The global financial crisis generated several criticisms of the accounting standards that guided the financial statements of the period. Two of the most salient criticisms were the undue reliance on fair value and the failure of existing frameworks to create sufficient credit loss buffers to absorb the losses arising from the crisis.
Both the International Accounting Standards Board (IASB), the authors of IFRS, and the Financial Accounting Standards Board (FASB), the authors of US GAAP, undertook to jointly review the accounting for financial instruments. Initially, this was part of their convergence program, but failure to agree on key components resulted in the parties developing divergent approaches to the issues of financial instrument recognition and measurement, impairment and hedging.
This article will focus on the key issue of the changes in the accounting for impairment for financial instruments as, while the details differ, both standards share a common objective of bringing forward the recognition of potential impairment losses in an entity’s holding of financial instruments. While no accounting framework could have created sufficient loan loss buffers to absorb the exceptional level of credit losses arising from the GFS, the experience did illustrate the weaknesses in the existing frameworks. This justifies the revision of the previous impairment framework.
While the intention of the review of impairment provisioning is to advance the timing of creation of provisioning, one must consider their net effect in conjunction with the changes in recognition and measurement criteria. Despite the criticism of the role of fair value in the severity of the global financial crisis, the revision of IFRS 9 now makes fair value through profit and loss the default criteria, replacing available for sale. The net effect may be to increase rather than reduce the role of fair value in financial instrument valuation, thus reducing the level of impaired provisions required. The net effect will not be apparent until entities start reporting under the new IFRS 9 that becomes compulsory as from Jan. 1, 2018. Within US GAAP, US ASU 2016-01, the modification to the recognition and measurement criteria broadly retain the framework existing at the time of the financial crisis.
Under both US GAAP and IFRS, the main change in impairment loss reporting has been a move from an incurred loss (backward looking) to an expected loss (forward looking) perspective. This change should have a material impact on banks operations and governance frameworks.
Similar objectives, different details.
While both sets of standards adopt a prospective loss approach, important differences exist between them. The biggest difference is in the timing of credit risk recognition. The revised US GAAP uses a current expected credit loss (CECL) based on the lifetime default risk to accrue the expected loss. It applies this test from day one of origination. The IFRS 9 expected credit loss (ECL) approach adopted by IFRS uses a three-stage approach to recognizing credit risk losses. This starts with a 12-month expected credit loss on asset origination that moves to the lifetime expected credit loss approach on a “significant increase in credit risk.” The third stage involves recognition of an impairment event and roughly equates to the existing incurred loss of IAS 39.
The two standards assume different minimum level of initial credit loss recognitions. US GAAP allows an initial recognition of zero for certain high-grade securities, while IFRS 9 expects the initial 12-month expected credit loss to be greater than zero. The scope of the US GAAP excludes debt securities classes as available for sale and is thus narrower than the IFRS 9 span that includes the AFS equivalent of fair value through other comprehensive income (FVOCI). Also, the two standards apply different treatments for purchased or originated credit impaired assets.
After 2018, these differences will reduce the comparability of financial statements produced under the two frameworks. The level of expected impairment losses that each generates will differ, and the progress of provisions through the credit cycle will vary. However, these differences should not be seen to disguise the common issues that face financial institutions in their implementation of these new frameworks.
The main issue required by both standards is that financial institutions must now have formal systems that model and calculate expected credit losses over the life of its financial assets, both on and off balance sheet. These systems cannot rely solely on historic data but are expected to include the results of the asset portfolios performance under different economic scenarios. While the standards make allowance for different levels of a bank’s sophistication and caveat the standards with remarks such as “without undue effort and expense,” the clear expectation is that existing credit risk management practices will not suffice beyond implementation.
This is requiring material efforts to create systems and processes that meet the new credit risk framework requirements. This is not just a matter of accounting, it is a pan-entity project. Hence, the implementation requires a proper project management. Implementation requires participation from at least the following units within the bank:
- Information technology
- Risk management
- Credit departments
- Internal audit
Several elements make the project worthy of board oversight. As described above, the project involves resources from across the whole entity. Second, the implementation of the prospective credit loss framework will have financial implications for the entity’s balance sheet. Finally, the standard requires materially expanded disclosures regarding the assumptions and methodology used in the estimation of credit losses. Each of these factors warrant oversight of the project at board level. In many cases the audit committee has identified this as a standing agenda item, while others have set up specific subcommittees to ensure successful implementation of the project.
