10 key steps to implement IFRS 9

In January 2018, banks will have to change the process that currently calculates their credit impairments. While 2018 may seem quite a long way off, there is a lot of work to be done in order to meet the requirements on time.

Those who start now to work towards implementation stand themselves in good stead to meet the requirements on time but also to establish a robust and efficient framework that leverages existing tools. Taking this approach and starting work now will leave risk and finance teams with a well thought out IFRS 9 infrastructure that will be straightforward to manage in the future.

Here are 10 key steps to a successful implementation:

1. Interpret the standards collaboratively:

The first step is to examine and interpret the IASB’s IFRS 9 standards and GAECL principles. Collaboration across functions provides a challenging aspect to the project and so developing strong ties between the risk and finance worlds is essential.

2. Assess your current situation:

It is imperative to assess your organisation’s data, system and model infrastructure for suitability under IFRS 9. Re-use of existing tools improves understanding of the solution, reduces complexity, cost, and aligns key risk and finance outputs.

3. Review historical data and existing risk models:

A thorough review of data items early in the project will pinpoint any shortfalls in the capture, storage or processing of data. This allows sufficient time to engage the wider business to organise and implement the necessary changes. 

4. Develop the right methodology:

There is no ‘one size fits all’ solution. Existing methodologies may not wholly solve the problem, so exploring alternative methodologies is important.

5. Test prototypes:

Building prototype models will enable the pros and cons of multiple methodologies to be identified in a manageable timeframe as well as enable you to carry out an early impact assessment of the provision figures under IFRS 9.

6. Refine methodology:

A review of the results from the prototype models and initial impact assessment stage may uncover issues with provision adequacy or stability. These issues can often be addressed by revising the modelling approach and so it’s important to have the time and opportunity to refine approaches.

7. Finalise modelling:

All modelling should adhere to already established standards with sound model validation and documentation to ensure they are robust, predictive and relevant.

8. Implement model governance:

Comprehensive quality documentation is the window through which the models will be judged and so is critical to demonstrating compliance.

9. Implement models and parallel runs:

The models must be implemented and run in parallel alongside IAS 39 models. This can create resource constraints as two distinct processes are run in parallel and so care will be needed to upskill existing teams ready for the challenge. 

10. Monitor regularly:

Regular monitoring will ensure the models remain fit for purpose. Ensuring a monitoring framework is in place means that models continue to be assessed for compliance to IFRS 9 and internal validation standards.

When choosing an external partner to work with, make sure any consultants you engage are fully aware of the requirements and can demonstrate solid experience in preparing data and building lifetime expected loss models. The correct implementation of the requirements under IFRS 9 requires data specialists and modellers who have direct experience and a deep understanding of forward-looking loss forecasting.