ratio, type of collateral, seniority of exposure, industry sector, the existence of recoveries on defaulted amounts, and other factors.
4.8.4. Banks are required to develop comprehensive methodologies for measuring ECL.
The calculation of ECL incorporates key risk components, including PD, LGD and EAD.
4.9. Probability of Default (PD) represents the likelihood that a borrower will enter
default over a defined time horizon, mainly 1 year.
4.9.1. Main components of PD model are:
4.9.1.1. Scoring mechanism:
4.9.1.1.1. The scoring mechanism assigns a numerical score to each borrower or facility
that reflects the likelihood of default over a specified horizon (typically 12 months for IFRS 9).
4.9.1.1.2. The scoring mechanism is an expert (expert-based score cards, etc.) or statistical
model (logistic regression, gradient boosting, etc.).
4.9.1.1.3. The scoring mechanism used in the PD model shall be developed at the level of
homogeneous risk segment (e.g., product type for lending or individual, segments for lending to legal entities).
4.9.1.1.4. For the development of the scoring mechanism the bank should consider the
usage of various risk drivers.
4.9.1.1.5. Not exhaustive list of risk drivers for corporate portfolios:

15
4.9.1.1.5.1. Financial information: Leverage, interest coverage, liquidity, profitability,
company size, etc.
4.9.1.1.5.2. Non-financial information: Industry, region, age of company, etc.
4.9.1.1.5.3. Type and value of collateral.
4.9.1.1.5.4. Externally available information (legal cases, credit bureau, etc.).
4.9.1.1.6. In case if corporate risk segment does not have enough statistics for model
development and calculation of calibration level, the external data could be considered to use.
4.9.1.1.7. For retail portfolios, a non-exhaustive list of risk drivers includes:
4.9.1.1.7.1. behavioural factors: missed or overdue payments (DPD), restructuring, credit
utilization, etc.
4.9.1.1.7.2. other factors: application factorss, LTV, DTI, and related measures.
4.9.1.1.7.3. externally available information (legal cases, credit bureau data, etc.).
4.9.1.1.8. for retail and SME segments, models used at loan origination as well as
behavioural models (for monitoring purposes) may be readily integrated into ECL calculations.
4.9.1.1.9. indicators based on days past due may be used as risk drivers within the
scoring mechanism or may act as separate early-warning signals that result in a deterioration of the PD rating and, consequently, an increase in PD.
4.9.1.2. The DR represents the frequency of borrowers in default calculated.
4.9.1.3. DR is a key metric used in the calibration of PD estimates and is determined
using the following formula:
DR =
Number of borrowers defaulting within 12 months from snapshot t Number of non − defaulted borrowers at snapshot t 4.9.1.3.1.Usage of EAD-weighted DR is not recommended due to the bias toward larger exposures:
4.9.1.3.1.1. averaging: to receive the calibration level for the PD model, DR points should
be averaged. The range for averaging to keep the PD model PIT compliant is 12 – 24 months.
4.9.1.3.1.2. snapshot dates: banks are free to decide and substantiate which snapshot date
frequency to use for DR averaging for calculation of PD model calibration level: month, quarter, half a year, 1 year. Any other snapshot frequency is not recommended.
4.9.1.3.1.3. treatment of Repeated Defaults: if the borrower defaults, then rehabilitates
(as per Item 4.5) and later defaults again during the year the second default is included in the DR numerator.
4.9.1.3.2. defaults may not be excluded from the statistics for DR calculation and further
calibration of PD models.
4.9.1.4. the numerical score calculated by scoring mechanism should be transformed
into the PD estimate for ECL calculations. Such transformation is facilitated by calibration.

16
4.9.1.4.1. in broader terms calibration is the mathematical task of fitting the coefficients
of calibration curve to existing level of defaults (calibration level) and discriminatory power of scoring mechanism.
4.10. LGD is loss incurred in the event of default, expressed as a percentage of the EAD.
4.10.1. In case banks have a lack of data to develop LGD model, the following
simplified approach can be used:
LGD = 1 − value of the collateral ∗ discount
EAD
4.10.2. Recommended approach for LGD estimation:
4.10.2.1. probability of credit quality recovery. For every default case, the bank may
consider two possible outcomes:
4.10.2.1.1. cure without economic loss (e.g., non DPD criteria of stage 3 were activated
but then borrower meets it’s contractual obligations in full and in time).
4.10.2.1.2. default outcomes accompanied with economic loss (e.g., discounted sale,
collateral repossession, write-off of fines and accrued interest, etc.).
4.10.2.2. the bank is recommended to predict the probability of the outcome for every
potential default case, taking into account available data and modelling capabilities. These probabilities are used to calculate weighted averaging of expected loss in the event of default (Loss Given Loss, LGL), which is then used to derive the LGD value applied in ECL calculations.
4.10.2.3. The central component in LGD modeling is LGL, it can be used both as
calibration level and a target variable for LGD model development, depending on the architecture of the model. № Type of collateral Discount 1 Residential and/or commercial real estate, including land sites, located in Baku 50% 2

