IFRS 9: Impairment rules
By Fridrich Housa ACCA and Lani Biggins CA
Following the GFC, the International Accounting Standards Board (IASB) and the US Financial Accounting Standards Board (FASB), supported by the G20 and regulatory bodies, embarked on the project of re-issuing the key accounting standards that apply to banks. The replacement of IAS 39 Financial Instruments: Recognition and Measurement with IFRS 9 Financial Instruments is set to be the biggest change in banks’ financial reporting since the introduction of IFRS. Impairment guidance under IFRS 9 is still being finalised with a re-exposed draft expected in the second half of 2012. While IFRS 9 is expected to be applicable for reporting periods beginning on or after 1 January 2015, banks and corporates are already starting to assess its likely impact and plan for its implementation.
This table summarises the current status of IFRS 9 and its expected implementation.
Previous Charter articles have covered Phase One – Classification and Measurement and Phase Three – Hedge Accounting (see Charter November 2009, ‘Milestone in Financial Instruments’ and Charter October 2010: ‘Rewriting the Hedging Rules’). This article focuses on Phase Two of the project – amortised cost and impairment of financial assets.
Phase Two development
Phase Two of the IFRS 9 project has been in development for some time now (as illustrated in the table on the opposite page), with major milestones as follows:
- June 2009 – the IASB issued a request for views on whether an expected cash flow model that incorporated expected losses in the estimated cash flows would be operational for banks to implement
- November 2009 – the IASB issued the first Amortised Cost and Impairment Exposure Draft (ED). The main underlying principle was that the expected loan losses would be recognised as the difference between contractual cash flows and expected returns over the life of a loan. The ED relied heavily on the application of the effective interest rate (EIR) method, which increased the complexity of the proposals and was not viewed positively by constituents. The ED introduced further management judgement into the assessment of allowances for impairment losses and introduced the concept of expected loss that until then was used only for regulatory purposes.
There was significant pressure from credit institutions to leverage off the existing credit systems and models used either for the calculation of allowances for impairment losses under IAS 39 (eg Incurred But Not Reported (IBNR) collective provisions) or for regulatory purposes (Basel II, in Australia Prudential Standards APS 112 and APS 113) and thus avoid complexity and implementation costs.
The main concerns expressed by constituents related to a) the determination and allocation of the initial estimate of expected credit losses on a financial asset using a probability-weighted outcome approach and b) the integration of credit loss estimates and interest recognition in the one model. Among other comments the ED did not specifically distinguish between assessing recoverable cash flows on an individual or portfolio basis.
As a response to the comments on the initial ED (mostly on significant operational challenges), the IASB and FASB issued a Joint Supplementary Document in January 2011. This document overcame the original differences between the IASB and FASB’s separate EDs in the timing of the recognition of credit losses and a compromise was reached where losses were recognised over the life of the loan unless certain conditions for earlier recognition were met.
Amortised cost and impairment – Phase Two
Phase Two responds to the challenges revealed by the GFC, which led to large credit losses being incurred in some segments of the financial markets. This has exposed some inadequacies with IAS 39’s loan impairment methodology.
While there are measures incorporated in IAS 39 to reduce procyclicality, such as the requirement to use historical loss experience adjusted on the basis of current observable data to reflect the effects of current conditions (eg by using probabilities of default (PD), and loss-given-default (LGD) estimates), the current ‘incurred loss’ approach recognises loan losses ‘too little, too late’. This approach has the effect of increasing allowances for loan losses in the middle of an economic downturn, which has a negative impact on banks’ regulatory capital and reduces banks’ capacity to lend. This in turn further contributes to a decline in economic activity and market uncertainty.
The amount of losses recognised for loans carried at fair value (using market rates as discount factors incorporating increased credit and liquidity spreads) were higher compared to those carried at amortised cost (using the original effective interest rate) due to the lack of objective evidence of loss. In fact, this contributed to the decline in value of assets marked to market during the global financial crisis, whereas losses on assets at amortised cost remained stable.
Due to the above criticisms of the ‘incurred loss’ model, IFRS 9 will introduce an expected loss model.
The joint supplementary document introduced the two-category (bucket) concept, being either a ‘good’ or ‘bad’ book. This required recognition of lifetime expected credit losses for performing financial assets (‘the good book’) on a time proportional basis with a minimum ‘floor’ amount of credit losses expected in the foreseeable future (which is no shorter than 12 months) and to recognise immediately lifetime expected credit losses in full for financial assets that are identified as non-performing (‘the bad book’).
There were concerns that the definition of the concept of the foreseeable future as well as inclusion of a floor based on credit losses expected in the foreseeable future was unclear. The approach proposed in the joint supplementary document in respect of the good book could result in the allowance being reduced during an economic downturn as a result of the foreseeable future time-horizon being shortened, resulting in lower expected losses in the good book.
In response to the comments received on the joint supplementary document, the IASB, together with FASB, is now working to develop an alternative three bucket impairment approach, which categorises financial assets into three buckets based on the level of credit deterioration (or in the case of purchased financial assets, with an explicit expectation of loss upon acquisition). Categorisation in the three buckets will govern when and how allowances for impairment losses are recognised:
- Bucket One includes the portfolio of assets that do not meet the criteria for either of the second two buckets. The latest proposal is for the loan loss provision to represent expected losses for the next 12 months from the reporting date. All financial assets at amortised cost are classified in Bucket One upon initial recognition (except for certain purchased financial assets)
- Bucket Two includes those financial assets that have been affected by deterioration in credit risk due to the occurrence of an observable credit event, which indicates a possible future default, but the specific assets that will default cannot be specifically identified
- Bucket Three is for an individual asset where information is available that specifically identifies that credit losses have, or are expected to occur for individual assets. For Bucket Two and Three lifetime expected losses would be provided for immediately.
What’s next and what are the challenges?
The IASB has acknowledged that implementation of its proposed expected loss approach may prove operationally challenging and costly and may require significant lead-time to implement. Operational issues may arise from:
- Models and systems to compute estimates of future cash flows and credit losses
- Decisions to assess impairment at an individual or portfolio level
- Availability of historical loss data or credit ratings information
- Estimation uncertainty and the need for subjective and complex estimates and judgements
- The need for management to update credit loss and other cash flow estimates on a periodic basis
- Interaction with regulatory requirements (especially Basel II and III requirements).
Financial institutions will be impacted most, especially in the areas such as provisioning, regulatory and economic capital, loan pricing and credit management. The impact on the pricing of financial assets especially (such as loans or derivatives) could have a flow-on effect to the wider economy.
The following are some of the issues that are already emerging under the proposed three bucket approach:
- Where the credit rating of a financial asset has deteriorated since origination, should the asset be transferred to Bucket Two or Three, particularly if no loss is expected and there is reasonable expectation that contractual cash flows will be collected?
- How can ‘through the cycle’ PDs, which are used for regulatory purposes, be adjusted to reflect a ‘point in time’ PD required for accounting purposes?
- How will the comparability between the banks be ensured, if the requirements for moving loans between Bucket One, Two and Three will be highly dependent on judgement and internal management processes?
- How will the composition of impaired portfolios be determined for the purposes of moving between Bucket One and Two?
- What will be the interpretation of the term ‘significant’ deterioration in credit quality by the constituents?
- How should collateral be incorporated into the calculation of PD?
Hopefully, these issues will be addressed when the re-exposure draft for Phase Two of IFRS 9 is issued later this year.
Fridrich Housa is a manager, assurance & advisory, Deloitte Touche Tohmatsu.
Lani Biggins is a director, assurance & advisory, Deloitte Touche Tohmatsu.
Article last updated 28 June 2012