- Laurent Birade, Senior Risk Consultant, SAS
- Geetika Chopra, Senior Principal Product Manager, Oracle
- David Grünberger, Deputy Head of the Division, Integrated Financial Markets, Austrian Financial Market Authority
- Martim Rocha, Director, Risk Business Consulting, SAS
- Nimesh Verma, Bank Advisory, Corporate & Institutional Banking, BNP Paribas
The impact of IFRS 9 is uncertain and could result in volatile financials, with a lack of both regulatory and market preparedness ahead of the January 2018 deadline. In a forum sponsored by Oracle and SAS, a panel discusses some of the key topics regarding the new regulation, including how banks and clients will be affected, who is best placed to ensure firms are prepared for implementation and the aspects the industry as a whole should be most concerned about.
What are the greatest financial impacts set to hit banks from the implementation of IFRS 9 and CECL?
Geetika Chopra, Oracle: The change to a forward-looking life-of-loan reserve calculation will, in most cases, lead to higher reserve allocation, affecting earnings for the transition years and thereafter. The fundamental reason for accounting bodies to adopt an incurred loss approach so far has been to dissuade earnings management. However, post-financial crisis, this viewpoint has changed.
Apart from the hit to earnings, financial institutions will have to spend money on tailoring their process and IT systems to adopt a forward-looking approach for reserving. In the case of CECL, where banks must provide for the life of the loan from its inception, the loan origination process will also undergo changes. Loans will be priced differently, considering the higher cost of capital due to enhanced reserve requirements. Data collection during loan origination will also add to the burden of doing business.
Laurent Birade, SAS: Without question, the changes in accounting standards will affect firms financially, and some initial estimates suggest the impacts will be significant, with overall reserve levels expected to increase by as much as 35–50%. Given that increased loss provisions represent the greatest impacts to balance sheets under stress scenarios, the effects of these accounting changes will likely ripple through future stress tests as well. In addition, while macroprudential stress tests have typically focused on the largest banks, these accounting standards apply to all financial institutions in their respective jurisdictions, regardless of size. But, along with these financial impacts, the changes required by IFRS 9 and CECL necessitate a much deeper level of modelling, analysis and reporting than before – and these changes are not insignificant. Firms may need to fundamentally adjust how they manage their loss allowance processes – including how they integrate their risk and financial data, design their analytics platforms and share information between departments. The scope of these changes can be substantial, depending on the complexity of firms’ balance sheets.
Nimesh Verma, BNP Paribas: Under IFRS 9, the carrying value of loan assets – which has been at amortised cost for hundreds of years – will change to incorporate modelled, forward-looking, volatile expected losses. As the accounting value of loans forms a key input to regulatory solvency, banks face immediate one-off hits to regulatory solvency when expected losses and lifetime expected losses increase significantly, relative to current provisioning and regulatory deductions. Interestingly, recent bank disclosure shows one-off hits to solvency to be manageable – a fall of 45 basis points in Common Equity Tier 1 ratio in the recent European Banking Authority (EBA) impact assessment, for example. This is lower than initially expected because of continued heavy provisioning and improved economic outlooks.
More importantly, IFRS 9 significantly increases volatility and procyclicality. It amplifies provision levels and drives large shifts in regulatory capital on an ongoing basis, as changes in economic outlook mean swaths of assets move between stages 1 and 2. At the same time, given amplified provisioning, we expect stress scenario impacts to swing wildly – with much higher impact for the same stress scenarios under IFRS 9 than under current standards. As a result – and as stress tests increasingly drive capital requirements – banks will need to either hold higher levels of spare capital headroom or use mitigants to address accounting volatility in future stress tests. In addition to product design, pricing and selective classification, we see an increased focus on active credit portfolio management. In particular, we expect to see development of hedges specific to IFRS 9, which target efficient reduction of stage-2 lifetime expected loss provisions, and new types of stress capital instruments.
How is the uncertainty around implementation of the new rules affecting clients’ preparation?