• Non-residential and non-commercial real estate.
• Residential and/or commercial real estate, including land sites, located
  outside Baku
  70%
  3 Vehicles 90%
  4 Equipment, inventory, products for sale 90% 5 Highly liquid securities (as in LCR) 10% 6 Cash collateral 0%

17
LGL = 1 − Total recovery% =
Balance at default − (Recovery − Costs of collection) Balance at default Where, costs of collection – legal, administrative, and other expenses incurred during recovery.
4.10.2.4. Important considerations for recovery calculations:
4.10.2.4.1. restructurings and write-offs are not considered as sources of recovery.
4.10.2.4.2. full recovery happens when the rehabilitation criteria are met.
4.10.2.4.3. in scenario of the sale of the loan, recovery should account the discount on
sale. For instance, if the balance at default is equal to 100, loan was sold for 40, then undiscounted recovery at the date of the sale is 40%.
4.10.2.4.4. in scenario of the repossession of the collateral, the amount and date of
recovery corresponds to the sale price and the date of the sale of the collateral not with the accounting transaction of taking the collateral on the institution’s balance sheet.
4.10.2.4.5. the CFs of recovery are discounted to the date of default based on the effective
interest rate of the specified contract.
4.10.2.5. One of the key components of LGD is the value of collateral. It is important to
ensure that there is proper governance around collateral valuation:
4.10.2.5.1. collateral should be sufficiently liquid—capable of being sold within a
reasonable timeframe and at predictable value. Illiquid assets (e.g., specialized machinery, company shares in private firms) should be excluded or heavily discounted.
4.10.2.5.2. collateral values should be updated regularly:
4.10.2.5.2.1. real estate: not less than every 3 years or more frequently in adverse
conditions.
4.10.2.5.2.2. marketable securities: mark-to-market daily or monthly.
4.10.2.5.3. the bank should have the right to take possession and liquidate the asset in
the event of default.
4.11. EAD modelling:
4.11.1. EAD represents the estimated outstanding amount of financial asset at the
time of default.
EAD = on balance amount ∗ expected repayments + off balance amount ∗ CCF Where, on balance amount – a sum of principal and accrued interest. expected repayments – a model estimate of repayments of current on balance amount before default; can be used by banks that have robust repayment models, internally validated, and used in the Asset and Liability Management (hereinafter – ALM) processes. off balance amount - a sum of credit related commitments of the institution such as undrawn credit line, letter of credit, guarantee, etc.

18
CCF - a model estimate of expected drawdown of off-balance sheet exposure (at reporting date) by the time of default.
4.11.2. scheduled payments and repayment modeling:
4.11.2.1. banks may use the following payment models for collective ECL calculations
in Stage 1, based on the approach described in Para 4.8.3 of these Guidelines:
ECL = ∑mPDt
T t=1
× LGD × EAD𝑡 × DF𝑡 = ∑cPDt
T t=1
× LGD × CF𝑡 × DF𝑡
4.11.2.2. Payment model for use in ECL calculations should:
4.11.2.2.1. be used by the bank in ALM processes.
4.11.2.2.2. successfully pass the independent validation which deem such model
applicable for intended use.
4.11.2.3. Payment models are not used for clients in Stage 2 and Stage 3, given that such
models are developed for portfolio with standard credit quality.
4.11.2.4. When the bank applies the simplified approach for collective ECL calculation
(ECL = PD * LGD * EAD), it does not use any factor that would reduce the current balance for EAD calculation (calculated based on expected payments or contractual schedule).
4.11.2.5. Simplified approach to CCF assessment:
4.11.2.5.1. If the bank does not have enough data for CCF modeling, the factors shown
in the table below can be used.
4.11.2.5.2. If a commitment falls into multiple categories, the lowest CCF may apply.
№ Type of credit related commitment CCF (%)
1
Direct credit substitutes, including guarantees of total debt obligations (including letters of credit that serve as collateral for loan and debt securities transactions). 100% 2 Sale and repurchase agreements and asset sales with recourse where the credit risk remains with the bank. 100% 3 Off-balance sheet items that are credit substitutes and not explicitly included in any other category. 100% 4 Note issuance facilities and revolving underwriting facilities regardless of the maturity of the underlying facility. 50% 5 Transaction-related contingent items (e.g. performance bonds, bid bonds, warranties and standby letters of credit related to particular transactions). 50% 6 Commitments, regardless of the maturity of the underlying facility, unless they qualify for a lower CCF. 40%

