1. David, can you tell us about yourself and your experience in the IFRS 9 field?
I’ve been crunching numbers for a long time now, first as an aerospace engineer, then as a validator and risk modeler in banking and insurance. Working in the financial world has been quite roller coaster ride during the past decade, so I haven’t regretted this career switch even once.
I’ve had the privilege of being head of KBC Group’s risk modelling team for several years. My team worked on all types of financial risks, including lots of lifetime credit impairment modelling.
When IFRS9 appeared, it was clear to me that it will fundamentally reshape both the finance and risk professions in banks. I joined CREDO in 2015 to become fully involved in this trend. The joint skill set of our team of experts allows us to offer end-to-end IFRS9 solutions to banks: modelling, ICT, software and governance. We are currently supporting numerous local banks and activities abroad are picking up speed.
2. Can you briefly tell the audience some factors that have emerged, which the IASB may not have anticipated, when releasing the standard?
I am regularly confronted with the limited opportunities for simplified modelling in smaller banks. The standard expects IFRS9 implementations to be commensurate to the size and complexity of financial institutions, but this did not consider the thin line that exists between ‘simplified’ and ‘plainly wrong’ modelling. In practice, there are not that many opportunities for simplification.
I also doubt whether the IASB anticipated just how low impairments can go when all conservatism is duly removed from the calculations. Unless rather pessimistic forward looking scenarios are included, under the current economic outlook banks may actually end up with impairments lower than their current IBNRs under IAS39. Given these low figures, I am puzzled by proposals for phased introduction of IFRS9 in reporting. I am also curious about the reaction of regulators, who will soon be confronted with the real final figures.
3. What are some of the possible modelling challenges risk managers could face?
There has been lots of talking about lifetime PD modelling, a field which still offers interesting opportunities of research. But I feel that LGD modelling has been forgotten in the process. Complex credit structures, in which various collaterals, guarantees and credits interact, are already quite a headache in regular LGD modelling under IRB. IFRS9 substantially raises the bar, by requiring to do the same in a forward looking context.
4. What are some of the key differences between implementation for smaller and larger banks, and the unique challenges each face?
While modelling poses a serious problem for smaller banks, they may have an edge over larger financial institutions in other respects. Their structure is simpler; therefore, their balance sheets are more straightforward to process, which makes a big difference when interpreting at results. ICT-wise, legacy systems and complex data flows also tend to be less of a problem.
5. In your opinion, what are some of the key challenges faced in the final preparation for implementation of IFRS 9?
Many banks underestimate the compliance requirements of IFRS9. As the focus shifts from modelling to disclosures, the need for a watertight process that forces the bank to stick to its reporting principles even when times will get rough is overlooked. This requires not only setting up robust governance, but also a software environment that guarantees the proper execution of this governance