Calculating expected lending losses and facing unexpected costs

The shift from making provision for losses suffered in the past (the incurred credit loss (ICL) model) to those likely to occur in the future (the expected credit loss (ECL) model), was intended to make banks safer by encouraging counter-cyclical, rather than pro-cyclical, behaviour. 

This switch to ECL was born of the 2008 financial crisis when its forerunner, ICL, was in full swing. The ICL rules effectively mandated banks to turn off lending, deepening the crisis. But despite policy-makers’ good intentions, the introduction of ECL in 2018 brought unintended consequences with implications for banks and the economy.

Specifically, it has become far more difficult for small and medium-sized enterprises (SMEs) to access bank lending. My research shows that affected banks’ SME loan holdings fell by 15–20% since ECL was introduced.

I have explored four possible explanations. First, rising loan-loss provisions reduce current earnings, making SME lending less attractive for banks. Second, there is a greater administrative burden in creating provision for all SME loans, many of which are risky and opaque in nature, which adds additional costs. Third and fourth potential explanations are that rising loan-loss allowances could dent banks’ regulatory capital, and that ECL could have spurred on banks to develop a better understanding of their lending practices, which may have resulted in lower levels of SME lending.

For SMEs, the problems are compounded by the paucity of alternative credit, as banks are the primary source of external finance.

While my research found neither bank shareholders nor the overall economy suffered directly from the introduction of ECL, it’s essential to recognise the importance of SMEs to society. In Europe and the US, SMEs represent more than 99% of non-financial companies, employing more than two-thirds of the non-financial workforce and generating more than half of total gross value added. So, what happens to SMEs matters greatly to us all.

Financial reporting

The adoption of the ECL has practical implications for banks, not least the effect on financial reporting.

While most of the world’s regulators gave banks a phase-in period to address the regulatory capital implications of ECL provisioning, the relief was temporary at best, leaving banks with much work ahead of them.

Banks should not assume that expected loss calculations for regulatory reporting will be sufficient for financial reporting. Allowance computations under Basel and International Financial Reporting Standards exhibit dramatic differences in terms of model inputs, assumptions and estimation horizons.

Similarly, interpreting the potential impact of ECL-based loan portfolios on banks’ regulatory risk-weight calculations has been an important lesson. In the past, regulators raised concerns about risk-weighting, as banks’ reassessment of their lending riskiness would be likely to push up risk-weighted assets.

All this is particularly important right now. We are increasingly able to judge the outcome of ECL reporting in terms of the growing volatility apparent in income statements and balance sheets. The events of 2020 have greatly magnified this, making bank resilience seem much weaker than would have been the case under the previous system of incurred credit loss reporting.

Peaks and troughs

Over the past six quarters, banks have reported massive peaks and troughs in their loan-loss allowances due to the rapidly changing economic, financial and political outlook.

Of course, rule-makers should appreciate that change is rarely cost-free. But who should be calculating and determining the right amount of bank lending and risk in the economy? Similarly, who can ensure that policy-makers possess the correct volume of timely data to make such judgements?

These are essential questions, which lead to a further conundrum. Does ECL reporting deliver what it’s meant to? In other words, does it help banks to fix the roof when the sun is shining so they can withstand a downpour when it comes?

The events of 2020 may not have provided ideal conditions in which to test this proposition, as the impact of Covid-19 was not so much a rainstorm but a once-in-a-lifetime blizzard.

Given the extraordinary circumstances, requiring banks to take into account all likely future problems may have led us to believe that the cure was worse than the disease.

So how best to navigate the ECL landscape? Managers should continue identifying best practices and invest in infrastructure and people. Loan-loss calculation under ECL is a journey, not a destination.

And bank investors — whether retail or institutional — should learn to live with ECL-based provisions. It may be tempting, and easy, to rely on pre-provision earnings, but this could prove recklessly short-sighted. Fundamentals such as loan portfolio quality must not be ignored. Looking ahead, bank investors should scrutinise ECL footnotes in bank financial reports to obtain necessary detail about the breakdown of bank portfolio structure. Forewarned is forearmed.

Aytekin Ertan is assistant professor of accounting at London Business School.