Four specific areas require detailed attention for this project:
- Data management
- IT systems
Prospective credit loss models rely heavily on historic data sets of sufficient longevity and granularity to provide a basis for modeling. Many banks, especially in emerging and transition economies, lack this data. Multiple data sets include instrument data, macro-economic data, historical performance, customer credit information, customer personal data and collateral. Any project needs to perform a gap analysis to identify data gaps. The BCBS had 239 sets of potential risk data aggregations.
Banks need IT systems for data management, modeling, measuring and reporting. These systems need to aggregate the different data types and interface with the various requirements of the framework. The project should ensure the integration and scalability of the IT environment to provide the functionality the framework requires to function effectively.
The new frameworks require banks to be able to model expected credit losses under a range of economic scenarios. Hence, looking at historic loss levels, while a starting point, is no longer sufficient of itself to justify any projected credit losses. Banks need to analyze existing credit risk management methods and assess new modeling options to meet the new requirements. The proposed frameworks recognize that banks and their markets are at different levels of sophistication and allows them to make their plans within the context of that environment. However, it seems a reasonable reading of the standards, especially IFRS 9, that there is an expectation of a continuing improvement in a bank’s credit risk management environment.
As with any forward-looking model, there is a large degree of judgment required when arriving at final positions. Any development of the modeling framework should reflect this and identify how the bank plans to factor the judgment into the modeling and what steps it will take to ensure its ongoing integrity. This is important as it becomes part of the reporting process and will require acceptance by the banks external auditors.
Finally, the modeling needs to cover a range of scenarios that capture material non-linearites in potential economic outturns. Such scenarios may include base, upside and downside scenarios that are appropriate for the bank’s asset portfolio, exposures to credit risk and level of sophistication.
The disclosures requirements regarding impairment follow the expanded requirements of the Enhanced Disclosures Task Force (EDTF) established by the Financial Stability Board following the GFC. The element of judgment involved in prospective estimates of future credit losses requires materially expanded quantitative and qualitative disclosures in the financial statements. The disclosures cover both on and off balance sheet exposures and include such issues as credit risk profile, credit policies, concentrations, impairment policies and risk mitigation policies. Disclosures are at a level of granularity that facilitates a substantive balance sheet analysis. These expanded disclosures, which include an expanded management commentary, require careful management to ensure they are consistent with the bank’s business plan and overall communication strategy.
IFRS 9 adoption deadline is 2018 while US GAAP adopters of ASU 2016-13 have until 2020 or 2021, depending on their status. The experience of early adopters provides valuable insights into the challenges
The standards have a strategic impact on the bank. They impact the bank’s business model and balance sheet structure and so require understating at board level.
The implementation process requires formal project management. The standards are not just a change in accounting. Their implementation will have an impact across the bank and hence require strategic impact on the bank’s business model.
The identification of data gaps is critical. Regardless of the level of sophistication, the new standard requires an enhanced risk modeling process that is data intensive.
Modeling the quantitative outcomes of the process is difficult but must be done to identify the potential balance sheet impacts. Any material increase in expected credit losses will impact the bank’s public profile and capital requirements. Hence, it is important for the bank to understand the potential impact and develop a strategy to manage the public disclosures.
Prospective credit loss modeling is highly judgmental. Banks need to develop practical policies and guidelines to inform these judgments. Appropriate monitoring and governance mechanisms of the judgment process are important elements.
The standards have material impacts throughout the bank. The need to document and disclose credit risk management policies and processes has impacts beyond just reporting. The impacts flow back through lending, credit risk management, accounting, underwriting and pricing.
The implementation is sufficiently material to provide opportunities for reconfiguration of processes to realize internal efficiencies. Changes to IT systems, policies and operating processes are offering opportunities to reconfigure the organization to comply with the standards. A reorganization may provide the opportunity for effective implementation of the standard without necessarily incurring excessive cost.