25 March 2013

FEE's comment letters to EFRAG and IASB on classification and measurement: limited amendments to IFRS 9

Generally, FEE supports the direction of the proposals in the ED. Like EFRAG, FEE welcomes the fact that the IASB considers the interaction with the project to revise IFRS 4 Insurance Contracts in the finalisation of IFRS 9 Financial Instruments. In respect of addressing accounting mismatches, the proposals in the ED represent a step forward but do not wholly address concerns.

The actual changes proposed could be considered in many respects more complex than current provisions in IAS 39 Financial Instruments: Recognition and Measurement and IFRS 9 (2010). There is a risk of issuing a new standard which is at least equally complex and not fully understandable, when one of the original aims for replacing IAS 39 was to reduce complexity. This risk needs to be kept in mind when finalising the standard and complexity reduction should be implemented wherever achievable.

Interaction with future IFRS on Insurance Contracts and Accounting mismatches

The introduction of the new measurement category of financial instrument Fair Value through Other Comprehensive Income (FVOCI) will help reduce inter alia the accounting mismatches between the measurement of financial assets and the related insurance contracts liabilities by recognising FV gains and losses in OCI. This provides a strong justification for the reintroduction of this measurement category following the IASB’s tentative decision on IFRS 4 Phase II, according to which, in order to remove the short-term volatility from the profit and loss line, gains and losses attributable to changes in market discount rates for insurance contracts shall be recognised in OCI.

However, as noted by EFRAG, FEE believes that there are still many concerns regarding the proposals in the ED.

The proposed new category of FVOCI will not avoid all accounting mismatches. As noted above, we understand the IASB tentatively agreed on a mandatory use of OCI for effects of changes in interest rates in insurance contracts. Therefore, accounting mismatches will remain, where assets of an insurer do not qualify for FVOCI under the proposals in this ED (e.g. derivatives used to hedge interest rate risks or instruments measured at amortised cost), unless solved through the hedge accounting proposals.

Many argue that the new categorisation of financial instruments proposed in the ED diminishes the clear impact of the “business model” as a primary characteristic to decide the category into which a particular financial instrument should be placed.

Therefore FEE supports the view that this category should be made available only if it reduces or mitigates accounting mismatch. It should not be mandated, since what might be right for banks in terms of accounting will not necessarily be right for insurance contracts accounting or general businesses and vice versa. Hence an option with respect to the use of the new measurement category would be helpful to avoid accounting mismatches.

For insurance contracts with participating features, where the mirroring approach applies, insurers should be allowed to present changes in insurance contract liabilities in OCI consistently with the presentation of changes in the directly linked underlying items. In any case, the IASB should provide for an appropriate interaction between the accounting for financial assets pursuant to IFRS 9 and the accounting for insurance contract liabilities.

The approach proposed in paragraph 7 is closely linked with the amortised cost category definition, which should ensure that standard debt instruments routinely used in traditional banking are eligible to remain in the amortised cost category. Debt instruments that are part of a wider held to collect model, measured at fair value whether through OCI or profit or loss, creates accounting mismatches and provides less useful information to users.

Amortised cost category definition

FEE agrees with EFRAG that there are still certain financial assets that do not pass the contractual cash flow characteristic assessment and for which an amortised cost (AC) measurement would provide more useful information than measurement at fair value.

FEE's concerns therefore remain regarding those assets that will not pass the test despite their consistency with the “held to collect” business model.

FEE agrees with EFRAG that the definition of interest in IFRS 9 should be revised and that it would make sense to have it aligned with the recent tentative decisions on the insurance contracts project, i.e. widening the definition and to clarify that it includes other inherent components (such as liquidity risk), if a definition is necessary at all.

In order to foster simplification and plain language, we recommend replacing the term “more than insignificant” by “significant” when assessing any contractual terms of an instrument, which should in theory not change the final result but would be clearer. It would also avoid the perception that there is a third category which is neither insignificant nor significant. FEE supports the view that only significant leverage, yield curve mismatch or any other significant deviation from the definition should require the fair value categorisation.

Regarding the business model considerations, FEE is of the view that more attention should be paid to the portfolios used to manage liquidity. FEE proposes that the definition of amortised cost should also encompass the business model applicable to liquidity portfolios with predetermined portfolio characteristics such as duration and credit risk profile, where individual purchases and sales are clearly justified by the goal to retain the predetermined portfolio characteristics.

In summary, the proposals may still be too restrictive and in some cases create unintended or unreasonable accounting mismatch or force traditional banking instruments held to collect contractual cash-flows into the fair value measurement category. Therefore, as described above it would be appropriate to further review the amortised cost definition and also to allow FVOCI on an optional basis for instruments outside the trading portfolio that otherwise would be in FVPL but to a limited extent also in amortised cost, in order to reduce or eliminate an accounting mismatch.

Bifurcation

Regarding the issue of “bifurcation” of hybrid financial assets, FEE’s position has been that the long term objective of a principles-based standard should be a single classification approach for hybrid contracts with financial hosts. Therefore, in order to meet the objective of reducing complexity in financial instruments accounting, FEE agreed in this respect with the direction of the IASB proposals to eliminate bifurcation of embedded derivatives for financial assets and currently retaining the rules for embedded derivatives in financial liabilities. However, from a principles point of view FEE questions why liabilities are not treated in the same way as assets.

Recycling

Regarding the issue of “recycling”, the amended proposal demonstrates the lack of principles in the various IFRSs and in this particular case also inside the same standard on financial instruments, since fair value changes in the equity securities classes are not recycled and fair value changes in the debt securities classes are proposed to be recycled on realisation or impairment. FEE would reiterate our longstanding position that IASB needs to formulate its principle-based position in the recycling debate.

Since FEE is not aware of any justifiable principle-based criterion to decide that some items should be recycled and others not, FEE concludes that recycling should be in general either required or prohibited. Since the IASB stresses the prominent position of the net income line in the performance statement, recycling of all relevant OCI items on realisation or impairment seems to be the logical response.

Comment letter to IASB

Comment letter to EFRAG

© FEE