International Accounting Standards Board

Mr Hans Hoogervorst

Chairman

30 Cannon Street

London EC4M 6XH

Submitted electronically to www.ifrs.org

5th July 2013

Dear Mr Hoogervorst

BBA RESPONSE TO IASB EXPOSURE DRAFT ED/2013/3 FINANCIAL INSTRUMENTS: EXPECTED CREDIT LOSSES

The British Bankers’ Association welcomes the opportunity to comment on the IASB’s Exposure Draft on Financial Instruments: Expected Credit Losses. We represent over 200 banks from 50 countries on UK and international banking issues.

The members of the BBA appreciate the considerable efforts made, including extensive outreach and consultation, by both the IASB and the FASB to develop practical proposals for the impairment of financial assets which use more forward looking information to address the perceived shortcomings of the incurred loss model exposed during the financial crisis.

Although there is some common ground between the current IASB and FASB proposals, it appears unlikely that there will be a converged approach to impairment accounting. The existence of two impairment models is, we believe, likely to have a number of unwelcome consequences:

· A lack of comparability of market information on changes in the credit quality of banks’ assets and a consequential impact on international competitiveness;

· A reduction in market confidence in financial reporting; and

· Market pressures upon international banks to provide information on both bases. This will significantly increase implementation and on-going costs, and increase the complexity of information provided without a proportional increase in its usefulness.

We therefore urge the Boards to consider the implications that two impairment models will have for users and preparers, and make a renewed effort to reach a converged solution.

The BBA believes that convergence towards an improved accounting standard for impairment losses should be the overriding objective. The differences between the IASB and FASB approaches provide an opportunity for stakeholders to consider which attributes are most important in providing decision useful information on expected credit losses, at a cost that the benefits justify. The BBA considers that the most important attributes of a new impairment model are that it should:

1. Reflect the economics of the business activity it purports to represent.

2. Be capable of alignment to internal credit risk management practices.

3. Be based on principles which are sufficiently clear to establish comparability across different organisations in different countries.

4. Be operational at a cost which justifies the benefit of the improvements in information provided.

5. Not give rise to unintended effects on lending activity, including disproportionate impacts on particular sectors of lending activity.

We discuss each of these in detail below.

1 Reflecting the economics of the business activity it purports to represent

The recognition of expected losses at the date of initial recognition, and the consequent overstatement of performance in following years, distorts the economics of lending relationships as it severs the economic link between the fair value of a loan or debt security on initial recognition and the credit quality of the instrument. Although both proposals contain this flaw, its effect is much more pronounced in the FASB proposal. We are therefore support the IASB’s approach as a compromise as it strikes a better balance between the operational challenges of implementation and the economics of providing credit.

It is important that the accounting model reflects the time value of money as this can have a material impact on the calculation of allowances. We disagree with the FASB model as certain methodologies permitted by the proposals do not take into account of the time value of money. While we support the IASB model for considering the time value of money explicitly, the wide choice of discount rates allowed by the Exposure Draft to improve the operability of the model, is likely to lead to significant measurement inconsistencies in practice. We suggest that the IASB explain the purpose of discounting and provide clear guidance on what discount rates are to be used.

2 Alignment to internal credit risk management practices based on expected loss

Any accounting model for the impairment of financial instruments, in order to reflect the economics of providing credit, must align with the entity’s credit risk management practices.

An approach to the recognition of impairment that is founded on the credit life cycle and credit risk management practices and which recognises increasing expected losses as credit quality deteriorates provides a faithful representation of portfolio credit quality and therefore of the underlying economics of lending. Current credit risk management practices collate and monitor data on the credit quality of assets. Alignment of financial reporting to credit risk management would enable preparers to make best use of existing data rather than having to create new data sets purely for accounting purposes. Furthermore, recognising impairments as credit risk metrics deteriorate provides users of financial statements information about the effects of changes in the credit quality of assets that would not be as transparent if lifetime losses for all financial instruments were recognised at origination. For financial assets which do not exhibit, and are not expected to exhibit, signs of significant deterioration, it is possible to adapt existing methodologies and data to estimate expected losses that may result from default events that are possible in the next 12 months. While the effort involved in adapting existing methodologies should not be underestimated, extending the estimation period beyond 12 months would not be supported by existing credit risk management practices and new methodologies would need to be developed.

The IASB model is therefore better aligned to internal credit risk management practices than the FASB model, which is significantly more complex to implement and requires a far higher level of management judgment. These factors impact on its reliability, operability, and the time it will take to implement. While the IASB model will still require significant adjustments to adapt the existing internal credit risk management practices to be compliant with the principles set out in the Exposure Draft, this is achievable within a reasonable time-frame.

3 Clear principles which establish comparability across different organizations in different countries

In seeking an approach capable of being implemented by a wide range of different entities, there is a balance to be struck between permitting different methods of achieving the desired outcome, and avoiding unacceptably wide diversity in practice. The BBA believes that both models require greater clarity of definition to support consistency in application while recognising that entities have different businesses and capabilities.

While the IASB model involves very significant judgments, it would be possible to achieve greater consistency in application of the concept of credit deterioration given its relevance to existing credit risk management processes than would be the case with the FASB model approach. Further work on the proposed disclosures is needed to ensure that they provide useful insights into the credit characteristics of financial instruments in the context of the impairment model.

