Maurizio Garro, Senior Lead, IBOR Transition Programme, Lloyds Banking Group
The IBOR transition represented one of the biggest revolutions in the history financial sector and, at the same time, brought about an opportunity to develop and enhance a framework to manage a model change. In this article, Maurizio Garro, Senior Lead, IBOR Transition Programme, Lloyds Banking Group, reviews the specific challenges financial institutions faced in managing the model risk that stemmed from the IBOR transition, and looks at future considerations that still need to be addressed.
The IBOR transition required the banking institutions to have robust plans to replace LIBOR with the RFR, term structure and arrears compounding. It is no secret that the model developers (quants) and model risk management function (MRM) have been playing a key role in the LIBOR transition groups and committees of the financial institutions.
In the last two years, quants and MRM teams at large institutions have worked heavily to change/update and validate/approve a large number of models for which LIBOR has been replaced. Of course, the first challenge was hiring new resources in a timely and cost-effective manner as finding the right candidates became more difficult with many institutions competing for a small pool of candidates (with obvious increase of the salary/fees).
From a model risk perspective, the first question has been to assess the materiality of impact of the change on the models. However, this must be performed across the full model life cycle and not just the development side (see figure 1 below).
The impacts must be assessed according to the nature of the models currently used in an institution like regulatory, pricing, internal etc. This is a critical classification because the plan and approach to model risk management is different across these models and may require specific actions.
In addition, there are other categories that should be included in the impact assessment due to the IBOR transition; these are the End User Computing (EUC) and Calculation platforms (despite the fact that they may not meet the model definition embraced by the financial institutions).
Certainly, inputs change and proxy (mainly for risk factors) have been one of the main changes (see also the fundamental review of the trading book, FRTB). In this case, a key consideration was to evaluate what could be the liquidity of the risk factors considered in the models due to the LIBOR transition. In some cases, we may have witnessed a perfect storm where LIBOR liquidity was lower while the market data and relevant transactions for the new RFR were not sufficient to give a satisfactory historical data.
Other aspects linked to the ones mentioned in the previous paragraph have been the monitoring of the model performance over time including the review of model limitations. The latter has been and will be a potential conflicting point between the business needs and the MRM function. In fact, the possibility to have a consistent number of limitations on the LIBOR-replaced models is something which can be faced by many financial institutions due to the short timeline allocated to the validation of the models.
Last, but not least, one fundamental challenge has been the timeline to get the model changes through the model governance according to the internal policy and the regulatory approval when applicable. It is worth to mention that the pressure on the quants and model validation to manage all the required model changes in such a compressed timeline has been huge, and this may have translated to an additional source of model risk.
A question that has been raised quite frequently was whether the financial institutions had to change their own model risk policy to accommodate for model changes triggered by the IBOR transition.
Of course, there is no right or wrong answer to this question as this depends on the stage where each institution is in the model risk management journey. However, there are some considerations that have come up as part of the IBOR transition:
What is the lesson that the LIBOR transition has taught us? From a model risk perspective, it was made clear that the approach chosen by the financial institutions in terms of model risk appetite and associated triggers play a fundamental role to balance the business needs and the model governance requirements. Of course, this has been and will be an important conversation between the relevant stakeholders and players like banks, regulators, insurance etc.
The LIBOR transition has been and could still be an opportunity to develop a more streamlined approach to manage model change in a timely manner and still comply with the principles of a good model risk management.
