IFRS 9 METHODOLOGY: HOW DO YOU MEASURE UP?

IFRS 9 METHODOLOGY: HOW DO YOU MEASURE UP? In July 2014, the International Accounting Standards Board finalised a move to simplify the accounting rules for recognising and measuring financial instruments. It replaced the International Accounting Standard (IAS) 39 with the International Financial Reporting Standard (IFRS) 9 in an effort to address issues that arose during the global financial crisis, such as delayed loan-loss provisioning. The mandatory effective date of IFRS 9 is 1 January 2018. Complying with IFRS 9 s new impairment methodology means most financial institutions will require between two and three years to implement the new standard. Structuring a programme of work that addresses the cross-functional dependencies is critical to a successful implementation. Whilst some progress has been made, the majority of the industry only started scaling-up their IFRS 9 programmes at the start of Q1 2015. Our research was conducted to understand how firms are progressing with their key deliverables for 2015; methodology definition, model development and data development. The results demonstrate that, for a significant number of firms, a large amount of work is still required in order to meet the 2018 deadline. HOW S IT GOING SO FAR? ARE YOU ON SCHEDULE TO SIGN OFF AND IMPLEMENT THE CHANGES? Only about half of respondents are on course for the 1 January 2018 IFRS 9 deadline, with over a quarter saying they are experiencing a significant delay in their change programme. The most quoted factors responsible for delay were: Resource constraints/priorities, complexity of the data and systems required and waiting for the final standard to be published before starting. Significant Delay 26% Small Delay 21% Yes 53% We are still at the conceptual stage and having difficulty in understanding how IFRS 9 would work practically for retail banking exposures. The Basel Committee Guidance on accounting for expected credit losses has challenged initial assumptions and thinking. PORTFOLIO ANALYSIS MANAGER HOW ADVANCED ARE YOU WITH METHODOLOGY DEFINITION AND MODEL DEVELOPMENT? With just two and half years until the deadline, 60% of the people we spoke to said they weren t even half way there with the IFRS 9 methodology definition and model development stages of their programmes. To stay on course for compliance, methodology definition should be complete by the end of 2015. Workstream completion % 0-20% 21-40% 41-60% 61-80% 81-100% Methodology Definition Model Development 0% 5% 10% 15% 20% 25% 30% 35% 1

WHAT ABOUT METHODOLOGY AND MODELS? IFRS 9 requires new capabilities to identify criteria indicative of significant credit risk deterioration, to calculate lifetime expected credit losses (ECLs) and to reflect changes in forward-looking economics, policy and regulation. WHICH MODELS ARE YOU LEVERAGING TO CALCULATE IMPAIRMENT UNDER IFRS 9? More than half of firms are leveraging their Basel models in one way or another, and many are using existing impairment models and forecasts, but over a fifth have started creating brand new models from scratch. Whichever option firms choose, they will all require an enormous amount of modeling and quantitative effort and expertise. 74% believe loan portfolios of < 1bn should not require a fully modelled solution All new models 21% Basel models only 33% At or below a billion, the level of data is not significant enough to offset implementation complexity DEPUTY CFO Other existing models 21% Basel and other models 25% For banks without IRB models, the change in methodology from IAS 39 is significant. Rather than calculating incurred loss, a lifetime expected loss model will be need to be developed for all lending products. For those banks with IRB models, leveraging these models will provide some efficiencies. However, whilst the definitions of ECLs are identical to the Basel definitions, the computation of ECLs under IFRS 9 will require key adjustments to remove the effects of the Basel rules: 1. One-year time horizon replaced by full behavioural lifetime. 2. Through-the-cycle and downturn default assumptions are removed in favour of point-in-time. 3. Probability of Default (PD) based on probability-weighted outcomes as opposed to best-estimate. 4. PDs and Loss Given Default (LGD) will include the full spectrum of information available to calculate forward-looking estimates, including the use of macro-economic forecasts. 5. Effective Interest Rate (EIR) calculations will need to include cash flows from the reporting date, not default date. 2

