Impairments of Greek Government Bonds under IAS 39 and IFRS 9: A Case Study

Abstract

International Financial Reporting Standard 9 (IFRS 9) 9 introduces new impairment rules responding to the G20 critique that International Accounting Standard 39 (IAS 39) results in the delayed and insufficient recognition of credit losses. In a case study of a Greek government bond for the period 2009–2011 when Greece’s credit rating declined sharply, this paper highlights the discretion that preparers have when estimating impairments. IFRS 9 relies more on management expectations and will lead to earlier impairments. However, these appear still delayed and low if compared to the fair value losses.

Keywords:

• financial instruments
• IAS 39
• incurred credit loss
• IFRS 9
• expected credit loss
• impairment
• loan loss provisioning

Acknowledgements

The comments by David Grünberger, Sue Lloyd, Michael Reiland, Yvonne Wiehagen-Knopke and Carsten Wittenbrink are highly appreciated. The paper is based on the study IP/A/ECON 2015-14 conducted at the request of the Committee on Economic and Monetary Affairs of the European Parliament. Permission to use the study for publication in academic journals has been granted with a letter D 319237 dated 13 November 2015 by the Acting Director of the Directorate-General for Internal Policies Directorate A – Economic and Scientific Policy of the European Parliament. The study is available from the website of the European Parliament: http://www.europarl.europa.eu/RegData/etudes/STUD/2015/563462/IPOL_STU(2015)563462_EN.pdf. The author is since April 2015 the academic member of the Technical Experts Group (TEG) of the European Financial Reporting Advisory Group (EFRAG). The paper has been prepared as an academic paper independent of EFRAG.

Disclosure Statement

No potential conflict of interest was reported by the author.

Notes

1 See IASB & FASB (2013) and O’Hanlon, Noor, and Weijia (2016) for the history of the standards development.

2 From a regulatory perspective, unexpected losses shall be absorbed by adequate levels of capital.

3 By earnings management preparers use judgement or accounting choices offered by standard setters to achieve earnings targets, for example, to avoid losses, to increase earnings above prior period earnings or to meet earnings expectations by analysts. See Healy and Wahlen (1999) for a review of the related literature.

4 IAS 39.45 requires an entity to classify financial assets at initial recognition in one of the four categories as defined in IAS 39.9: (1) held-to-maturity investments are non-derivative financial assets with fixed or determinable payments and a fixed maturity for which the entity has the intention and ability to hold them to maturity; (2) loans and receivables are non-derivative financial assets with fixed or determinable payments and a fixed maturity which are not quoted in an active market; (3) financial assets at fair value through profit and loss are financial assets held-for-trading, derivatives, or financial assets designated as at fair value under the fair value option; (4) available-for-sale financial assets are non-derivative financial assets which have not been classified as either (1), (2), or (3). If classified as (1) held-to-maturity or as (2) loans and receivables subsequent measurement of these financial assets is at AC using the effective interest method (IAS 39.46). Measurement of all other financial assets is at fair value. Changes in fair value are recognised in net income except for financial assets classified as (4) available-for-sale for which changes in fair value are recognised in other comprehensive income.

5 See Novotny-Farkas (2016) for an analysis of the interaction of the IFRS 9 impairment rules with bank supervisory regulations.

6 See O’Hanlon et al. (2016) for a comparison of the impairment rules in IFRS 9 and the proposed US GAAP impairment rules.

7 The Basel Committee on Banking Supervision (2015, p. 32), however, cautions that investment grade-rated financial assets ‘cannot automatically be considered low credit risk’.

8 See Grünberger (2013, pp. 382–386) for a discussion of the literature on the credit spread puzzle.

9 The increase between rating grades is not linear so that an absolute comparison of default probabilities is not adequate, see Grünberger (2013, pp. 34–36, 140–146) for a discussion.

10 Different maturities are used here because by 30 June 2010 the residual maturity has shortened to nine years.

11 For the assessment of changes in credit risk IFRS 9.B5.5.17 provides a list of potentially relevant information. This includes item:

(l) significant changes, such as reductions in financial support from a parent entity or other affiliate or an actual or expected significant change in the quality of credit enhancement, that are expected to reduce the borrower’s economic incentive to make scheduled contractual payments

With the rescue package in place it will be difficult to argue that there has been a change in the support of the other euro area countries as of 30 June 2010.

12 The full fair value model as proposed by the Joint Working Group (2000) would measure all financial assets (and liabilities) at fair value and thus measure identical financial assets at identical amounts. Despite heavy criticism in particular from the financial industry IASB and FASB declared a full fair value model as their aim for a long-term solution for accounting for financial instruments up to 2008, see, for example, IASB (2008). In the wake of the global financial crisis this aim has been dropped.

13 See, for example, the press reports SPIEGEL ONLINE (2011a) and SPIEGEL ONLINE (2011b).

14 See Zettelmeyer et al. (2013, pp. 5–6) for a report on the discussion.

15 In November 2011 an ESMA Public Statement similarly argued that (a) Greece was in financial difficulties and (b) that the 21 July 2011 agreement on a private sector involvement would have indicated concessions as of 30 June 2011; see ESMA (2011, pp. 6–7).

16 An exception is the Portuguese Banco BPI which ‘believes that the restructuring plan of the Greek debt will be successful’ and therefore ‘does not expect to incur a loss on the Greek Government Bonds portfolio, so it has not recognized impairment on these exposures in the financial statements of June, 30, 2011’. See Banco BPI (2011, p. 203).

17 Each of the expected cash flows may itself represent a probability distribution to be represented, for example, by the probability weighted average of the possible cash flows (i.e. in statistical terms the expected value). Sometimes cash flows are used that the entity expects to receive with certainty.

18 Wimmer and Kusterer (2010, pp. 455–457) describe an assumption of a 100% probability of default as the prevailing practice for specific impairments (Einzelwertberichtigungen) under national German accounting requirements and for impairments booked by German banks reporting under IFRS.

19 Except for the addition in parentheses ‘(excluding future credit losses that have not been incurred)’ in IAS 39.63.

20 IAS 39.64 requires for individually assessed financial assets which are not found to be impaired that an entity includes them ‘in a group of financial assets with similar credit risk characteristics and collectively assesses them for impairment’. For sovereign debt with a good credit rating, the collective assessment will probably also not result in impairments.

21 The IASB argues in the Basis for Conclusions (IFRS 9.BCE.90–92) that the decision for the IFRS ECL impairment model as the only model for debt instruments was at the request of constituents referring in particular to the Financial Crisis Advisory Group. The report of the Financial Crisis Advisory Group (2009) indeed criticised the boards because of the existence of multiple and inconsistent impairment approaches. It did ask for more forward-looking impairments but not for a specific model.

22 Comparability of the measurement of FV-TOCI assets is not affected as it is still at fair value.

23 Bischof and Daske (2016) argue as a result of their discussion of the EU endorsement criteria that ‘the standard cannot reasonably be rejected on grounds of the IAS Regulation’.

24 See, for example, Cruces and Trebesch (2013) and Sturzenegger and Zettelmeyer (2005). Moody’s (2011, p. 13) provides a list of recovery rates.

25 See Tomz and Wright (2013, pp. 16–17).