19
4.12. Application of macroeconomic indicators:
4.12.1. the impact of macroeconomic factors is taken into account at the level of the PD
component within risk parameters. The bank evaluates the impact of macroeconomic data on the credit portfolio, which consists of customer segments with similar risk characteristics predetermined by macroeconomic data. The forecasted data is considered as follows:
4.12.1.1. the DR or the PD component is selected as the target variable (hereinafter y).
4.12.1.2. the explanatory variable (hereinafter x) is selected from macroeconomic
indicators based on the criteria defined in Paras 4.12.2 and 4.12.3 of these Guidelines.
4.12.2. Only factors with forecast indicators published in publicly available official
external sources are included in the list, as the analysis requires the calculation of forecast indicators for the default level.
4.12.3. Factors published more than once a year are selected for inclusion in the full list,
as the modeling process requires a sufficient number of observations:
4.12.3.1. The functional dependence between y and x is determined through the
assessment of regression between PD component indicators of various frequencies (monthly or quarterly) and macro factors. Various transformations of these variables can be used to find the optimal dependence between y and x.
4.12.3.2. A series of indicators must be tested to verify the quality of the model. These
indicators include, but are not limited to:
4.12.3.2.1. Economically correct signs of regression parameters (e.g., an increase in GDP
leads to a decrease in PD).
4.12.3.2.2. The p-value of the model’s F-test does not exceed 1%.
4.12.3.2.3. The p-value of t-tests for all model parameters does not exceed 1%.
4.12.3.2.4. The R^2 indicator exceeds 30%.
4.12.4. After completing all analyses and tests, banks should compile the final regression
equation.
4.13. Governance:
4.13.1. IFRS 9 ECL estimates and its components (PD, LGD, EAD, staging, etc.) are
integrated into risk and business processes, where it is applicable and reasonable from a 7 Short-term letters of credit (with a maturity of below one year) arising from the circulation of goods, issued and confirmed by banks. 20% 8 Commitments that grant the bank a unilateral right not to fulfill the obligation in full, or that are automatically cancelled upon deterioration of the borrower’s creditworthiness . The domestic supervisory authority evaluates various factors that may effectively constrain the banks’ ability to exercise these non-fulfillment rights and, where appropriate, considers applying a higher CCF to certain commitments. 10%

20 business standpoint, to ensure consistency between risk measurement, reporting, and strategic decision-making. These processes include, but are not limited to:
4.13.1.1. Loan pricing, as a cost of risk component.
4.13.1.2. Portfolio and risk monitoring - tracking deterioration trends, risk migrations,
and portfolio quality assessments.
4.13.1.3. Risk appetite monitoring and limit setting. ECL is considered as one of the
elements of risk appetite framework
4.13.1.4. Stress testing and scenario analysis. Forward-looking information engine is
considered as a part or core mechanic of stress testing and sensitivity analysis to simulate capital and P&L outcomes under adverse conditions.
4.13.1.5. Strategic and business planning – ECL forecasts (as of the reporting date)
should be considered as an important input into strategic and business planning.
4.13.1.6. External and internal reporting. Consistent use of IFRS 9 metrics should be
ensured for external reporting (IFRS reporting, investors presentations, central bank reporting, etc.) and internal reporting used for portfolio steering and KPI setting/measurement.
4.13.2. Model governance and validation: a robust governance framework should be
established for all IFRS 9 models (PD, LGD, EAD) and ECL calculations:
4.13.2.1. Documentation standards – Models should be well-documented with clearly
defined segmentation, data sources, risk drivers, and methodologies, and be reproducible.
4.13.2.2. Validation process – PD, LGD, and EAD models should undergo independent
validation before deployment and be subject for regular model validation.
4.13.2.3. Statistical validation of PD, LGD, and EAD includes the following components:
4.13.2.3.1. Predictive power – checking whether model forecast accuracy exceeds that of
random guessing (for PD and LGD, using ROC AUC tests and relevant methods).
4.13.2.3.2. Calibration – assessing the magnitude of model estimates is in line with
observable data (using Binomial tests for PD models and the LCR method for LGD models).
4.13.2.3.3. Stability – check that model shows stable performance over time (for instance
PSI or evolution of model metrics over time).
4.13.2.4. ECL model backtesting – retrospective assessment of the ECL model against
realized losses.
4.13.3. Oversight by relevant business unit heads and the Management Board: relevant
business unit heads should maintain oversight over the ECL process, ensuring that it operates effectively, consistently, and in alignment with the institution’s risk appetite, accounting policies, and regulatory expectations. This includes:
4.13.3.1. ensuring that the ECL framework — including data sourcing, staging, model
outputs, forward looking information and final provisioning — is governed by clear policies and regularly reviewed for relevance and robustness.

21
4.13.3.2. establishing effective control mechanisms over the application of manual
adjustments and management overlays, ensuring traceability, and respond appropriately to internal risk indicators and external factor changes.
4.13.3.3. ensuring alignment between risk, finance, and business functions in terms of
data ownership, staging, ECL calculation and usage of ECL components in business processes.
4.13.3.4. monitoring key ECL indicators (e.g., stage migration rates, segment-level ECL
trends) and challenging significant deviations or unexplained volatility.
4.13.3.5. Ensuring that ECL figures are communicated clearly to the Management Board
and external parties where required, with an appropriate level of detail.
5. Submission of reports
5.1. Reports specified in Annex 1 to these Guidelines are prepared electronically and
submitted to the Central Bank via the Electronic Services Portal. Reports are approved with the enhanced electronic signature of the Chairperson of the Bank’s Management Board or the Chief Financial Officer (for a local branch of a foreign bank, the branch head or chief accountant (the person performing these functions). Reports should be submitted within 10 (ten) business days following the end of the reporting period as follows:
5.1.1. monthly for systemically important banks.
5.1.2. quarterly for other banks with total assets of AZN 1 (one) billion or more.
5.1.3. semi-annually for banks with total assets of AZN 1 (one) billion or less.
5.2. Reporting format and content requirements:
5.2.1. Reports are prepared as of the last business day of the reporting period (reporting
date), including all transactions up to that date. The ‘Reporting Date’ field indicates the last business day of the period. Items for branches, departments, and representative offices are presented in consolidated form for both on-balance and off-balance sheet obligations;
5.2.2. All figures in reports are presented in thousands of AZN. Amounts in foreign
currency are converted to AZN at an official exchange rate determined by the Central Bank on the reporting date.
5.2.3. Amounts up to AZN 500 are rounded to zero; amounts AZN 500 and above are
rounded to the nearest AZN 1,000. Rounding is applied consistently to ensure alignment between totals and individual schedules in reports.
5.2.4. If any indicator equals zero, the report explicitly shows “0” in the corresponding
field.
5.2.5. No field in the report is left blank. Each field contains either a figure or a zero.

