Arguably one of the most significant challenges around IFRS 9 implementation is the volume and depth of data required, not to mention the analysis of that data. New research reveals the data gaps that many financial institutions are struggling to plug in advance of the January 2018 deadline.
IFRS 9 carries a much larger data requirement than its predecessor, IAS 39. One of these data hurdles is the ‘solely payments of principal and interest’ (SPPI) test for classifying and measuring financial instruments. Under the SPPI test, over 70 different attributes of a security must be checked to determine whether it is eligible to be held at fair value or amortised cost. That is an enormous amount of data, and across a number of assets it can quickly become unmanageable, or too large for legacy systems to process.
The other main data hurdle banks face as a result of IFRS 9 is the expected credit loss (ECL) calculation. According to new research conducted by Regulation Asia in partnership with S&P Global Market Intelligence, 23% of respondents have concerns around meeting the data requirements to support ECL modelling, with more than half reporting partial to substantial data gaps in ECL estimation.
Almost one-fifth (19%) of respondents worry about the overall development of ECL models, and a similar proportion attach importance to developing an internal systems infrastructure for ECL calculations and reporting. However, a decent number (43%) of respondents expect to have adequate internal capability to implement a compliance governance framework to support ECL modelling.
Looking at the most common data challenges associated with the ECL calculation, there are subtle differences between banks in different geographical regions. In Europe, corporate lending data was cited as the most significant challenge (27%), followed by SME data (23%), and securities data (21%). In the Middle East and North Africa regions, however, SME data was seen to be most challenging (34%), followed closely by corporate lending data (32%).
Respondents also highlighted significant concerns around reconciliation between financial reporting and credit risk data, as well as data integrity. Part of the challenge around data integrity arises from the fact that different institutions are using different sources of information for the underlying assumptions in their ECL models.
According to the research, around 40% of respondents are relying on internal resources to estimate forward-looking macroeconomic assumptions for their ECL model, with 32% turning to external forecasts and 29% to regularly available forecasts. Due to having limited information in-house, many small- to mid-sized institutions will likely take a simpler approach and rely on external economic forecasts.Whichever assumptions are used in a bank’s ECL model, they must be incorporated into an institution's corporate governance framework and receive board and regulatory approval. What’s more, given the subjectivity and judgment required, banks must ensure any assumption or calculation methodology – including data sources – is transparent and auditable.
Given that many banks have their historical data scattered across multiple applications and databases, and separate from their risk data, IFRS 9 implementation is no walk in the park. Nevertheless, the work required will not be in vain. In addition to demonstrating regulatory compliance, banks will be able to use the enhanced level of data required by the new accounting standard to better inform their day-to-day operations, as well as any strategic decisions.