Martim Rocha, SAS: IFRS 9 and CECL are primarily principles-based. As such, the implementation guidelines are not prescriptive, and will likely change as consensus is built and clearer guidance is provided. Given the evolutionary nature of these standards, institutions may need iterative model development cycles before transitioning to IFRS 9 and CECL reserving. That might be too narrowly focused and could result in firms underestimating the amount of work needed to cover the full spectrum of required activities. Firms that start too slowly or become fixated on only a single aspect of the programme, with limited dedicated resources, may find their timelines for the overall implementation significantly compressed. For example, if we look at the history of regulatory stress testing as a guide, many firms initially approached implementation by simply adding head count to meet compliance requirements and deadlines, then attempted to refine their processes later. This can be a very inefficient and costly proposition for firms. In the case of IFRS 9 and CECL, with the financial-impact bottom line being more direct and transparent, it is paramount that firms are more deliberate and forward-thinking in how they approach their overall implementations.
Geetika Chopra: The standards are not prescriptive and do not recommend any single approach. Furthermore, the Financial Accounting Standards Board allows institutions to adopt approaches based on their complexity and size. There are no standard benchmarks to ascertain what constitutes a complex business model – much is left to the judgement and discretion of the auditor. Though the US Federal Reserve Board published a set of frequently asked questions on the adoption of CECL, such guidance is largely missing from central banks of most other jurisdictions.
Uncertainty has led to institutions adopting tactical solutions. Some are extending the models built for the Comprehensive Capital Analysis and
Review/Dodd-Frank Act stress tests/capital calculations to meet CECL requirements. However, whether that would sufficiently meet the standards and pass the audit test is debatable.
To what extent are financial institutions prepared for the technological challenges of implementing these rules?
Geetika Chopra: Most institutions find themselves underprepared – these challenges include the computation of effective interest rates and the amortisation of fees and cost over the expected life of a loan, for example. This significantly complicates the accounting process, and institutions will need to maintain separate sets of books for actual interest accruals and effective interest rate-based accruals. This is a huge challenge, considering that computation and its reconciliations are now a pre-close activity for all institutions.
Laurent Birade: Any time a principles-based rules framework requires implementation, institutions need to be ready to evaluate alternative interpretations of the standard and consider what their peers are doing to ensure that they are not alone in their interpretations. This is the crux of the challenge institutions face – implementing a technology framework that can adapt as interpretations of rules change. The EBA, for example, recently published an impact assessment raising concerns that up to 20% of banks have either reduced or completely eliminated the parallel run from their go-live plans. After getting the right data, models and vetted outputs, the setup and implementation of a repeatable production platform and process is critical. This platform will need to support a variety of runs on a month-end basis with all required approvals and documented explanations as to why the allowance moved the way it did in a limited timeframe. Having a highly flexible yet robust platform is likely to be where institutions will face the most difficult technological challenges.
How can supervisors respond to IFRS 9, and what are the most important responses banks can expect?
David Grünberger, Austrian Financial Market Authority: Both capital markets and banking supervisors are currently responding to IFRS 9. The European Securities and Markets Authority is co-ordinating national accounting enforcements, and many topics have already been discussed and decided upon – especially those from Austria. Publication of these decisions, which concern critical areas of stage transfers and loss quantification, is in the pipeline.
Although many critical implementation issues have not yet been publicly challenged, accounting enforcers and banking supervisors are now actively pursuing these issues to pre-clear some implementation issues and prevent outlying ones. The EBA has conducted its second impact study and published its final IFRS 9 implementation study. New stress-testing methodology now focuses on IFRS 9 provisions, which are the most crucial amendment.
The most complex advancement is probably the new stress-testing methodology – the EBA has established a forward-looking simulation in which banks calculate forward-looking expected credit losses (ECLs), as per forward balance-sheet dates. The probabilistic nature of stress tests will render simulations of IFRS 9 – the staging, for example – similarly probabilistic. In addition, the ‘perfect foresight assumption’ aligns the specifications of stress scenarios with those of accounting scenarios.
The triggers used to determine stage transfers is another key topic. A variety of approaches has appeared recently, and auditors seem hesitant to communicate where the boundaries are set – although new commentaries from the ‘Big Four’ audit firms have defined some limits. Accounting enforcers have brought up key topics such as the so-called ’absolute triggers’ and organised EU-wide supervisory clearance. A similar cleared topic was the future treatment of incurred but not reported losses. Banks will need to check whether their methods comply with decisions and expectations.
