Banks in pandemic turmoil

As the COVID-19 spread and policy reactions have disrupted markets, bankers on both sides of the Atlantic have called for relaxation of accounting standards introduced in the wake of the Great Financial Crisis, known as expected credit loss provisioning. These calls, like much bank lobbying on capital regulation, should be ignored by public authorities and accounting standard-setters. There is no perfect accounting thermometer for credit risk in banks’ loan books, but breaking the current thermometer in the midst of a crisis would do far more harm than good. 

Since there are two main sets of accounting standards in the world, the debate on expected credit loss provisioning is actually two different debates echoing each other. In the United States, accounting standards are set by the US Financial Accounting Standards Board (FASB), a non-profit body under oversight by the US Security and Exchange Commission. The relevant FASB standard is Accounting Standards Update (ASU) 2016-13, “Measurement of Credit Losses on Financial Instruments” (The credit loss component is also referred to in US accounting discussions as “Topic 326” or “ASC 326,” where ASC stands for Accounting Standards Codification.). ASU 2016-13 just entered into force for large listed banks, since it is to be applied on financial statements starting on or after 15 December 2019. Following a further update in November 2019 by the FASB, the corresponding date for smaller banks (all large US banks being publicly listed) is 15 December 2022. In the rest of the world, most large banks use the International Financial Reporting Standards (IFRS) set by the International Accounting Standards Board (IASB), a global standard-setting body hosted by the non-profit IFRS Foundation. The relevant IFRS standard is IFRS 9 on Financial Instruments, issued by the IASB in November 2013 and endorsed three years later by the European Union, among other jurisdictions. IFRS 9 has been implemented for some time since it became effective for annual periods starting on or after 1 January 2018. Whereas IFRS 9 and ASU 2016-13 are not identical, both are variations of the same principle of expected loss provisioning. 

The principle’s adoption by FASB and the IASB came in the early 2010s, in response to prodding from public authorities through the Financial Stability Forum (now the Financial Stability Board, FSB) as early as March 2009 while the Great Financial Crisis was raging. The feeling at that time among central banks and financial ministries was that the prior established method, known as incurred loss provisioning, was leading to excessive procyclical effects when losses were indeed incurred. Better instead, the concept went, to book a provision early on, as soon as the loss is foreseeable, even if no repayment has been missed yet. There was controversy from the start on whether that might lead to a different but equally problematic pattern of procyclicality. After several years of debates, however, the two standard-setters obliged, and research published in 2017 by the Bank for International Settlements (which hosts the FSB Secretariat in Basel) concluded that this had indeed been the right thing to do. 

Predictably, however, the banks were never enthusiastic about having to book losses earlier than under the previous methods, and lobbied heavily against it on both sides of the Atlantic. Such lobbying has been predictably revived by the COVID-19 shock, together with calls for more general suspension of credit loss provisioning – in other words, the recognition that loans are becoming bad, namely non-performing loans (NPLs). These calls have been relayed by various political and institutional figures, including at the EU level the President of the European Economic and Social Committee, an advisory body. In the US, the Federal Deposit Insurance Corporation (FDIC) has endorsed them in a letter to FASB on 19 March, which immediately received support from the Conference of State Bank Supervisors. In its letter, the FDIC suggests both further delaying the implementation date of ASU 2016-13 for smaller banks, and giving larger banks an “option to postpone [its] implementation”. The other major federal bank regulators, the Federal Reserve and the Office of the Comptroller of the Currency, appear not to have echoed this call, at least in the public sphere – worth noting as other recent statements on prudential policy have been made jointly by all federal regulators, as is customary. 

Such calls should not be heeded. By mid-March in both the euro area and the United States, banks have been granted very significant capital relief – namely, they can let losses eat into their capital buffers as these have been built up significantly over the last decade in application of the global prudential accord known as Basel III. This welcome action implies that banks have a considerable capacity to absorb losses in the near future, without being in breach of their regulatory and supervisory capital requirements. The ECB has estimated the corresponding leeway for the core capital measurement at €120 billion, a massive amount. No miracles will (or in fact, should) happen in terms of credit expansion, given the parlous economy, but credit will not unduly contract as an effect of procyclical regulatory constraints, and financial stability is protected. Those decisions were made swiftly – not least in the euro area, where ECB Banking Supervision has acted at a pace that would have been impossible in the pre-banking union era given challenges of coordination and stigma effects. Banking union remains dangerously unfinished and has not achieved its stated aim of breaking the bank-sovereign vicious circle, but even in its halfway-house status this episode demonstrates its tangible benefits. 

Because the capital relief allows banks to absorb significant losses, changing the loan loss provisioning method now is unnecessary. It is also undesirable, because there is significant value in reassuring investors and the broader public that banks are not allowed to hide bad news and move stealthily towards “zombie” status. At a more basic political level, this is not the time to give the banks special favours if these can be avoided. Longstanding experience suggests that supervisory forbearance – namely, allowing banks to pretend they meet regulatory requirements when from an economic standpoint they don’t – should only be considered as a last resort. It is not justified under the present circumstances (Loan forbearance by the banks themselves, namely letting borrowers miss scheduled payments without triggering collateral execution, is a separate matter.). 

In the euro area, there is an additional dimension given the ongoing discussion about cross-border risk-sharing and solidarity. While political difficulties abound, there is willingness to share some of the burden of fighting the pandemic and preventing economic collapse – but that cannot be expected to extend to bank rescues. Italy is at the centre of current concerns, both because of its high sovereign indebtedness and its tragic exposure to the pandemic. The Italian Banking Association has lobbied for supervisory forbearance. Italy’s true priorities do not lie there. 

The ECB has complemented its March 12 capital relief decision with additional guidance on March 20, allowing the banks to provision as few losses as possible within the constraints of IFRS 9 and providing favourable interpretations on related matters, eg, not classifying loans in arrears as NPLs if they are covered by a government guarantee. Simultaneously, the ECB has made it explicit that it rejects “[supervisory] forbearance for NPLs” and that “[i]t remains crucial, in times of distress, to continue identifying and reporting asset quality deterioration and the build-up of NPLs in accordance with the existing rules, so as to maintain a clear and accurate picture of risks in the banking sector”. In the UK, the Bank of England has taken a similar stance. In the US, the FASB should resist the FDIC’s pressure to exempt large banks from implementation of ASU 2016-13. Smaller banks are a less critical concern, with potential space for compromise on implementation schedules. 

The last few days have spectacularly illustrated the ancient wisdom that bank lobbying on capital and accounting should generally be resisted in the public interest. Banks had lobbied tooth and nail against Basel III, its multiple buffers, and related constraints on capital, leverage, and liquidity. But thank God for Basel III: Having these buffers in place is precisely why authorities were able to provide a credible response (so far) to concerns about banking sector stability in the pandemic crisis. Authorities should implement the accord’s remaining items in due time, and steadfastly resist appeals for undue supervisory forbearance. The banking system is critical to society and requires attention and support. In doing so, however, tough love is preferable to complacency.