Pool Party – Examining the potential of pool models

Pool Party – Examining the potential of pool models

By Tobias Noll, Senior Relationship Manager/Deputy Head of Marketing & Sales, RSU Rating Service Unit

Can you please tell the Risk Insights readers a little bit about yourself, your experiences and what your current professional focus is?

I have been working in the banking industry for over 12 years now, starting my career in investment banking one year before Lehman collapsed. At that time, working in a front office position, I naturally had a highly ambivalent attitude towards “risk”, as it usually appeared to be a business spoiler. Over time, in the post Lehman era, this perception as a whole changed and coping with all the different kinds of risk became daily business.

In my current role as a Senior Relationship Manager at RSU Rating Service Unit, dealing with credit risk and all of the related supervisory requirements is the main focus of my daily business. My role is to support RSUs clients in the German and European banking industry with regards to our internal rating models. Being a provider of internal pool models in times of fast changing and constantly increasing supervisory requirements, professional life does not often appear to be a pool party at all.

What, for you, are the benefits of attending a conference like Risk EMEA 2019 and what can attendees expect to learn from your session?

Information is key. Attending a conference like Risk EMEA 2019 gives all delegates an opportunity to meet their peers from around Europe and to share insights and experiences on relevant risk related topics. A cross boarder exchange of information becomes more and more crucial, especially since the supervisory requirements are laid out across Europe in a homogeneous way.

There are different ways of coping with all the upcoming supervisory requirements regarding internal models. My session will give the audience some insight on the collaborative approach of pooling data from different banks and running internal models on that very data pool. This approach is rather unknown in Europe, but has worked very well for the past 15 years in Germany. Expect to get hooked and feel free to join our pool party!

What are some current challenges financial firms are experiencing in relation to internal models?

The implementation of the EBA Guidelines on PD and LGD estimation (EBA-GL-2017-16) alongside with stricter regulatory requirements in general and additionally the TRIM-exercises are most certainly the biggest challenges banks are facing not only right now, but also for the upcoming years. In addition, across the industry attention has turned towards cost efficiency, which means that keeping models running at any price is not an option many banks are willing to pursuit.

What are some possible solutions for the current challenges within internal models?

Frustrating as it may seem, the supervisory landscape is not likely to change any time soon. Hence, banks either have to fulfil the supervisory requirements, or they need to keep their eyes open for alternative solutions.

Some 45 banks decided to join RSU Rating Service Unit and use our pool models for their wholesale business. They simply outsourced the development, validation and the operation of the internal models for various types of obligors, offering the banks flexibility and a cost-efficient way of focussing on their core businesses. One of the many positive side effects is that RSU as the service provider, copes with most of the supervisory requirements for example implementing the EBA Guidelines on PD and LGD estimation. Where necessary, clients are involved, but the main part of the heavy supervisory-workload remains at RSU Rating Service Unit.

This solution could easily be adopted in other European jurisdictions, first movers can be found in Austria and Luxembourg, where several banks already joined our pool party.

Can you give a brief overview of what pool models are and why they should be used?

Basically, our pool models work like any other internal model, only more accurately. This is because all participating banks contribute their individual data to one growing data pool.

RSUs models rely on that pool which benefits all contributors: meaningful statistical analyses require a certain amount of data. The more – the better, and the more accurate the forecasts. Where data is scarce, it will also be difficult to provide statistical evidence of the validity of the systems. All of our models are approved for the use under the IRBA.

Find out more: www.rsu-rating.de/en/

risk emea series