22
Annex 1 to the ‘Methodological Guidelines on the
implementation of International Financial Reporting Standard 9 for asset classification and loan loss provisioning’
TABLE U1
Title page
CENTRAL BANK OF THE REPUBLIC OF AZERBAIJAN
ADDITIONAL REPORT on the implementation of IFRS 9
TABLE Ü 1 - TITLE PAGE
Bank name: License number:
Bank address:
Reporting date:
Completion date:
This report is submitted for the purpose of clarifying the information contained in the previously submitted reports:
Yes
SIGNATURE AND CERTIFICATE

23
I confirm that this report has been prepared in accordance with the instructions of the Central Bank of the Republic of Azerbaijan, and that the information presented in the report regarding the status and results of the bank’s operations is completely accurate. Member of the Management Board authorized to use an electronic signature, (first, middle, last names) Electronic signature right valid to (final date):
Reporting unit:
Reporting person:
Name (first, middle, last names):
Position:
E-mail address:
Phone number:
Fax:
TABLE A1
A.
Loan portfolio at amortized cost (all loans, including foreign-currency-denominated loans) (in thousand manats) Types of loans Total amount Expected credit losses Total Stage 1 Stage 2 Stage 3 POCI Total Stage 1 Stage 2 Stage 3 POCI 1 2 3 4 5 6 7 8 9 10 11

24
Loans to customers, total
A
0

1. Business loans A1
   0
   1.1 Industry A1_1
   0
   0
   1.2 Agriculture A1_2
   0
   0
   1.3 Construction A1_3
   0
   0
   1.4 Transport A1_4
   0
   0
   1.5 Information and
   communication
   A1_5
   0
   0
   1.6 Loans to commercial
   enterprises
   A1_6
   0
   0
   1.7 Other non
   -production and service sectors
   A1_7
   0
   0
2. Consumer loans A2
   0
   2.1 Residential property
   renovation
   A2_1
   0
   0
   2.2 Car purchase A2_2
   0
   0
   2.3 Purchase of household
   appliances
   A2_3
   0
   0
   2.4 Credit cards A2_4
   0
   0
   2.5 Other A2_5
   0
   0
3. Mortgage loans A3
   0
   0
4. Other loans A4
   0
   0

25
B. Loan portfolio at amortized cost
(in foreign currency)
(in thousand manats)
Types of loans
Total amount Expected credit losses
Total Stage 1 Stage 2 Stage 3 POCI Total Stage 1 Stage 2 Stage 3 POCI 1 2 3 4 5 6 7 8 9 10 11 Loans to customers, total B 0

1. Business loans B1 0
   1.1 Industry B1_1 0 0
   1.2 Agriculture B1_2 0 0
   1.3 Construction B1_3 0 0
   1.4 Transport B1_4 0 0
   1.5 Information and communication B1_5 0 0
   1.6 Loans to commercial enterprises B1_6 0 0
   1.7 Other non-production and
   service sectors
   B1_7 0 0
2. Consumer loans B2 0
   2.1 Residential property renovation B2_1 0 0
   2.2 Car purchase B2_2 0 0
   2.3 Purchase of household
   appliances
   B2_3 0 0
   2.4 Credit cards B2_4 0 0
   2.5 Other B2_5 0 0
3. Mortgage loans B3 0 0
4. Other loans B4 0 0

26
C. Loan portfolio at fair value through other comprehensive income
(all loans, including foreign currency denominated loans) (in thousand manats) Types of loans Total amount Expected credit losses Total Stage 1 Stage 2 Stage 3 POCI Total Stage 1 Stage 2 Stage 3 POCI 1 2 3 4 5 6 7 8 9 10 11 Loans to customers, total C 0

1. Business loans C1 0
   1.1 Industry C1_1 0 0
   1.2 Agriculture C1_2 0 0
   1.3 Construction C1_3 0 0
   1.4 Transport C1_4 0 0
   1.5 Information and communication C1_5 0 0
   1.6 Loans to commercial enterprises C1_6 0 0
   1.7 Other non-production and
   service sectors
   C1_7 0 0
2. Consumer loans C2 0
   2.1 Residential property renovation C2_1 0 0
   2.2 Car purchase C2_2 0 0
   2.3 Purchase of household
   appliances
   C2_3 0 0
   2.4 Credit cards C2_4 0 0
   2.5 Other C2_5 0 0

