latest-news-headlines Market Intelligence /marketintelligence/en/news-insights/latest-news-headlines/covid-19-uncertainty-turns-banks-loan-loss-provisioning-into-guesswork-57853787 content esgSubNav
In This List

Coronavirus uncertainty turns banks' loan loss provisioning into guesswork

Blog

Banking Essentials Newsletter: July 10th Edition

Blog

Enabling Growth in the EMEA Loan Market

Podcast

Masters of Risk | Episode 12 - Women's Panel

Blog

Banking Essentials Newsletter: June 26th Edition


Coronavirus uncertainty turns banks' loan loss provisioning into guesswork

Expected loss estimates under the European IFRS 9 accounting regime will be more guesswork than model-driven prediction as no existing approach can gauge the impact of the new coronavirus pandemic on the economy.

Although banks can use the flexibility provided in the accounting standard to limit the impairments they would have to recognize on loans that come under COVID-19-related stress, the economic outlook plays a significant role in these estimates. Banks will have to tap in the dark for a big part of their loss modeling as economists and market observers are still only theorizing about the pandemic's hit to the real economy.

READ MORE: Sign up for our weekly coronavirus newsletter here, and read our latest coverage on the crisis here.

ECL arithmetic

The new International Financial Reporting Standards regime, dubbed IFRS 9, changed the way banks provision for credit losses, requiring capital to be set aside for expected rather than already incurred losses. Banks criticized the expected credit loss, or ECL, model before IFRS 9 came into force Jan. 1, 2018, over concerns it would enhance procyclicality as it requires higher provisions in a downturn and lower provisions in good times.

SNL Image

SNL Image

As the COVID-19 pandemic spread, banks called on regulators for relief regarding IFRS 9 loan loss provision reporting. Regulators responded by advising lenders to carefully assess the short- and long-term effects of the crisis before raising their provisions.

The European Securities and Markets Authority, or ESMA, said issuers should take into account the impact of economic support packages and banks' own forbearance measures provided to borrowers before they assign a higher risk to financial instruments in their portfolios.

Under IFRS 9, if a loan has experienced a significant increase in credit risk since initial recognition, banks have to switch provisioning for that loan to a lifetime ECL, or stage 2, from a 12-month ECL, or stage 1. By making the switch the bank assumes there is a heightened probability of default throughout the loan's whole lifetime instead of only within the next 12 months.

SNL Image

SNL Image

If banks decide that the effects of the pandemic on borrowers' ability to repay will be limited in the long term, they do not automatically need to move to a lifetime ECL classification, according to Osman Sattar, accounting specialist and director EMEA financial institutions at S&P Global Ratings.

Even if there is a significant increase in credit risk for some loans and banks have to recognize lifetime ECLs, provisions for these lifetime losses might be lower if government guarantees on the exposures are taken into account, as ESMA has advised, Sattar said in an interview.

COVID-19 fits no model

The Basel Committee on Banking Supervision said April 3 that it expects lenders to continue to apply the relevant frameworks.

"Banks should use the flexibility inherent in these frameworks to take account of the mitigating effect of the extraordinary support measures related to COVID-19," said the committee, which oversees global prudential regulation of banks.

The Bank of England said March 26 that ECL estimates should be based on "the most robust, reasonable and supportable assumptions possible in the current environment" as this will help reduce the risk of a significant overstatement of ECLs that could lead to an unnecessary tightening in credit conditions.

Even if such assumptions are applied, the BoE said "any changes made to ECL to estimate the overall impact of COVID-19 will be subject to very high levels of uncertainty as so little reasonable and supportable forward-looking information is currently available on which to base those changes."

COVID-19 presents a unique challenge to future loss modeling because there are still more questions than answers about the impact of the outbreak on the global economy. "No forward-looking model could have predicted the features of the COVID-19 medical situation and its implications," Deloitte risk advisory experts wrote in a March 20 blog post.

The existing models for the recognition of expected losses on financial assets take into account the interaction between macro-economic indicators and expected loss prediction elements, they said. However, to assess the economic fallout of COVID-19 there first needs to be clarity about when the lockdowns will be lifted and business activity can resume, where production will be positioned and at what pace will consumer demand recover, they said. The pace of recovery in certain sectors will be slower and others will need to completely transform as a result of the health crisis, Deloitte said.

Guesswork

Economists are still guessing what shape the post-COVID-19 economic recovery will take and what the effect of the mitigating measures undertaken by regulators and governments will be. In this environment, it is hard to gauge the full blow to banks' balance sheets.

Although nonperforming loan levels are likely to increase post-COVID-19 given expected company failures, it is too early to predict the size of that increase, according to Sattar. Bad debt levels will be higher if the economic stress lasts longer or the recovery is weaker than expected, Alexandre Birry, head of financial services analytics and research at S&P Global Ratings, said in a webinar March 27.

Currently, Ratings assumes a strong economic rebound in Europe in the latter half of 2020 and into 2021, which suggests banks' short-term forbearance activity now may limit credit losses later.