The International Accounting Standards Board (IASB) finally signed off on its marathon IFRS 9 project in July 2014. The release of the new International Financial Reporting Standard (IFSR) for financial instruments marks the final milestone in the board’s ponderous response to the global financial crisis. IFRS 9 unites all aspects of the financial-instruments accounting – classification and measurement, impairment and hedge accounting – under one roof. And from January 2018 it will replace the IASB’s existing rule book, IAS 39, on the topic.

Although IFRS 9 introduces a new classification and measurement approach, as well as a new model for hedge accounting, the real change under the hood is the standard’s impairment model. In contrast to IAS 39, IFRS 9 requires entities to estimate and book expected credit losses on amortised-cost financial instruments upfront at stage one of a new three-stage impairment method.

The IASB hopes that this significant shift will address the problem of banks delaying recognition of credit losses until the last possible moment. Indeed, writing in an investor-focused newsletter, IASB vice-chairwoman Sue Lloyd, the so-called Mother of IFRS 9, argued: “These changes, in the timing of recognition and the consideration of reasonable and supportable forward-looking information, are important changes from existing IFRS, which only allowed impairment losses to be recognised when a loss had been ‘incurred’. 

“Even then, only the effect of events that had already occurred could be considered in measuring those impairment losses. The new requirements should help to address concerns by many investors about the recognition of impairment being ‘too little, too late’.”

Where credit risk has deteriorated significantly, the new standard forces preparers to recognise full lifetime losses at stage two of the new model. The holder of the instrument would, however, continue to recognise interest revenue calculated on the effective interest rate on the gross carrying amount. And, if an asset is also credit impaired, it moves into a new stage-three bucket where the entity must recognise not only full lifetime losses but can only book interest revenue on the gross carrying amount, less the loss allowance.

As for how entities will calculate that loan-loss provision, recent developments suggest that European financial institutions are endeavouring to align IFRS 9’s requirements with regulatory technical standards such as the Bank Recovery and Resolution Directive.

IFRS 9 Timeline

July 2014: IASB signs off IFRS 9 project.

November 2016: EU formally adopts IFRS 9.

January 2018: IFRS 9 replaces IASB’s existing rule book.

The early hints suggested that there were two methods in the running. One was to take the lifetime probability of default (PD) and scale it to arrive at a proxy for full lifetime equivalents. The other method is the expected cash-flow approach, where entities estimate cash flows (and therefore implied losses) over time.

Posting to his blog on the topic recently, PwC valuations partner Atul Karir writes: “In theory, either of these two approaches arguably could be used as a starting point. To estimate hold value, lifetime losses would need to be estimated for all loans, that is, not just loans that have shown a significant increase in credit risk, but performing loans too.

“However, when it comes to estimating exit value, interest-rate risk will need to be reflected for the fixed-rate portion of the book. This presents a problem for those following the lifetime PD approach as that approach does not have cash flows.”

The blog continues: “It’s important to bear in mind that the [Regulatory Technical Standard] explicitly states that the valuations should estimate cash flows, and there is good reason for this. Crucially, the expected cash flows approach provides a view on the timing of future losses. This is critical in assessing the viability of the restructured bank post-resolution as it impacts how quickly capital levels will be depleted by impairments as the bank ramps up its profitable activities.”

Karir concludes that, in recent months, practitioners in his firm have seen a shift away from a lifetime scaler approach, “with banks focussing on modelling losses over time and under varying macroeconomic scenarios”.

Back in Brussels, the EU formally adopted IFRS 9 in November 2016. This endorsement came despite rumblings from some quarters that the European Parliament would block the standard. In fact, as IPE reported in September 2015, IFRS 9 endorsement was a done deal. 

“If a regulatory system purports to be regulating banks on a going-concern basis, then there can be no merit in the pretence that losses will be borne by creditors”

Tim Bush

During the course of 2015, well-placed sources told IPE they do not expect the Parliament to block IFRS 9 – despite recent sabre rattling. Speaking on condition of anonymity, the sources explained that, because IFRS 9 was a delegated act under EU procedure, the Parliament was unable to amend it. Instead, it could either accept it or veto it.

And although the European Parliament has shifted from being a consultative assembly to a co-legislator within the EU, a veto was never realistically on the cards. Instead, IPE reported that the Parliament would, in all likelihood, vote on a non-binding own-initiative resolution of its ECON Committee that could be highly critical of both the IFRS Foundation and the IASB’s standards.

That resolution – also supporting the adoption for IFRS 9 – duly emerged in September 2016. It also called on the European Commission (EC) to examine the possibility of introducing a phase-in regime for the impairment requirements of IFRS 9 to avoid any sudden unwarranted impact on institutions’ capital ratios and lending. In turn, the EC responded in November 2016 as part of its CRR II/CRD V proposals with transition proposals.

Warning from the European Banking Authority on EC approach

But, it recently emerged, the new standard’s woes were far from over. The European Banking Authority (EBA) has now stepped into the ring to warn that the EC’s transition approach could mean banks recognise lower loan-loss provisions or impairments than at present. The effect of any such move, the EBA warns, is to unpick any of the incremental improvements brought about by IFRS 9 in relation to impairment.

More damningly, the EBA has argued that the EC’s proposal “as it currently stands, could be interpreted as allowing institutions to add back [expected credit losses] in stage three under IFRS 9”, which would “result in the neutralisation” of the provisioning now in place under IAS 39.

Worse still, the EBA argued: “However, if an institution decided to apply the transitional arrangements, it would be able to add provisions

back to [tier-one provisions] and therefore have a positive impact due to IFRS 9.”

Someone else who remains unconvinced of IFRS 9’s benefits is Tim Bush, head of governance and financial analysis at Pensions & Investment Research Consultants (PIRC). A long-standing IFRS sceptic, he recently argued in evidence to the UK Parliament’s Treasury Select Committee: “Given that IFRS 9 itself is already defective, due, in my opinion, to regulatory capture, the EBA paper is revelatory in terms of what it says about the approach of the EU Commission, to allow banks the discretion to reverse out the impact on bank capital numbers.

“The EBA paper merits reading in its entirety. However, I will state my observation that the Commission proposing that some IFRS 9 losses being ‘taken to Tier 2 capital’ [that is, subordinated debt] is repeating the same conceptual flaw as Basle/IAS 39. Losses, in reality, are borne 100% by shareholders (on a going-concern basis), unless the bank is in run off (gone concern).

“If a regulatory system purports to be regulating banks on a going-concern basis, then there can be no merit in the pretence that losses will be borne by creditors. Losses only reach creditors once they have been borne by shareholders.”