27
3. Mortgage loans C3 0 0
4. Other loans C4 0 0
D. Loan portfolio at fair value through other comprehensive income (in foreign currency) (in thousand manats) Types of loans Total amount Expected credit losses Total Stage 1 Stage 2 Stage 3 POCI Total Stage 1 Stage 2 Stage 3 POCI 1 2 3 4 5 6 7 8 9 10 11 Loans to customers, total D 0

1. Business loans D1 0
   1.1 Industry D1_1 0 0
   1.2 Agriculture D1_2 0 0
   1.3 Construction D1_3 0 0
   1.4 Transport D1_4 0 0
   1.5 Information and communication D1_5 0 0
   1.6 Loans to commercial enterprises D1_6 0 0
   1.7 Other non-production and
   service sectors
   D1_7 0 0
2. Consumer loans D2 0
   2.1 Residential property renovation D2_1 0 0
   2.2 Car purchase D2_2 0 0
   2.3 Purchase of household
   appliances
   D2_3 0 0
   2.4 Credit cards D2_4 0 0

28
2.5 Other D2_5 0 0
3. Mortgage loans D3 0 0
4. Other loans D4 0 0
E. Loan portfolio at fair value through profit or loss (all loans, including foreign currency denominated loans) (in thousand manats) Types of loans 1 2 Loans to customers, total E 0

1. Business loans E1 0
   1.1 Industry E1_1
   1.2 Agriculture E1_2
   1.3 Construction E1_3
   1.4 Transport E1_4
   1.5 Information and communication E1_5
   1.6 Loans to commercial enterprises E1_6
   1.7 Other non-production and service sectors E1_7
2. Consumer loans E2 0
   2.1 Residential property renovation E2_1
   2.2 Car purchase E2_2
   2.3 Purchase of household appliances E2_3
   2.4 Credit cards E2_4
   2.5 Other E2_5
3. Mortgage loans E3
4. Other loans E4

29
F. Loan portfolio at fair value through profit or loss (in foreign currency) (in thousand manats) Types of loans 1 2 Loans to customers, total F 0

1. Business loans F1 0
   1.1 Industry F1_1
   1.2 Agriculture F1_2
   1.3 Construction F1_3
   1.4 Transport F1_4
   1.5 Information and communication F1_5
   1.6 Loans to commercial enterprises F1_6
   1.7 Other non-production and service sectors F1_7
2. Consumer loans F2 0
   2.1 Residential property renovation F2_1
   2.2 Car purchase F2_2
   2.3 Purchase of household appliances F2_3
   2.4 Credit cards F2_4
   2.5 Other F2_5
3. Mortgage loans F3
4. Other loans F4

30
TABLE A1.1
A1. Regarding changes in the loan portfolio balance across stages (all loans, including foreign currency denominated loans) (in thousand manats) Types of loans Total amount Total Stage 1 Stage 2 Stage 3 POCI Restructured portfolio (from Stage 2) 1 2 3 4 5 6 7 Loans to customers, total A1 0

1. Business loans A1.1
   Balance at the beginning of the year A1.1_1 0 Issued during the period A1.1_2 0 Repaid during the period A1.1_3 0 Movement to Stage 1 A1.1_4 0 0 Movement to Stage 2 A1.1_5 0 0 Movement to Stage 3 A1.1_6 0 0 Loans recovered from previously written-off amounts A1.1_7 0 Loans written off A1.1_8 0 Closing balance of the year A1.1_9 0
2. Consumer loans A1.2
   Balance at the beginning of the year A1.2_1 0 Issued during the period A1.2_2 0 Repaid during the period A1.2_3 0

31
Movement to Stage 1 A1.2_4
0
0
Movement to Stage 2 A1.2_5
0
0
Movement to Stage 3 A1.2_6
0
0
Loans recovered from previously written
-off amounts
A1.2_7
0
Loans written off A1.2_8
0
Closing balance of the year A1.2_9
0
3. Mortgage loans A1.3
Balance at the beginning of the year A1.3_1
0
Issued during the period A1.3_2
0
Repaid during the period A1.3_3
0
Movement to Stage 1 A1.3_4
0
0
Movement to Stage 2 A1.3_5
0
0
Movement to Stage 3 A1.3_6
0
0
Loans recovered from previously written
-off amounts
A1.3_7
0
Loans written off A1.3_8
0
Closing balance of the year A1.3_9
0
4. Other loans A1.4
Balance at the beginning of the year A1.4_1
0
Issued during the period A1.4_2
0
Repaid during the period A1.4_3
0
Movement to Stage 1 A1.4_4
0
0
Movement to Stage 2 A1.4_5
0
0
Movement to Stage 3 A1.4_6
0
0
Loans recovered from previously written
-off amounts
A1.4_7
0

32
Loans written off A1.4_8 0
Closing balance of the year A1.4_9 0
A2. Expected credit losses across stages across stages (all loans, including foreign currency denominated loans) (in thousand manats) Types of loans Expected credit losses Total Stage 1 Stage 2 Stage 3 POCI Restructured portfolio (from Stage 2) 1 2 3 4 5 6 7 Loans to customers, total A2 0

