Bank loss provisioning rules: a convenient scapegoat in the Covid-19 crisis?
Better coordination between prudential regulators and accounting standard-setters will help banks withstand the crisis, writes Lucas Mahieux.
Accounting and the Covid-19 crisis
As a consequence of the upcoming economic crisis triggered by the Covid-19 situation, US and European banks are expected to book more than $50bn of charges on bad loans in the first quarter of 2020. The accounting rules that dictate the way banks provision for loan losses are crucial because they directly affect the ability of banks to continue lending to the real economy. Importantly, new accounting rules for the provisions of credit losses were introduced in the wake of the global financial crisis, and those new rules are now being blamed for aggravating the current situation. Why do provisioning rules for banks play such an important role?
New provisioning rules
Every time there is a major crisis, accounting rules for banks are at the heart of the debate. During the 2007-08 financial crisis, the accounting rules for the provision of loan losses were criticised for increasing pro-cyclicality and thereby exacerbating the crisis. Those rules relied on the concept of incurred loss and were blamed for excessively delaying the recognition by lenders of credit losses. As a result, the US Financial Accounting Standard Board and the International Accounting Standard Board recently introduced new provisioning rules: the Current Expected Credit Loss standard (CECL) in the US and IFRS 9 in the rest of the world. Those new standards use the concept of expected credit loss: financial institutions are require to provision for expected future credit losses using all information available. The coronavirus crisis challenges this new concept of expected credit loss.
While standard-setters argue that the new accounting rules provide more timely information to stakeholders, financial institutions have heavily criticised the concept of expected credit loss. Most notably, banks are concerned that the forecasts of future credit losses are unreliable and inappropriate loss recognition may lower bank capital ratios. Given those arguments and the emergence of the coronavirus crisis, the US Congress decided to provide via the CARES Act an option for lenders to delay the implementation of CECL until 31 December 2020 or the end of the coronavirus national emergency, whichever comes first. Given that CECL is barely in its first year, it was easier to roll it back, whereas IFRS 9 came into force in 2018 and is more established. As a result, banks outside the US have no choice but to comply with IFRS 9.
Insights from my research
In a recent working paper with Haresh Sapra at the University of Chicago and Gaoqing Zhang at the University of Minnesota, we shed some light on this debate and assess the desirability of the new expected loss provisioning models. To that end, we develop an economic model to examine the impact of moving from an incurred loss provisioning model to an expected loss provisioning model on the banks’ behaviours and on their capacity to lend to the economy. Our results provide two interesting insights that may be useful to both accounting standard-setters and banking regulators.
First, our research highlights the crucial role of expected loss models for prudential regulators and banks’ stakeholders. In particular, we show that timely loss recognition under an expected loss model is beneficial because it allows regulators to intervene early and prevent banks from engaging in excessive risk-taking. More importantly, we show that this benefit always dominates the potential false-alarm costs caused by the imprecise information inherent in the expected loss model. The reason is that a rational regulator fully internalises the imprecision of the expected loss models and always takes the action that results in the highest surplus.
Second, bankers fear that the application of the new expected loss models during the Covid-19 pandemic may lead to a sudden significant increase in credit loss provisions, which would result in an erosion of banks’ capital. Nevertheless, such concerns are only valid if prudential regulators set banks’ capital requirements independently of the accounting standards used to provision for loan losses. Our paper shows that capital ratios and loan loss models should be jointly determined: if banks change the estimation of credit losses, then banking regulators should also adjust banks’ capital requirements. In another working paper with Jeremy Bertomeu at University of California San Diego and Haresh Sapra, we show that accounting measurement and capital requirements are indeed complementary tools that affect the capacity of banks to lend to the real economy. We therefore call for a better coordination of prudential regulators and accounting standard-setters.
A key policy implication of my research is that changing the provisioning rules for estimating credit losses requires prudential regulators to adjust capital adequacy ratios at the same time. In particular, my coauthors and I show that relative to capital requirements under the previous incurred loss model, capital requirements under the new expected loss model would be looser when the precision of estimating early credit losses is relatively high and/or when banks’ risk-taking incentives are not too severe. Looser capital requirements will in turn spur lending by banks. Hence, the new accounting rules may actually benefit lenders as well as the whole economy. Consistent with our predictions, most of the publicly traded US banks did not use the option to delay the implementation of CECL in the first quarter of 2020.
Recent regulatory actions and conclusion
Policymakers in the US and in Europe have acknowledged the importance of the connection between accounting rules and capital requirements. It is noteworthy that the insight from my research on the complementarity between accounting rules and capital ratios is consistent with the recent actions taken by prudential regulators. Indeed, in the US, the banks who decide to comply with CECL benefit from temporary capital relief by the US banking regulators. Similarly, in Europe, banks do not have the option to delay the implementation of IFRS 9 but are likely to face looser capital requirements in the next few years.
In a nutshell, while my research shows that policymakers and accounting standard-setters should refrain from changing the expected loss standards, it provides a rationale for the banking regulators’ decisions to loosen banks’ capital ratios. Banks are key players in the economy as they are in charge of providing credit to firms and households. The drastic increase in loss provisions by banks is a sign of the economic pain to come. However, a better coordination between prudential regulators and accounting standard-setters will help banks to withstand this coronavirus crisis, which will in turn benefit the world economy.