Guidance Note on Implementation of IFRS 9

REGULATORY GUIDANCE ON IMPLEMENTATION OF IFRS 9
1.0 INTRODUCTION
The International Financial Reporting Standards (IFRS) were adopted in Tanzania in July 2004 as part of measures to improve reporting practices, transparency and disclosures by reporting entities. Adoption of IFRS implies that all revisions to existing standards as well as new accounting standards issued by the International Accounting Standards Board (IASB) must be adopted by all reporting entities. In July 2014, the IASB issued the final version of IFRS 9 (Financial Instruments) to replace IAS 39 (Financial Instruments: Recognition and Measurement) requiring all reporting entities that have adopted IFRS to implement the new accounting standard by January 1, 2018. Bank of Tanzania is interested in high quality implementation of IFRS 9 (the Standard) that results in a reliable measurement of capital and enhances market discipline through greater transparency. In this regard, Bank of Tanzania has issued this guidance in order to ensure robust and consistent implementation of IFRS 9, especially in areas where banks are expected to exercise considerable judgment and/or elect to use simplifications and other practical expedients permitted under the Standard. This guidance is not intended to override IFRS 9 but endeavours to promote consistency of application, facilitate greater comparability across the banking sector and address supervisory concerns. Apart from this guidance, Bank of Tanzania expects banks and financial institutions to adhere to the guidance issued by Basel Committee on Banking Supervision titled Guidance on Credit Risk and Accounting for Expected Credit Losses issued in December 2015 and

Regulatory Treatment of Accounting Provisions-Interim Approach and Transitional Arrangements issued in March 2017.
2.0 IMPLEMENTATION OF IFRS 9
In order to ensure proper and consistent adoption of IFRS 9 requirements in the Tanzanian banking sector, Bank of Tanzania expects banks and financial institutions to observe, at minimum the regulatory guidance prescribed below:
2.1 Measurements of Financial Instruments
2.1.1 Assessment of Significant Increase in Credit Risk
The standard requires an entity, at each reporting date, to assess whether the credit risk on a financial instrument measured at Amortised Cost and Fair Value through Other Comprehensive Income (FVOCI) has increased significantly since initial recognition, using amongst other factors the change in the risk of a default occurring over the expected life of the instrument. Bank of Tanzania expects banks and financial institutions to put in place policies and systems as well as governance arrangements and controls to identify instances where their exposures have suffered significant increase in credit risk.

In assessing significant increase in credit risk, banks and financial institutions shall consider quantitative, qualitative and ‘backstop’ (30 days past due presumption) indicators. In using quantitative elements, banks and financial institutions should consider the change in lifetime Probability of Default (PD) by comparing the lifetime PD at the reporting date with the lifetime PD at initial recognition. The criteria for relative quantitative increases in PD indicative of a significant increase in credit risk should be defined and documented by banks and financial institutions. The assessment by banks and financial institutions shall, among others, consider changes in credit risk at counterparty and individual credit level. Generally, most qualitative factors indicative of a significant increase in credit risk are reflected in PD models and therefore, are included in the quantitative assessment. However, where it is not possible to include all current information about qualitative factors in the quantitative assessment, banks and financial institutions should recalibrate PDs or adjust

estimates when assessing significant increase in credit risk or calculating Expected Credit Losses (ECLs). IFRS 9 requires shift of credit exposure from stage 1 (where 12-month ECL is required) to stage 2 (when lifetime ECL is considered) when there is a significant increase in credit risk. In addition to the criteria provided in paragraph B5.5.17 of the Standard, banks and financial institutions are required to consider the following qualitative factors in assessing significant increase in credit risk:
 Credit scoring of the obligor by any of the licensed private credit bureaux;  Deterioration of collaterals pledged as securities for the credit exposures;  Expectation of restructuring due to financial difficulties;  Evidence that full repayment of interest and principal is unlikely in the future, regardless of the number of days past due;  Deterioration in credit worthiness of the borrower such as changes in business location without notification, failure to keep promises and poor management of collateral; and  Consider macroeconomic indices and sector/industry/geographical idiosyncrasies. Where a bank or financial institution sets its transfer threshold for groups of financial assets, it is important that all financial instruments in that portfolio must have similar credit risk characteristics at initial recognition such as a credit rating within a relatively narrow band. To ensure appropriate identification of significant increase in credit risk, banks and financial institutions should avoid applying absolute PD or credit rating threshold to all exposures in a portfolio except where exposures are of a similar credit risk at initial recognition. The use of absolute threshold is only permitted if it would appropriately capture significant increase in credit risk since initial recognition in a manner consistent with the requirements of IFRS 9. Banks and financial institutions should consider both counterparty and individual exposures of the obligor and connected obligors, in determining significant increase in credit risk. This