1. Business loans A2.1
   Balance at the beginning of the year A2.1_1 0 Closing balance of the year A2.1_2 0
2. Consumer loans A2.2
   Balance at the beginning of the year A2.2_1 0 Closing balance of the year A2.2_2 0
3. Mortgage loans A2.3
   Balance at the beginning of the year A2.3_1 0 Closing balance of the year A2.3_2 0
4. Other loans A2.4
   Balance at the beginning of the year A2.4_1 0 Closing balance of the year A2.4_2 0

33
Note: In movements to Stage 1, Stage 2, and Stage 3, the relevant fields should be populated with a negative value When filling in the ‘Loans granted during the period’ rows, loans granted during the reporting period should be recorded under the column corresponding to their stage as of the reporting date. Restructured portfolio due to financial difficulty
TABLE A1.2
A. Loan portfolio breakdown by stages and days past due (all loans, including foreign currency denominated loans) (in thousand manats) Types of loans Total Days past due Current 1-30 days 31-90 days over 90 days 1 2 3 4 5 6 Loans to customers, total A1 0 Business loans A1.1 0 Stage 1 A1.1_1 0 Stage 2 A1.1_2 0 Stage 3 A1.1_3 0 Consumer loans A1.2 0 Stage 1 A1.2_1 0 Stage 2 A1.2_2 0 Stage 3 A1.2_3 0 Mortgage loans A1.3 0 Stage 1 A1.3_1 0 Stage 2 A1.3_2 0

34
Stage 3 A1.3_3 0
Other loans A1.4 0
Stage 1 A1.4_1 0
Stage 2 A1.4_2 0
Stage 3 A1.4_3 0
B. Loan portfolio breakdown by stages and days past due (in foreign currency) (in thousand manats) Types of loans Total Days past due Cari 1-30 days 31-90 days over 90 days 1 2 3 4 5 6 Loans to customers, total B1 0 Business loans B1.1 0 Stage 1 B1.1_1 0 Stage 2 B1.1_2 0 Stage 3 B1.1_3 0 Consumer loans B1.2 0 Stage 1 B1.2_1 0 Stage 2 B1.2_2 0 Stage 3 B1.2_3 0 Mortgage loans B1.3 0 Stage 1 B1.3_1 0

35
Stage 2 B1.3_2 0
Stage 3 B1.3_3 0
Other loans B1.4 0
Stage 1 B1.4_1 0
Stage 2 B1.4_2 0
Stage 3 B1.4_3 0

36
TABLE A2
A. Securities portfolio
(all securities, including foreign currency denominated securities) (in thousand manats) Balance items Total amount Expected credit losses Total Stage 1 Stage 2 Stage 3 POCI Total Stage 1 Stage 2 Stage 3 POCI 1 2 3 4 5 6 7 8 9 10 11 Securities portfolio, total A 0

1. Securities measured at amortized
   cost
   A1 0
   1.1 Securities of central banks A1_1 0 0
   1.2 Government treasury bills and
   other securities of central government authorities A1_2 0 0
   1.3 Investments in term/preferred
   shares
   A1_3 0
   1.3.1 Financial sector A1_3_1 0 0
   1.3.2 Non-financial sector A1_3_2 0 0
   1.4 Other debt liabilities A1_4 0
   1.4.1 Financial sector A1_4_1 0 0
   1.4.2 Non-financial sector A1_4_2 0 0
2. Securities measured at fair value
   through other comprehensive income
   A2 0
   1.1 Securities of central banks A2_1 0 0

37
1.2 Government treasury bills and
other securities of central government authorities A2_2 0 0
1.3 Stocks A2_3
0
1.3.1 Financial sector A2_3_1
0
0
1.3.2 Non
-financial sector A2_3_2
0
0
1.4 Other debt liabilities A2_4
0
1.4.1 Financial sector A2_4_1
0
0
1.4.2 Non
-financial sector A2_4_2
0
0
3. Securities measured at fair value
through profit or loss
A3
0
1.1 Securities of central banks A3_1
1.2 Government treasury bills and
other securities of central government authorities A3_2
1.3 Stocks A3_3
0
1.3.1 Financial sector A3_3_1
1.3.2 Non
-financial sector A3_3_2
1.4 Other debt liabilities A3_4
0
1.4.1 Financial sector A3_4_1
1.4.2 Non
-financial sector A3_4_2

38
B. Securities portfolio
(in foreign currency)
(in thousand manats)
Balance items
Total amount Expected credit losses
Total Stage 1 Stage 2 Stage 3 POCI Total Stage 1 Stage 2 Stage 3 POCI 1 2 3 4 5 6 7 8 9 10 11 Securities portfolio, total B 0

1. Securities measured at amortized
   cost
   B1 0
   1.1 Securities of central banks B1_1 0 0
   1.2 Government treasury bills and
   other securities of central government authorities B1_2 0 0
   1.3 Investments in term/preferred
   shares
   B1_3 0
   1.3.1 Financial sector B1_3_1 0 0
   1.3.2 Non-financial sector B1_3_2 0 0
   1.4 Other debt liabilities B1_4 0
   1.4.1 Financial sector B1_4_1 0 0
   1.4.2 Non-financial sector B1_4_2 0 0
2. Securities measured at fair value
   through other comprehensive income
   B2 0
   1.1 Securities of central banks B2_1 0 0