HOW WILL YOU MEASURE SIGNIFICANT DETERIORATION? Significant deterioration is a key concept within IFRS 9 that triggers the switch from using 12 month ECL to lifetime ECL. No specific or mechanistic approach is imposed by the Standard and will vary depending on the sophistication of the bank and the financial instruments it holds. European regulators are expecting banks to look widely and holistically for information when assessing significant deterioration and the Basel Committee on Banking Supervision s consultative document on sound credit risk assessment and valuation of loans (SCRAVL) has highlighted that relying only on past due information would constitute a very low quality implementation. WHAT METRICS HAVE YOU USED TO DETERMINE SIGNIFICANT DETERIORATION? Our research suggests respondents have taken the Standard and regulators requirement on board and are using a wide range of metrics in order to determine if significant deterioration has occurred. Given that the Standard requires firms to measure deterioration in probability of default, it is unsurprising that over 80% or respondents will make use of existing default models. Using delinquency and forbearance as a backstop to identifying stage 2 accounts is also a favoured approach. Probability of Default Delinquency Internal Credit Ratings and Scores Forbearance Qualitative Assessment High Risk Account Management External Credit rating and Scores Value of Collateral/Finance Rates or existing instrument terms Operating results for the borrower Credit Prices 0% 20% 40% Wholseale 60% 80% Retail HOW ARE YOU CLASSIFYING FORBEARANCE? 80% of respondents expect forbearance to be limited to only one of either Stage1, Stage 2 or Stage 3. This is surprising and suggests some confusion amongst firms as to the imperfect relationship between FINREP and IFRS definitions. Depending on the type of forbearance and the reason for asset modification, there may be a stage it will predominantly be attributed to. Stage 1 Only 41% However, both non performing and performing forbearance as defined in Stage 2 Only 34% FINREP could be allocated to any of the 3 stages. Given the operational issues in capturing forbearance and modified asset type accurately, this adds a Stage 1, 2 & 3 Stage 2 & 3 11% 9% further layer of complexity to calculating significant deterioration as it does to systems and processes relating to these exposures. Stage 3 Only 5% 3

HOW WILL YOU HANDLE MODIFIED ASSETS? Every bank will have to use substantial levels of judgement when modelling their provision. These judgements will not be standardised across banks and so will require appropriate levels of disclosure to explain the assumptions and methodologies used. As discussed above, categorisation and disclosure of the requirements for modified assets are different to those required for regulatory returns (FINREP, Basel III etc.). Ensuring that correct information is captured will have an impact on the operational functions that process forbearance. WHAT PROBLEMS DO THE DEFINITIONS OF MODIFIED ASSETS IN IFRS 9 PRESENT YOU WITH? Almost 9 in every 10 respondents expected problems with the definitions of modified assets. No Problems forseen 12% Expect Problems 88% Measurement of modified financial assets is a major challenge in the financial services industry HEAD OF CORPORATE REPORTING Half the respondents expect issues with the measurement to be a major challenge, saying it added another layer of complexity to an area that is already difficult for the financial services industry to report on. A lack of data on modified assets was cited as the most common reason for measurement issues. Problems caused by consistency, definition and reporting were largely as a result of the difference in definitions and treatment between forbearance for FINREP and Basel III and the ability to reconcile between the two. Measurement 50% Reporting 38% Definition 36% Consistency 30% Unknown/unsure 16% 4