would ensure that the impact of multiple exposures to the same obligor originated at different periods with different initial PDs have been taken cognisance of in compliance with IFRS 9.
2.1.2 Rebuttable Presumptions
IFRS 9 allows banks and financial institutions to use more than 30 days past due rebuttable presumptions on the basis that there has not been a significant increase in credit risk. In any case the number of days past due should not exceed 60 days. However, the Bank of Tanzania requires banks and financial institutions to ensure that, the presumption is supported by thorough analysis evidencing that 30 days past due is not correlated with significant increase in credit risk. Such analysis should consider both current reasonable and supportable forward looking information that may cause future cash short falls to differ from historical experience. Information regarding thorough analysis evidencing that 30 days past due is not correlated with significant increase in credit risk shall be availed to the Bank of Tanzania whenever required.
2.1.3 Staging and Migration
At transition, banks and financial institutions are expected to place financial instruments without significant increase in credit risk in the 12-months ECL bucket irrespective of the obligor’s credit risk rating at origination. However, where significant increase in credit risk has been observed, such credits shall be moved to Lifetime ECL. IFRS 9 allows credit exposures to migrate from higher credit risk categories to lower credit risk categories, that is, from stage 3 to stage 2 and from stage 2 to stage 1. Migration from stage 3 to stage 2 should consider criteria for upgrade of credit accommodations as follows:
 in the case of overdraft facilities, the account has satisfactorily performed for a minimum period of two consecutive quarters; and  in the case of term loans, the obligor has timely paid four consecutive installments.

On the other hand, credit exposures may migrate from stage 2 to stage 1 when there is a significant improvement of the credit exposure. The following considerations should be used in determining whether an exposure should shift backward from stage 2 to stage 1:
 Up-to-date with payments: All outstanding payments on the credit facility are made on time and no payments are in arrears; and  Improvement of the quantitative and qualitative factors that caused significant increase of the credit risk. Upgrade from stage 2 to stage 1 shall be subject to a monitoring period of 90 days for conventional loans and 30 days for Microfinance loans to confirm if the risk of default has decreased sufficiently before upgrading such exposure.
2.1.4 Impairment of Financial Instruments
Banks and financial institutions shall put in place appropriate policies to ensure sound credit risk assessment and measurement processes as well as systems, tools and data that appropriately aid the assessment of credit risk and computation of ECLs. In accordance with paragraph 5.5.1 of the Standard, banks and financial institutions are required to determine ECL for financial assets measured at amortised cost or FVOCI, lease receivables, contract asset, loan commitment and financial guarantee contract.
2.1.4.1 Expected Credit Loss (ECL) Model
Banks and financial institutions shall put in place a sound ECL model to estimate expected credit losses at both the individual lending exposure and overall portfolio levels. Models used in the ECL assessment and measurement process should consider the impact of changes to borrower and credit risk-related variables such as changes in PDs, LGDs, exposure amounts, collateral values, migration of default probabilities and internal borrower credit risk grades based on historical, current, and reasonable and supportable forward-looking information, including macroeconomic factors such as interest rates, unemployment rates and commodity prices published by National Bureau of Statistics, BOT, and other reputable local and international institutions. Banks and financial

institutions shall adopt sound ECL methodologies commensurate with their size, complexity, structure and risk profile. Banks and financial institutions are required to compute ECLs on significant exposures (non-retail) and credit-impaired loans individually while ECLs for retail exposures that have less borrower-specific information may be measured on collective basis. To measure ECL on collective basis, banks should have credit risk rating processes in place to appropriately group exposures on the basis of shared credit risk characteristics. ECL shall be computed in accordance with the requirements of IFRS 9. Banks and financial institutions are required to carry out credit review and update their ECLs, at least quarterly, to reflect changes in credit risk since initial recognition. The methodologies and assumptions underlying the ECL methodology should be reviewed at least annually. In computing ECL, banks and financial institutions shall consider the following:
(a) Probability of Default (PD)
In estimating PD, banks and financial institutions are required to consider quantitative and qualitative indicators. The qualitative indicators to be considered at minimum should take into consideration factors used when establishing whether there is significant increase in credit risk. In determining PDs, banks and financial institutions are also required to ensure that all key risk drivers and their predictive power are identified and calibrated based on historical data for 5-10 years. Those that do not have 5 to 10 years historical data may utilise data sourced externally. Banks are not permitted to assume a constant marginal rate of default over the remaining lifetime of a financial instrument without appropriate supporting analysis. As provided under paragraph B5.5.5 of the Standard, only exposures that have similar credit risk characteristics can be grouped together for the purpose of calculating the PDs.