39
1.2 Government treasury bills and
other securities of central government authorities B2_2 0 0
1.3 Stocks B2_3
0
1.3.1 Financial sector B2_3_1
0
0
1.3.2 Non
-financial sector B2_3_2
0
0
1.4 Other debt liabilities B2_4
0
1.4.1 Financial sector B2_4_1
0
0
1.4.2 Non
-financial sector B2_4_2
0
0
3. Securities measured at fair value
through profit or loss
B3
0
1.1 Securities of central banks B3_1
1.2 Government treasury bills and
other securities of central government authorities B3_2
1.3 Stocks B3_3
0
1.3.1 Financial sector B3_3_1
1.3.2 Non
-financial sector B3_3_2
1.4 Other debt liabilities B3_4
0
1.4.1 Financial sector B3_4_1
1.4.2 Non
-financial sector B3_4_2

40
TABLE A2.1
A1. Regarding changes in the securities portfolio balance across stages (all securities, including foreign currency denominated securities) (in thousand manats) Balance items Total amount Total Stage 1 Stage 2 Stage 3 POCI 1 2 3 4 5 6

1. Securities measured at amortized cost A1
   Balance at the beginning of the year A1_1 0
   Acquired or purchased during the period
   A1_2 0
   Repaid or sold during the period A1_3 0
   Movement to Stage 1 A1_4 0 0
   Movement to Stage 2 A1_5 0 0
   Movement to Stage 3 A1_6 0 0
   Closing balance of the year A1_7 0
2. Securities measured at fair value
   through other comprehensive income
   A1
   Balance at the beginning of the year A1_1 0
   Acquired or purchased during the period
   A1_2 0
   Repaid or sold during the period A1_3 0
   Change in fair value A1_4 0
   Movement to Stage 1 A1_5 0 0
   Movement to Stage 2 A1_6 0 0

41
Movement to Stage 3 A1_7 0 0
Closing balance of the year A1_8 0
A2. Expected credit losses across stages
(all securities, including foreign currency denominated securities) (in thousand manats) Balance items Expected credit losses Total Stage 1 Stage 2 Stage 3 POCI 1 2 3 4 5 6 Securities portfolio, total A2 Balance at the beginning of the year A2_1 0 Closing balance of the year A2_2 0 Note: For Stage 1, Stage 2, and movements to Stage 3, the corresponding cells should be filled with a negative value When filling in the ‘Loans granted during the period’ rows, loans granted during the reporting period should be recorded under the column corresponding to their stage as of the reporting date.
TABLE A3
A. Off-balance sheet liabilities
(all off-balance sheet items, including foreign currency denominated off-balance sheet items) (in thousand manats) Off-balance sheet items Total amount Expected credit losses

42
Total Stage 1 Stage 2 Stage 3 Total Stage 1 Stage 2 Stage 3 1 2 3 4 5 6 7 8 9 Off-balance sheet items, total A 0

1. Credit instruments A1 0.0
   1.1 Credit liabilities A1_1 0.0 0.0
   1.2 Unused credit lines A1_2 0.0 0.0
   1.3 Other öhdəliklər A1_3 0.0 0.0
2. Guarantees and similar
   commitments
   A2 0.0
   2.1 Guarantees A2_1 0.0 0.0
   2.2 Other similar commitments A2_2 0.0 0.0
3. Letters of credit A3 0.0
   3.1 Standby letters of credit A3_1 0.0 0.0
   3.2 Documentary letters of
   credit
   A3_2 0.0 0.0
4. Other off-balance sheet
   liabilities
   A4 0.0 0.0
   B. Off-balance sheet liabilities
   (in foreign currency)
   (in thousand manats)
   Off-balance sheet items
   Total amount Expected credit losses
   Total Stage 1 Stage 2 Stage 3 Total Stage 1 Stage 2 Stage 3

43
1 2 3 4 5 6 7 8 9
Off-balance sheet items, total B 0

1. Credit instruments B1 0.0
   1.1 Credit liabilities B1_1 0.0 0.0
   1.2 Unused credit lines B1_2 0.0 0.0
   1.3 Other öhdəliklər B1_3 0.0 0.0
2. Guarantees and similar
   commitments
   B2 0.0
   2.1 Guarantees B2_1 0.0 0.0
   2.2 Other similar commitments B2_2 0.0 0.0
3. Letters of credit B3 0.0
   3.1 Standby letters of credit B3_1 0.0 0.0
   3.2 Documentary letters of credit B3_2 0.0 0.0
4. Other Off-balance sheet
   liabilities
   B4 0.0 0.0
   TABLE A3.1
   A1. Regarding changes in the off-balance sheet liabilities balance across stages (all off-balance sheet items, including foreign currency denominated off-balance sheet items) (in thousand manats) Off-balance sheet items Total amount Total Stage 1 Stage 2 Stage 3 1 2 3 4 5 Off-balance sheet items A1

