Revisiting Constituents’ Reflections on the Incorporation of Day-one Losses into IFRS 9

ABSTRACT

IFRS 9 requires the recognition of expected credit losses from the inception of a financial instrument, resulting in so-called day-one losses. The incorporation of day-one losses caused considerable controversy among the IASB members and its constituents. With a focus on the constituents’ positions and reasoning, this study portrays the discussions held in the comment letters received by the IASB during the drafting process. We find that most constituents initially rejected day-one losses as conceptually unsound and/or as inappropriately affecting investors’ and preparers’ decision-making. Despite these continuing concerns, the majority of constituents eventually accepted day-one losses as a pragmatic approximation of expected credit losses in the absence of superior alternatives. Considering the technical and political nature of standard setting, our analysis provides insights into the constituents’ assessment of departures from the Conceptual Framework and the constituents’ views on the standard setters’ responsibilities regarding financial stability after the financial crisis.

Keywords:

• conceptual framework
• day-one losses
• expected credit loss model
• financial stability
• IFRS 9

JEL Classification:

• G01
• G21
• G23
• G28
• M41
• M48

Acknowledgements

The authors would like to thank two anonymous reviewers, Andreas Barckow, Kirstin Becker, Niclas Hellman, Christoph Kuhner†, Anne McGeachin, Stuart McLeay, Martin Messner, Araceli Mora, Christoph Pelger, Richard Pucci, Walter Schuster, and participants in the 14th Workshop on European Financial Reporting (EUFIN) in Stockholm, the 35th Annual Conference of the French Finance Association (AFFI) in Paris, the British Accounting and Finance Association 2019 Annual Conference (BAFA) in Birmingham, the 42nd Annual Congress of the European Accounting Association (EAA) in Paphos, the 81st Annual Meeting of the Verband der Hochschullehrer für Betriebswirtschaft e.V. (VHB) in Rostock, and the 8th EIASM Workshop on Accounting and Regulation in Siena as well as faculty members of the Stockholm School of Economics for valuable comments on prior versions of this paper. Furthermore, we thank the DAAD for funding a research stay at the Stockholm School of Economics in 2019 during which this paper was presented.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 For comprehensive overviews of the IFRS 9 drafting process and its underlying dynamics, see Hashim et al. (2016, 2019) and Pucci and Skærbæk (2020).

2 Furthermore, a shorter average time horizon for lending in the US (compared to European banks) might have driven the FASB’s preference for the immediate recognition of lifetime ECLs (Pucci & Skærbæk, 2020, p. 10).

3 An exception is provided by a dissenting opinion on the final US impairment standard of two FASB members, who criticised day-one losses basically for the same conceptual shortfalls as the IASB (FASB, 2016a, pp. 235–240).

4 We analysed the responses to the DP 2008 to gain a comprehensive understanding. As the IASB did not propose a specific impairment methodology in the DP, we exclude it from the numerical overviews for reasons of consistency.

5 To ensure a consistent analysis, we limit this part of the examination to those IASB proposals that incorporated day-one losses, i.e., the joint SD of 2011 and the ED of 2013.

6 The appendix shows the details of all cited CLs.

7 Beyond active supporters (explicitly referring to day-one losses), we also classify CLs as supportive of day-one losses that basically agreed with the immediate recognition of lifetime ECLs (the CECL model), the floor (SD 2011), or the 12-month allowance (ED 2013) and did not raise any explicit concerns about day-one losses.

8 Furthermore, academic institutions and users exhibited high rates of support in 2011. However, in total, these constituent groups sent only 10 CLs in 2011, which limits the informative value of these findings.

9 Notably, in 2013, many of the US respondents opposing day-one losses (six out of 14 CLs) simultaneously rejected the turn away from the incurred loss model in general.

10 In the case of commenting institutions, we control only for the institutions and not for the individuals signing the CLs.

11 Note that changes from or to support might stem from our classification system, as we also classify CLs as supportive of day-one losses if they simply agreed with the overall approach, even if they did not explicitly refer to day-one losses.

12 Each CL could have presented several arguments.

13 In the development of IFRS 9, the Conceptual Framework of 2010, which the IASB and FASB have jointly drafted, served as the theoretical basis for standard setting.

14 Please note that only five US constituents supported day-one losses in 2013, which limits the informative value of the 100% usage rate of the loss allowance adequacy argument in Table 5.

15 In contrast to constituents who accepted day-one losses with concerns, supporters of day-one losses drew on the approximation argument without expressing any reservation.

16 Please note that the alleged ‘double-counting’ does not relate to the total credit losses recognised over the life of the financial asset.

17 Note that the joint Conceptual Framework of 2010 deliberately excluded prudence as a desirable characteristic of financial reporting and emphasised the merits of neutrality for accounting information to be useful (IASB, 2010a, QC12, QC14, BC3.27–29).

18 Building on the dichotomy of ex-ante and ex-post conservatism (Mora & Walker, 2015), the feature of day-one exhibits characteristics of ex-ante conservatism because the recognition of ECLs from the outset unconditionally understates the economic value of financial assets irrespective of any triggering event (see in detail Giner & Mora, 2019, p. 739; Hashim et al., 2019, p. 709). Non-financial asset impairment, in contrast, basically follows ex-post conservatism (Barker & McGeachin, 2015, p. 194). The incurred loss model of IAS 39 was (more) consistent in this regard with the impairment requirements for non-financial assets, as it was similarly based on a threshold for recognising impairment, that is, objective evidence of a loss event, and thus also represented ex-post conservatism (Barker & McGeachin, 2015; Mora & Walker, 2015).

19 For an in-depth analysis of the interaction of IFRS 9 and prudential capital requirements, see Novotny-Farkas (2016).

20 Shortened lending horizons and higher pricing of higher-risk assets and very long-term financing were also mentioned by some banks, participating in a field test study on IFRS 9 conducted by EFRAG, as potentially arising negative consequences (EFRAG, 2015, par. 14).

21 In this regard, some opposing respondents noted that the financial assets typically held by non-bank institutions did not cause the financial crisis (IASB, 2010b, par. 31(a)).

22 For example, CL 19c, 2013, p. 1; CL 45b, 2013, pp. 3–4; CL 183, 2013, p. 8.

23 For example, CL 123, 2013, p. 4.

24 For example, CL 41, 2011, pp. 1–2; CL 138, 2013, p. 7.

25 For example, CL 123, 2013, p. 4.

26 Other constituents, in turn, criticised the 12-month period introduced by the IASB as an ‘arbitrary’ proxy (CL 91, 2013, p. 2; CL 133, 2013, p. 2) lacking any conceptual foundation apart from the fact that it broadly aligns with the existent Basel regulatory requirements (IASB, 2014c, BCE.152).