(b) Loss Given Default (LGD)
Banks and financial institutions are required to analyse relevant macroeconomic indicators that influence LGD or its components to aid estimation of collateral values when modelling the term/structure of LGD. The banks and financial institutions that already have or intend to use Basel LGD values should effect necessary adjustments to comply with IFRS 9. The methodology for determining LGD where appropriate should be designed at a component level with the calculation of LGD broken down into drivers. For secured exposure, the amount of expected recovery should be calculated based on the present value of the amounts expected to be received from foreclosure of collateral less cost of recovery, and based on the estimated future value of the collateral and other reliable resources according to documented experience of the bank or financial institution. At minimum, the bank or financial institution should consider forecasts of future collateral valuations (including expected sale discount), time to realisation of collateral (and other recoveries), allocation of collateral across exposures where there are several exposures to the same borrower, recovery rates, haircuts, type of collateral and external costs of realisation of collateral. The following table provides guidance on key assumptions for computations of Net Present Value (NPV) of expected recovery from collateral:
Assumption Description
Forward collateral projection
Banks and financial institutions should determine the value of the collateral at expected time of foreclosure of the collateral. Haircut Banks and financial institutions should apply reduction to the value of the collateral due to forced/distressed foreclosure. The haircut applied shall base on the banks or financial institution’s past experience and the credit rating of the guarantor. Banks and financial institutions may consider to use full amount of the following collaterals: cash or treasury bills, notes or bonds, or other instruments as the Bank of Tanzania may approve. Cost of recovery The costs incurred to possess and sell the collateral, including legal fees and agent fees should be deducted when measuring the amounts recoverable. Time to recovery Banks and financial institutions should be guided by their past experiences as to the time required between default and recovery of collateral. Interest/yield rate The discount rate as prescribed in item (d) below.

(c) Exposure at Default (EAD)
Banks and financial institutions are required to ensure that the period of exposure in the model is not shorter or longer than the maximum contractual period over which the entity is exposed to credit risk. To determine the period of exposure that equals the historical average life of loans, banks and financial institutions are required to evaluate whether it is consistent with forward-looking expectations based on reasonable and supportable information. For revolving credit facilities, a bank or financial institution shall estimate 12-month ECL based on its expectations of the portion of the loan commitment that will be drawn within 12 months of the reporting date while lifetime ECL is calculated based on the expected portion of the loan commitment that will be drawn over the expected life of the loan commitment. In determining the period of exposure for revolving credit facilities, banks and financial institutions shall take cognisance of their expected credit risk management measures which serves to mitigate credit risk including terminating or limiting credit exposure. Banks and financial institutions are permitted to use the legally enforceable contractual period for revolving credit facilities provided that analysis of historical data shows that, in practice, management action consistently limits the period of exposure to the contractual period. Banks are expected to consider all relevant historical information that is available without undue cost and effort when determining the exposure period of a revolving credit facility. ECL computation for financial guarantee contracts shall consider expected payments to reimburse the holder for a credit loss that it incurs, less any amount banks expects to receive from the holder, debtor or any other party. Banks and financial institutions are required to demonstrate that their EAD models are fit for purpose under IFRS 9. The basis for inputs and adjustments should be documented.

Banks and financial institutions should not use 12-month EAD as a proxy for lifetime EAD without appropriate justification. (d) Effective Interest Rate and Discount Rate Banks and Financial Institutions are required to make appropriate adjustments in contractual interest rate in attaining effective interest rate used to compute IFRS 9 ECL/LGD. In making these adjustments, banks should consider all charges associated with the credit facility. Banks and financial institutions are required to reflect the effect of time value of money in ECL by ensuring that the discount rate used approximates the effective interest rate (EIR). For a financial guarantee contract, the discount rate should reflect the current market assessment of time value of money and risks specific to the cash flows. Assumptions about prepayments, extensions and utilisation during the period of exposure used in the ECL calculations shall be updated to reflect currently available information consistent with that used in estimating interest income. (e) Definition of Default Default means any credit accommodation for which contractual obligation for repayment is past due for more than 90 days for banks and financial institutions and 30 days for Microfinance banks or qualitative criterion by which the bank or financial institution considers that the obligor is unlikely to pay its credit obligations in full, without recourse to actions such as realising security (if held)” (“unlikeliness to pay” events).
2.1.4.2 Validation of ECL Model
Banks and financial institutions shall put in place policies and procedures to validate models used to assess and measure ECL. Banks and Financial Institutions should ensure that the validation process allows for systematic evaluation of robustness, consistency and accuracy of the model as well as its relevance to the underlying portfolio.