4 Operability at a cost which justifies the benefit of the improvements in information provided

The IASB has rightly placed great emphasis on the operability and the effects of the proposals.

Some of our members have modelled the impact and the performance of the IASB and FASB models on a sample of their portfolios based on common economic scenario. We would draw the Board’s attention to the following key findings of these exercises:

a) Both models deliver earlier loss recognition and higher impairment allowances and are more responsive to deteriorating economic change than the incurred loss approach in IAS 39.

b) Alignment to existing risk management approaches was instrumental to completing the exercise.

c) Both models require extensive data, even where this is available, it needs to be gathered from different sources. Portfolios within the Advanced Internal Rating-Based (AIRB) approach for regulatory purposes tend to be data rich, with behavioural scores and other measures of credit quality available. Banks using the Basel Standardised Approach will find the new requirements more challenging as they have less modelling capability and will have issues in terms of data availability and quality.

d) The models require many more assumptions and a substantially higher level of estimation than is the case under IAS 39, particularly in respect of economic forecasting. Where forecasts are available with the granularity required by the models, they tend to be over a five year period at most and typically revert to a long run average. It needs to be appreciated that economic forecasts are not available in the required detail in many countries.

e) The FASB lifetime expected loss approach produces a higher impairment allowance but is significantly more complex in operation, requires an increased level of estimation and relies to a greater extent upon management assumptions.

Our overall assessment is that the IASB model is operable, but that it will require a great deal of estimation and considerable resources and time to implement to the high standards required for financial reporting. We do not consider the FASB model to be operable to the required standard.

5 Avoid negative economic effects on lending activity which result purely from the accounting model adopted

Many commentators have concluded that the FASB lifetime expected loss will have negative consequences for lending activities due to the effects of recognizing lifetime losses on initial recognition of loans and loan commitments, and the general reduction in regulatory capital which may result from the model. Commentators have also noted that the impacts on certain forms of lending, such as long-term or higher risk lending, including lending to the SME sector, will be particularly marked. We share their concerns.

While the effects of the IASB model are likely to be much less severe than the FASB model it still shares some of these characteristics, notably the potential impact on regulatory capital and the profitability of individual portfolios. The impact on equity, and therefore, regulatory capital, might also be volatile and this could affect banks’ ability to lend in worsening economic conditions.

We estimate that the impacts will be:

· Reduced availability and higher cost of loans with longer maturities or to lower-rated borrowers such as small and medium sized enterprises. The required return for these longer dated or lower rated loans may become higher than is warranted by the economic risk.

· Reduced diversification of banks’ balance sheets as banks seek to de-emphasise longer maturity and higher credit risk lending.

· Incentives for greater distortion in the structure of the financial services industry encouraged by arbitrage opportunities between banks (seeking to manage the impact of heavily front-loaded provisions) and shadow-banking market participants for whom the interplay between the accounting and regulatory factors is less significant.

These significant potential impacts on the wider economy and the financial system need to be carefully considered by standard setters, regulators and governments. In view of the above, we are supportive of the IASB model under which the severity of the negative economic effects on lending as described above will be lower. We are also concerned that a failure to converge will result in similar distortive impacts between credit markets internationally.

6 Lead-time for implementation

Whilst the proposals can be implemented using approaches based on the calculations that banks perform to meet regulatory capital requirements, these models are very different in practice, as can be seen from our comments on operability, above. For many portfolios and assets, existing models will need to be substantially adapted to address the more forward looking nature of the requirements, and the adjustments necessary to probability of default, loss given default and discounting. For other portfolios, entirely new impairment models will have to be developed. Above all else, methodologies for estimating lifetime expected loss will need to be derived.

The BBA recognises the importance of finding a solution that can be implemented within a timescale acceptable to stakeholders. However, there are likely to be significant operational challenges in implementing a more forward looking impairment accounting model. In addition, there will be complex interactions with regulatory capital and taxation that will require careful analysis and engagement with the relevant authorities, as well as the assessing any business impacts and the implementation of governance and control frameworks. Furthermore, public companies will want to communicate and explain the likely effects of the new accounting standard on their business to investors and other market participants well in advance of the transition date.

Given the significance of the changes, and confirmed by our members modelling work, the BBA believes that a minimum of three years will be needed from the date of publication of the final standard to the effective. Any shorter timescale for a project of this complexity would involve unacceptable operational risk.

It is important that the final requirements, including disclosures, are available by the end of 2013 so that companies can start their implementation projects having certainty as to the requirements, which is particularly important given the potentially high costs of the implementation programmes.

Our responses to the questions set out in the IASB ED are provided in the Appendix. We hope you find these comments are of value. Please feel free to contact us if you would like to discuss any of the points in more detail.

Yours sincerely

Paul Chisnall
Executive Director
T +44(0)20 7216 8865
E

Appendix I: Questions for respondents

Question 1

(a) Do you agree that an approach that recognises a loss allowance (or provision) at an amount equal to a portion of expected credit losses initially, and lifetime expected credit losses only after significant deterioration in credit quality, will reflect:

(i) the economic link between the pricing of financial instruments and the credit quality at initial recognition; and

(ii) the effects of changes in the credit quality subsequent to initial recognition?

If not, why not and how do you believe the proposed model should be revised?

(b) Do you agree that recognising a loss allowance or provision from initial recognition at an amount equal to lifetime expected credit losses, discounted using the original effective interest rate, does not faithfully represent the underlying economics of financial instruments? If not, why not?