Regime change is rarely pretty to watch – even in the staid world of accounting. The ongoing shift to an expected credit loss framework is causing consternation among banks worried about the impact of higher loan loss reserves on capital.
The concerns are twofold. The first has to do with ambiguity over what constitutes a valid forecast for loss provisions. The other has to do with how the US’s implementation of the new standards align with the Fed’s stress-testing rules.
Both IFRS 9 and its US counterpart, the Current Expected Credit Loss regime, require estimates of lifetime losses be based on “reasonable and supportable” information, but there is no agreement as to what those words actually mean. Depending on their interpretation, banks may take greater or lesser provisions against lifetime expected losses – something which will have a direct impact on their capital.
The problem as always with allowing firms an element of judgement is that self-interest can creep in. Banks will surely choose the course of action that works in their favour – and who could blame them?
Let’s say a bank’s forecasts show a downturn in economic conditions two years hence. The bank may deem it expedient to declare that its “reasonable and supportable” forecast period only extends to 18 months. Hey presto, the hit on loan loss provisions is mitigated.
The amount of discretion available to banks is even greater under CECL, which allows firms to revert to historical information to estimate lifetime losses for periods that fall outside the “reasonable and supportable” area. With greater discretion comes greater room for interpretation and subjectivity.
All of this puts the onus on external auditors to scrutinise a bank’s credit risk models to make sure the firm is not “shifting the goalposts”, as one expert puts it. Failure here could serve to undermine the credibility of what amounts to a global step-change in accounting practices for financial institutions.
Banks are irked these two regulatory regimes seem to be pulling them in different directions with regards to calculating losses
In the US, banks are also stressing about the regime’s impact on stress testing. CECL requires banks to set aside provisions to cover potential losses on all loans using a bank’s own macroeconomic forecasts and information about past loss history. The aim is to arrive at a realistic view of future losses. The Federal Reserve’s Comprehensive Capital Analysis and Review (CCAR) stress test, on the other hand, requires banks to estimate losses based on supervisory scenarios that are typically more severe than the banks’ own forecasts. The scenarios are deliberately unrealistic because they are meant to cover unexpected losses. This means the amount of capital banks would need to set aside to cover loan losses under CCAR could be significantly higher than the amount projected under CECL.
Under CECL, each bank will revise its loan loss estimates based on the data available at the time. If, for example, it lends to the energy sector, and a geopolitical event occurs that affects oil and gas prices, it will have to revise its estimates upwards or downwards. Under CCAR, a firm will need to make these estimates for each of the nine quarters in the CCAR cycle, based solely on the information they have at the beginning of the cycle. Therefore, there’s a high likelihood that their accounting estimates for CECL will deviate from the capital estimates for CCAR.
Banks are irked these two regulatory regimes seem to be pulling them in different directions with regards to calculating losses. They are also hoping authorities will recognise some overlap between the two regimes and issue further guidance.
Although the Fed has stated it will provide guidance on CECL for the 2020 CCAR, a solution needs to be found quickly in order to give banks time to reconfigure their CCAR processes. Watch this space.
The week on Risk.net, September 8-14, 2018Receive this by email