Model validation should be conducted when the ECL models are initially developed or adopted and when significant change are made to the models. The scope for validation should include review of assumptions, inputs, design and outputs. The scope of validation should also establish thresholds for model performance which, if breached, should lead to remedial actions such as model recalibration or redevelopment. Model validation should be performed independently of the model development process and by experienced personnel(s) with requisite expertise. The outcome of the validation process should be documented and subjected to review by the bank’s and financial institution’s internal and external auditors. The initial ECL model validation report should be submitted to the Bank of Tanzania by 31st March 2018 for review and clearance. The findings and outcomes of model validation should also be reported in a prompt and timely manner to senior management and Board. Further, banks and financial institutions shall conduct reviews to ensure that their ECL models are appropriate at least annually. The Bank of Tanzania shall periodically evaluate the effectiveness of the banks’ and financial institutions’ credit risk management practices to ensure among others, that the methods used for determining ECLs are in line with the Standard.
2.1.4.3 Documentation of ECL Model
Banks and financial institutions are required to document processes and procedures used in determining ECL; model validation framework, processes and procedures; changes to the model methodology and tools; the range of data used; validation results; and remedial action taken when necessary. Banks and financial institutions are required to ensure that documentation is regularly reviewed and updated.
2.1.4.4 Low Credit Risk Exemption
IFRS 9 permits financial instruments that are considered to have low credit risk on the reporting date, should not be assessed for significant increase in credit risk since its initial recognition. The Bank of Tanzania expects banks and financial institutions to exercise this

exemption on limited basis. Accordingly, banks and financial institutions may use this exemption on the Loan or Overdrafts or portion of the Loan or Overdrafts secured by Cash or Government Securities.
2.1.4.5 Interest Revenue
Paragraph 5.4.1 of the standard stipulates that Interest Revenue shall be calculated by using the effective interest method, and it shall be calculated by applying the effective interest rate to the gross carrying amount of a financial asset except for credit impaired financial assets where entities shall apply the credit adjusted effective interest to the amortized cost of the financial assets from initial recognition. However, banks and financial institutions are required to abide to the relevant existing regulations guiding recognition of interest income from non-performing loans.
2.1.4.6 Write-off of Non-Performing Financial Assets
Paragraph 5.4.4 of the Standard requires an entity to write-off the portion of the gross carrying amount of a financial asset for which the entity has no reasonable expectations of recovery, with such write-off constitutes a de-recognition event. However, banks and financial institutions are required to abide to relevant existing regulations guiding write￾off of non-performing financial assets.
2.1.4.7 Restructured Financial Assets
The contractual terms of a financial asset may be restructured or renegotiated for a number of reasons, including factors not related to current or potential credit deterioration of the customer (e.g. changing market conditions, customer retention, etc.) For financial assets restructured while under lifetime ECL should not be moved to 12- months ECL unless it complies with the requirements of item 2.1.3 (Staging and Migration) of this guidance. For financial assets which have been restructured while under 12-months ECL and there is evidence of significant increase credit risk should be treated as originated credit impaired financial asset.

3.0 TRANSITION ARRANGEMENT FROM IAS 39 TO IFRS 9
Banks and Financial institutions shall apply IFRS 9 prospectively. The transition from IAS 39 to IFRS 9 is likely to result into an increase in impairment for credit losses resulting into reduced profitability for banks and financial institutions. This increase in impairment loss may ultimately result into significant reduction of capital. Bank of Tanzania is aware of this potential “ shock” on capital. Accordingly, it has deemed necessary to introduce a transitional arrangement to avoid “capital shock”. The objective being to smooth out the impact on capital resulting from the introduction of Expected Credit Loss (ECL) accounting. In the first adoption, when IFRS 9 impairment is greater than BOT provision, for the purpose of computing Core Capital banks and financial institutions shall spread the excess impairment equally over three years. BOT will keep the current approach of comparing the IFRS impairment with BOT provision, as well as General provision of 1% for all loans classified as current.