DO YOU HAVE THE DATA? European regulators expect banks to invest in appropriate resources for a full and robust consideration of forward looking information, including macro-economic forecasts, in their calculation of ECLs and significant deterioration. What is deemed appropriate will depend on the size of the bank. However for the larger national and international banks the regulators expectations are that obtaining the relevant information will not involve undue cost or effort. WHAT ECONOMIC FORECASTS ARE YOU USING AND WHERE IS THE INFORMATION USED? Simple base case used in planning and stress testing Multiple bespoke scenarios Qualitative adjustments Other economic forecasts/undecided In individual calculation As an overlay Unsure 22% 36% 33% 85% Given the significant data challenges posed with modeling lifetime expected credit losses, most respondents are trying to use as much of their existing information as possible, particularly making use of current base cases and stress scenarios. Over half expect to apply the forecasts at account level although many pointed out that use of an overlay is also being considered. Calculations will need to be made more precise and consistent with Individual Capital Guidance calculations. Therefore we will need to incorporate more granular, supported data HEAD OF MARKET AND LIQUIDITY RISK DO YOU HAVE ALL THE ADDITIONAL DATA REQUIRED TO DEVELOP AN IFRS 9 COMPLIANT METHODOLOGY? Less than 1 in 10 IFRS 9 programmes have all the data they need, with more than half either having none or only some of the additional data they require. No 26% Partially 31% Yes 8% Mostly 35% Over half (54%) of respondents are using data never previously used for impairment calculations, forecasting or stress testing Firms are having to mine for new data they have not gathered, sourced or modelled before. Respondents told us about their need for more granular data, extra macroeconomic forecasts and extended time horizons. In addition, the burden on data architectures is compounded by the requirement to compare current credit quality to credit quality at origination. This will require data to be stored over the full behavioural lifetime of the financial instrument. 5

ARE YOU HAVING TO MAKE CHANGES TO YOUR DATABASES AND/OR SYSTEMS IN ORDER TO CLOSE THOSE DATA GAPS? Almost 9 in 10 programmes are having to redevelop their databases and/or systems, with two thirds of respondents saying this redevelopment is a high impact problem. Yes - Low impact 23% No 12% Yes - High impact 65% All the challenges we re facing are as a result of higher complexity in calculations and dependency on additional data that needs to be made available in the systems HEAD OF GROUP ACCOUNTING POLICY A significant challenge is that even when data is available, in principle, it may not yet be fed into the appropriate IT systems. Additionally, the new data required will be held in a number of different systems and integration of their output data is a complex problem to solve. This will require, in most cases, significant effort to capture, verify and pass on the data in a systematic way in time for compliance. CAN YOU FIND THE RESOUCES? HOW MANY PEOPLE ARE WORKING ON THE WHOLE PROGRAMME AND HOW MANY RESOURCE GAPS DO YOU HAVE? Number of people working on the IFRS 9 Programme 0-10 11-20 20-30 31+ Resource Gaps on the IFRS 9 Programme 0-5 6-10 10-20 21+ 0% 10% 20% 30% 40% 50% 60% % of respondents Despite IFRS 9 first being published in 2012 before being finalised in July last year, the majority of firms only started building their IFRS 9 programmes from Q1 2015 onwards. This has created a surge in demand for institutions who need to secure the right technical resources and experienced change professionals to deliver this critical programme. Firms of varying sizes have responded to the survey, however 50% had resource gaps in excess of 6 people, with 1 in 10 requiring more than 21 people. This poses the danger that a limited supply of expertise, combined with high concurrent demand, could disrupt even the best planned initiatives. 6

