As post-2008 accounting rules face their first real test amid the coronavirus pandemic, banks need to be transparent about their expected credit loss models to help the market understand how well they are faring, according to analysts.
The crisis triggered a spike in first-quarter loan loss provisions across the board, but amounts diverged widely between European banks. Among a sample of the largest European banks by assets, Spain's Banco Santander SA recorded provisions of €3.93 billion, while Finland-based Nordea recorded just €154.0 million.
Lack of comparability
Although an "enormous" amount of work has gone into comparing first-quarter loan loss charges, it is still very difficult to tell whether a large number indicates that one bank is being more cautious or that its loan portfolio is just worse than its peers', UBS equity analyst Jason Napier told journalists in a May 28 conference call.
This lack of comparability is a key reason why the cost of equity that investors expect for banks will remain high, he said. To bolster investor confidence in the sector as a whole, there needs to be a conversation about a potential standardization of banks' disclosure practices after the current downturn, Napier said.
Banks themselves are still getting to grips with their expected credit loss, or ECL, modeling given the high level of uncertainty regarding the full economic impact of COVID-19 and the shape of the post-crisis recovery. But the more the market knows about the assumptions and methodologies they apply to estimate ECLs, the better it can assess their performance, according to Dierk Brandenburg, head of the financial institutions group at Scope Ratings.
It is important that banks disclose the GDP and unemployment forecasts that go into their ECL models, Brandenburg said in an emailed comment.
"This will provide the necessary transparency and allow for appropriate market discipline," he said.
European banks' first-quarter disclosures of macroeconomic assumptions were patchy, with only a handful giving full-year guidance for ECL provisions or the amount of loans at significant risk of default, which further complicates the analysis, according to Brandenburg.
Regulators and standard-setters should, therefore, push for improvements to the disclosure of banks' ECL assumptions, he said.
Among Nordic banks, differing macro assumptions have led to widely different levels of provisions.
Meanwhile, representatives of banks and regulators in Poland are establishing a panel to develop a common approach to loan loss provisions and banks there want the central bank to publish a macroeconomic forecast they can use, according to local newspaper Rzeczpospolita.
IFRS 9 challenge
Since the implementation of the International Financial Reporting Standard, IFRS 9, on Jan. 1, 2018, European banks have been required to provision for expected rather than incurred losses. Although this facilitates the early recognition of loan impairments and therefore improves financial stability, IFRS 9 has been criticized for amplifying procyclical effects.
Implemented on a straightforward basis, the standard would produce significant loan losses very early on because all loans coming under stress would have to be classified as "Stage 2" and would require lifetime ECL provisions. This would make it harder for banks to generate the revenue to absorb those charges, Napier said.
Under IFRS 9, loans classified as Stage 1 are performing and Stage 3 are nonperforming. Loans considered to be at significant risk of default are moved into Stage 2 from Stage 1.
To mitigate the negative effect of COVID-19, European regulators advised banks to take a long-term view of loan performance and use the flexibility provided by the standard to carefully assess which loans have a good chance of quickly recovering from the initial, short-term shock.
Comparability has also been raised as a potential issue with IFRS 9 before. Depending on how each bank decides to tackle "the non-linear relationship between ECL and forward-looking information such as macroeconomic scenarios," comparability across entities may be "significantly jeopardized," BBVA analysts warned ahead of the IFRS 9 implementation in late 2017.
No flaw in standard
But according to accounting experts, comparability issues do not necessarily mean the IFRS 9 standard is flawed.
If banks adequately disclosed the macroeconomic assumptions and weightings they are using, the risk appetite and lending methodologies they are deploying in different portfolios and therefore the policy choices they are making, this would improve comparability across institutions, Damian Hales, a partner and U.K. IFRS 9 target operating model lead at Deloitte said in an interview.
Banks' loan losses will vary due to different operating environments and lending profiles, but comparability could be improved if banks present the information in a way that investors can understand, Tom Millar, a partner in Deloitte's banking and capital markets audit practice, also told S&P Global Market Intelligence.
Although the economy may bounce back quickly after the initial COVID-19 shock, banks see the current crisis as another 10-year event, Hales and Millar said. The process of working through the bank assets affected by the pandemic will likely take seven to eight years, similar to the aftermath of the 2008 financial crisis, Millar said.