Looking ahead to FRTB implementation and understanding the requirements

Looking ahead to FRTB implementation and understanding the requirements

Xavier, can you tell the Risk Insights’ readers about yourself and your professional experiences?

In 2005 I became one of the founders and managing directors at ActiveViam. I manage our UK operations and I am in charge of business development in the finance industry across EMEA and Asia-Pacific. I am constantly looking for ways to expand our offering to meet the needs of financial institutions all over the world.

Can you outline the top critical challenge for banks with regards to FRTB implementation?

I see 3 main challenges:
– Defining desk granularity and boundaries
– The complexity of P&L attribution
– Maintaining data availability and quality all throughout

The common thread across all those challenges is that they demand a level of analytics capabilities that few banks possess today. In order to meet these requirements, many will have to upgrade their architecture. The fourth and final challenge for IT managers will be to find a solution that can accomplish all that and does not become a time and money sink to implement.

How can risk managers better prepare for the transition process from the existing regulatory regime to the FRTB regime?

A key aspect of the transition is that whatever model the bank ultimately chooses (Standard Approach or Internal Model Approach), this model will need to be tested thoroughly and over a fairly long period of time. Risk managers will need solutions that enable them to keep driving the business as usual, while being able in parallel to test and adjust their models and to ramp up their capabilities progressively.

Can you explain some of the potential challenges for banks of running different models for different regulators?

It’s a triple challenge of accuracy, speed and practicality. First, you need to be certain of the accuracy of the data under different models, and this means not just the raw data but also the calculation results. Then there is the question of speed: any model involves complex calculations that each require significant computing power, having them run on parallel platform would be slow and costly. Finally, having to juggle with several systems that have each their own user interface and quirks would be highly impractical for end users.

The best approach therefore is to have a single platform that can support several models for different agents, and ideally is even able to switch from one to the other at will. The only technology that can do this today is in-memory computing, where the data is stored unaggregated, and all calculations and aggregations are performed on the fly.

How do you see the role of the market risk professional changing over the next 6-12 months?

I expect they will spend more time preparing reports for the regulator. Ultimately the transition to FRTB will be truly successful if such reports can be produced with accuracy, but also taking too much time away from your regular tasks, those that actually move the business forward.