WHAT S THE OVERALL PICTURE? WHAT ARE YOUR BIGGEST CHALLENGES IN SOLVING THE IFRS 9 CONUNDRUM? Three quarters of people told us that models and methodology were their biggest headache, with only a couple fewer citing data as a major problem. These challenges are driven by the complexity involved in the practical application of what is in essence a simple set of principles. Models & Methodology Data IT Resourcing Other 14% 41% 76% 73% 68% This interplay between simple guidelines and complex application raised the question as to what a robust solution would look like. Understandably, there was a difference between those that are responsible for modeling the solution, many of whom felt that the standard demanded complexity, and those who had to explain the data whose preference was for simplicity, reliability and ease of explanation. However, most respondents acknowledged that the final solution will lie somewhere between these two viewpoints and should be no more complex than necessary. 83% believe a robust solution to be a simple, reliable and easily understood process Too much complexity cannot be sustained HEAD OF RISK METHODS & MODELS 17% believe a robust solution to be a complex and exhaustive implementation For the solution to be fully auditable, we would like a simple approach. However, I think some complexities are unavoidable. PORTFOLIO ANALYSIS MANAGER It s challenging enough explaining the current Expected Loss methodology to governance committees; additional complexity may add to precision, but in previous similar situations we have found additional complexity does not necessarily add to accuracy HEAD OF MARKET AND LIQUIDITY RISK IN CONCLUSION In 2013, the BBA predicted that the complexity and cost of implementing IFRS 9 would be second only to the Basel II implementation and broadly equivalent to the entire first time adoption of IFRS in 2005. Experience to date, as borne out in this research, suggests this to be an accurate prediction, even where the solution is being kept as simple as possible. Whilst 2018 may seem a long way off, failure to make progress now will jeopardise the chances of a timely implementation. 7

RESEARCH METHODOLOGY The research was undertaken by the Centre For Financial Professionals (CFP). CFP conducted an online questionnaire, receiving 227 responses between 23 April and 22 May 2015. The most appropriate professionals in the CFP and Parker Fitzgerald databases were surveyed. The raw data was provided to Parker Fitzgerald for further analysis and interpretation. Of the most relevant responses used to compile this data, more than half were from retail and commercial banking entities. About two thirds of those surveyed work within the Risk function, with the rest offering a finance perspective. Retail & Commercial 57% Universal 16% Corporate Bank 11% 37% Risk Investment Bank 5% Finance 63% Insurance 5% Asset Finance 5% Asset Management 2% ABOUT PARKER FITZGERALD Parker Fitzgerald is an award winning professional services firm specialising in the delivery of risk and regulatory transformation within the financial services sector. We partner with the world s leading financial institutions to manage the strategic impacts of new financial regulation across the enterprise and deliver market leading capabilities for the risk function. With extensive experience working in risk management for many of the world s leading financial institutions. This includes credit risk, loss forecasting, econometric and stochastic analysis. The firm has extensive experience in the design, build and validation of both capital and impairment models to support both retail and wholesale product sets. We have been intimately involved with our clients IFRS 9 programme delivery since July 2013. Whichever stage of the IFRS 9 implementation you have reached, Parker Fitzgerald has an experienced team of IFRS adoption, credit risk and change practitioners to assist you in assessing the strategic and operational impacts on your organisation and delivering the necessary change across the enterprise. Disclaimer The information contained in this document has been compiled by Parker Fitzgerald and includes material which was obtained from information provided by various sources but has not been verified or audited. Except in the general context of evaluating our capabilities, no reliance may be placed for any purposes whatsoever on the contents of this document or on its completeness. No representation or warranty, express or implied, is given and no responsibility or liability is or will be accepted by or on behalf of Parker Fitzgerald or by any of its partners, members, employees, agents or any other person as to the accuracy, completeness or correctness of the information contained in this document or any other oral information made available and any such liability is expressly disclaimed. CENTRE FOR FINANCIAL PROFESSIONALS The Centre For Financial Professionals is an international research and business to business event organizer. The focal point for financial risk professionals to advance through renowned thoughtleadership, knowledge sharing, unparalleled networking, industry solutions and lead generation. Driven by and dedicated to high quality and reliable primary market research, the Centre for Financial Professionals provides financial institutions, industry professionals and companies with unparalleled knowledge sharing through research projects and reports, peer-to-peer conferences, exhibitions, news and reports. Whether your objective is thought-leadership and networking at an international conference and exhibition, a focused training course to advance your knowledge and skillsets or an independent research project to base your next thought-leadership report, the Centre for Financial Professionals can provide the right service for you. 8