What is the most common mistake made by institutions trying to restructure their businesses to implement IFRS 9 and CECL?
Martim Rocha: The most common mistake is underestimating the work associated with implementing the standard – almost every bank in the world is struggling to meet the deadline. For example, when the first IFRS 9 surveys were conducted, around 70% of all firms stated they would do a parallel run for one year. Now, with the clock ticking and firms approaching the halfway point on the planned one-year run, few have been in parallel run for the first half of the year. Furthermore, many firms are considering no parallel run at all – or perhaps just in the last quarter with a limited set of portfolios. While several factors have contributed to this, underestimating the effort involved was clearly the main factor. Firms have found that developing the necessary new models proved more difficult than expected. Additionally, difficulties with collecting the necessary data – sometimes due to deficiencies in the existing systems or the maturity of the markets involved – extended the work associated with calibrating these new risk models. And now, even with many firms completing their initial-model development cycle, a new challenge is arising: how to assemble all of this into a process that can support the monthly/quarterly production demands.
Geetika Chopra: Most businesses realise that risk, finance and financial reporting now all have interdependencies, which require them to connect and communicate in the same language. Organisationally, this is a big challenge – ownership of ECL compute now lies with both of these functions, along with the overarching responsibility of model risk management.
Now that stakeholders of this process are located far and wide, it is imperative that institutions ensure there is single ownership of IFRS 9 and CECL programmes within a bank.
How will the European Commission’s recent moves to reduce the capital impact on banks by phasing in the requirements help in the long run?
David Grünberger: The phasing-in period is temporary and there is no strong downturn in sight – in fact, the rules will probably expire before having a severe impact. The complexity of these rules is striking, and many implementation issues will remain unclear well beyond January 2018. One expectation is that the phasing-in period will see reduced stress-test capital requirements. However, stress tests will also include a fully fledged implementation version, and analysts and markets usually prefer fully fledged stress impact. Thus, these temporary rules are something of a regulatory safety net to protect banks’ capital ratios, in case we see an unexpected credit crisis within the next few years. Therefore, the most likely scenario is that these rules will never have any real impact.
Martim Rocha: These moves aim to help banks smooth the transition to IFRS 9 in the short run and level the playing field between the internal ratings-based and standardised approaches to capital. Quite a few banks in Europe are struggling to survive and, in many cases, these moves assist particular countries with economies in slow recovery. These moves are therefore welcomed by many to help banks comply, but questions remain in the market as to their effectiveness – including the final impacts of the standard. The European Commission is opening the door to revising – during the transitional period – how IFRS 9 will be implemented in Europe. This can add to the uncertainty around the regulation going forward and can create potential misalignment with other jurisdictions, undermining the initial objective of globally harmonising the rules.
Geetika Chopra: It may not help in the long run, but it would help reduce the immediate impact of adopting the standards.
How much of a concern is the volatility around CECL estimates?
Laurent Birade: After the one-time hit to bank capital, volatility is the single largest financial concern banks will have. Moving from a framework that relates common metrics – such as days past due and other likely default indicators – to one where the lifetime loss estimates can vary based on multiple, less transparent factors is going to take some getting used to. For example, under the new CECL paradigm, things such as the length of the forecastable and supportable future are key elements of the level of expected loss. And expectations for that future are partially based on whether firms are currently in a benign or an adverse environment. Outside of that forecastable period, a long-term average is used. The methods used to determine and transition to that average can have serious impacts on the expected loss. These levers, and their impacts on the lifetime losses based on current conditions, certainly create the potential for increased volatility.
Geetika Chopra: Most institutions would change their business models and loan origination processes to minimise the impact to provision and reserve, but considering that macroeconomic, qualitative and environment factors are an overlay on ECL, it could also become an instrument to ensure reduced volatility. Nonetheless, these factors would come under higher scrutiny from auditors for the same reason.
Who are best placed to implement new accounting standards – accountants, risk managers or a combination of both?
Geetika Chopra: It has to be both together. The need to converge risk management and basic accounting principles is the very foundation of the standards, and would necessitate such a world view within the organisation as well.