44
Balance at the beginning of the year
A1_1 0
Issued during the period A1_2 0
Decrease during the period A1_3 0
Movement to Stage 1 A1_4 0 0
Movement to Stage 2 A1_5 0 0
Movement to Stage 3 A1_6 0 0
Closing balance of the year A1_7 0
A2. Expected credit losses across stages
(all off-balance sheet items, including foreign currency denominated off-balance sheet items) (in thousand manats) Off-balance sheet items Expected credit losses Total Stage 1 Stage 2 Stage 3 1 2 3 4 5 Off-balance sheet items A2 Balance at the beginning of the year A2_1 0 Closing balance of the year A2_2 0 Note: For Stage 1, Stage 2, and movements to Stage 3, the corresponding cells should be filled with a negative value When filling in the ‘Issued during the period’ rows, loans granted during the reporting period should be recorded under the column corresponding to their stage as of the reporting date.

45
TABLE A4
A. Claims on banks and other financial institutions (all assets, including foreign currency denominated assets) (in thousand manats) Asset items Total amount Expected credit losses Total Stage 1 Stage 2 Stage 3 Total Stage 1 Stage 2 Stage 3 1 2 3 4 5 6 7 8 9 Amounts due from banks and other financial institutions, total A1 0
1.1 "Nostro" accounts A1_1 0.0
a) Resident banks A1_1a 0.0 0.0 b) Non-resident banks A1_1b 0.0 0.0
1.2 Short-term interbank financial
instruments (up to 7 days)
A1_2 0.0 a) Resident banks A1_2a 0.0 0.0 b) Non-resident banks A1_2b 0.0 0.0
1.3 Deposits held with banks and
other financial institutions
A1_3 0.0
1.3.1 Deposits held with banks A1_3_1 0.0
a) Resident banks A1_3_1a 0.0 0.0 b) Non-resident banks A1_3_1b 0.0 0.0
1.3.2 Deposits held with other
financial institutions
A1_3_2 0.0

46 a) Other resident financial institutions
A1_3_2a 0.0 0.0 b) Other non-resident financial institutions A1_3_2b 0.0 0.0
1.4 Loans to banks A1_4 0.0
a) Resident banks A1_4a 0.0 0.0 b) Non-resident banks A1_4b 0.0 0.0
1.5 Other maliyyə institutlarına
kreditlər
A1_5 0.0 a) Other resident financial institutions A1_5a 0.0 0.0 b) Other non-resident financial institutions A1_5b 0.0 0.0 B. Claims on banks and other financial institutions (in foreign currency) (in thousand manats) Asset items Total amount Expected credit losses Total Stage 1 Stage 2 Stage 3 Total Stage 1 Stage 2 Stage 3 1 2 3 4 5 6 7 8 9 Amounts due from banks and other financial institutions, total B1 0
1.1 "Nostro" hesabları B1_1 0.0
a) Resident banks B1_1a 0.0 0.0 b) Non-resident banks B1_1b 0.0 0.0

47
1.2 Short
-term interbank financial instruments
(up to 7 days
)
B1_2 0.0 a) Resident banks B1_2a 0.0 0.0 b) Non -resident banks B1_2b 0.0 0.0
1.3 Deposits held with banks and
other financial institutions
B1_3 0.0
1.3.1 Deposits held with banks B1_3_1 0.0
a) Resident banks B1_3_1a 0.0 0.0 b) Non
-resident banks B1_3_1b 0.0 0.0
1.3.2 Deposits held with other
financial institutions
B1_3_2 0.0 a) Other resident financial institutions B1_3_2a 0.0 0.0 b) Other non -resident financial institutions B1_3_2b 0.0 0.0
1.4 Loans to banks B1_4 0.0
a) Resident banks B1_4a 0.0 0.0 b) Non
-resident banks B1_4b 0.0 0.0
1.5 Loans to other financial
isntitutions
B1_5 0.0 a) Other resident financial institutions B1_5a 0.0 0.0 b) Other non -resident financial institutions B1_5b 0.0 0.0

48
TABLE A4.1
A1. Amounts due from banks and other financial institutionsin stagelərarası dəyişməsinə dair (all interbank claims, including foreign currency denominated interbank claims) (in thousand manats) Interbank claims Total amount Total Stage 1 Stage 2 Stage 3 1 2 3 4 5 Amounts due from banks and other financial institutions A1 Balance at the beginning of the year A1_1 0 Placed during the period A1_2 0 Repaid during the period A1_3 0 Movement to Stage 1 A1_4 0 0 Movement to Stage 2 A1_5 0 0 Movement to Stage 3 A1_6 0 0 Closing balance of the year A1_7 0

49
A2. Expected credit losses across stages
(all interbank claims, including foreign currency denominated interbank claims) (in thousand manats) Interbank claims Expected credit losses Total Stage 1 Stage 2 Stage 3 1 2 3 4 5 Amounts due from banks and other financial institutions A2 Balance at the beginning of the year A2_1 0 Closing balance of the year A2_2 0 Note: For Stage 1, Stage 2, and movements to Stage 3, the corresponding cells should be filled with a negative value