Laurent Birade: The new standards affect so many areas that their implementation requires a cross-functional team to address. Most commonly we find the finance department driving implementation, but it is really dependent on the organisation and can be tough to generalise. We have seen instances where risk has led the programme and perhaps focused too exclusively on the loss modelling aspect, neglecting some of the issues around controls and production efficiencies. Also, it’s not enough to simply calculate ECL – the work to get losses posted to appropriate journal entries and feed the regulatory reports is also critical. On the other hand, accounting may not always be in the best position to assess the suitability of the credit loss models. It really depends on the skill and experience of the respective teams. In addition to the risk and finance teams, IT, audit and others should also be involved in the implementation process. Data, technology infrastructure and reporting, as well as appropriate governance and controls, all play a key role in addressing the standards.
What aspects of IFRS 9 or CECL are of greatest concern to you?
Nimesh Verma: For one, the lack of regulatory preparedness – no fundamental regulatory rethink is in train yet. IFRS 9 represents a sea change in key inputs to regulatory solvency, with the new expected loss provisioning having a significant impact on the carrying values of the largest component of banks’ balance sheets – loans. The current regulatory capital framework is calibrated assuming inputs based on incurred loss accounting standards. This calibration incorporates one-year expected loss and one-year unexpected loss at a 99.9% confidence level. Moving to expected loss provisioning – especially beyond a one‑year horizon – necessitates a recalibration or even a rethink of regulatory capital rules. This will ensure the frameworks and metrics are consistent. Otherwise, capital requirements will be double-counted by overlaying differing time horizons and confidence levels. If this double-count is not addressed, forward-looking provisions will lead to earlier and increased bank failures, which could damage system-wide stability.
It is therefore disappointing that a fundamental regulatory rethink to adapt to IFRS 9 has not yet begun. There has been insufficient study of how IFRS 9 will affect regulatory solvency – especially during stress. One would have expected at least a position paper from the Basel Committee on Banking Supervision on how accounting provisions and regulatory capital will complement each other – but this has not yet happened. As a matter of urgency, the Basel Committee needs to deliver a coherent framework that takes into account the transformative nature of the changes to the accounting standards.
Another aspect of concern is the lack of market preparedness, as investors and analysts are far from ready for IFRS 9. To date, both bank disclosure and comprehensive industry-wide studies of the impact and effects of IFRS 9 have been very limited. While data has been disclosed, this has merely been regarding the one-off impact of IFRS 9 upon implementation. There has been no focus on the ongoing challenge – how solvency will vary through the cycle and the potential level of volatility that may be experienced in profitability and solvency. Nor has the market understood how IFRS 9 will affect the meaning and relevance of key financial metrics used by investors and analysts.
Under IFRS 9, accounting profits and losses will become an even weaker proxy for value creation. Figures will not be comparable, and will be based on complex internal models (at a time when the Basel Committee is moving away from them). The reported profitability, financial statements and regulatory solvency will be completely different under IFRS 9 than previous standards. IFRS 9 is a new lens through which to view financial performance, and investors and analysts need to adapt their analytical methodologies and recalibrate their financial models. They may even need to develop an entirely new suite of analyses. However, from our discussions with market participants, it appears that this has barely begun. This is often blamed on the lack of disclosure and communication around IFRS 9 – we have to wait for bank communication and results first. Unfortunately, for many banks, this will first be forthcoming in 2018 reporting, once IFRS 9 is live. All of this gives scope for market confusion, not just next year but also when we hit the next period of stress.
Martim Rocha: IFRS 9 and CECL represent such a dramatic change to loss accounting, and the impacts are so public and meaningful to the financial statements that the transition brings a lot of uncertainty. Banks and investors will see major differences in reported numbers compared with historical financial reports and across peers, and they will collectively need to rationalise the extent to which these differences are due to accounting rule changes versus something more fundamental to the portfolio. As banks refine their modelling techniques, they introduce yet another source of volatility to the reported financials post-transition. It is going to take time to make sense of it all.
Geetika Chopra: The assumptions made of ECL models, qualitative factor adjustments and macroeconomic overlays must be adequately transparent and auditable. A non-transparent process, such as an Excel tool-based approach or a black-box solution, may be prone to errors and would become the scapegoat for any perceived earnings management by an institution. The fundamental reason for not allowing a forward-looking approach could become its own peril.