The FRTB regulations will mean a huge change for banks and their capital regimes. Any great change to a regulatory regime comes with a lot of challenges and we wanted to look at the 10 greatest FRTB challenges facing financial institutions as a result.

FRTB Challenges

  1. Strategy – Setting a clear strategic path to getting ready for FRTB will be a big challenge. The institution will also need to allocate a sizeable budget to the delivery of this large and extensive program that will touch all aspects of the banks processes – from front to back. Studies indicate that anywhere from 40 – 120 million EUR will need to be spent to get ready for FRTB. This will especially be the case for the medium to large institutions that will have to have a very clearly defined roadmap that leads them to being ready for the regulatory prescribed go ready date in 2019. This will also require a clear understanding of the regulations and how it affects the institution in question and what processes and systems have to be changed in order to get ready for FRTB.
  2. Technology – It’s quite important to realise that while there are some quantitative challenges to implementing FRTB, this is fundamentally a data challenge. This comes down to the fact that under FRTB there will be loads and loads of data produced. So much so that we are in Big Data territory. This will mean that banks have to move from a traditional approach of using RMDBS to store data to technologies like Hadoop, Spark, Scala. In short the whole technology stack needs to be changed and very specialist skills will be needed – these don’t come cheap and are not easy to find. You can read a lot more detail about the challenges around data volumes here.
  3. Staff – Banks will need to make sure they retain their best staff and also hire on top to deliver what will be a complex program of work over a couple of years. People will also need to be trained up to learn new skills and improve their knowledge. This will especially be true for institutions that have not had to worry about things like curvature risk in their current models, but they will need to account for this even with the Standardised approach under the new model.
  4. Desk Structure – This will be a very important aspect to implementing FRTB in a bank. Fundamentally banks will need to adhere to a desk structure that complies with the rules the regulatory have set. Given some of the subtleties of the rules, and the fact that liquidity horizons form part of the capital calculation, we will no doubt see a reduction in the desks that trade complex products as these will attract a much higher and more punitive capital charge. Fundamentally there will be some tough choices that need to be made by management in banks as to which business lines are going to be persisted with and which ones will be wound down.
  5. Banking / Trading Book Boundary – This is a cornerstone of the FRTB rules and rules will be put in place to make the boundary more clear with rules for inclusions and exclusions. There will also be limitations on reclassifications which will mean arbitraging will be much harder between the trading and banking books.
  6. Reference Data – FRTB rules require a lot of classification and management of data according to market parameters like sectors, industry, capitalisation etc. This type of data is easy to get from systems like Reuters and Bloomberg (and others), however the challenge for banks will be to bring this together in a single system where enrichment and transformation is managed in a consistent and uniform way. Again this is something that can be done in a Datalake type big data store and transformation and enrichment can happen using different technologies.
  7. PnL Attribution – This is a big deal. This requires that the finance and risk functions become much more joined up. This can vary from institution to institution, but the overwhelming state of play is that there is a disconnect between these 2 crucial functions. This will requite closer cooperation in terms of people, systems and data models. Also consider that this will need to be done at the desk level and this really will be a tough challenge to overcome.
  8. Data Quality – While we have talked about the amount of data that will be produced and the types of data that will be needed for enrichment and from a reference data point of view, we have not talked about the data quality issues. This will be fundamental, the regulations go so far as to state that there will be a requirement for a single golden source of data that can be relied upon across the institution. Add to this the fact the lineage of data has to be clearly tracked and audit trails can be maintained (potential uses for Blockchain technology perhaps?). There will also be a need to ensure that sound and solid data models are maintained across aa institution.
  9. Market Data and Non Modellable Risks – The FRTB regulations are very prescriptive about the time series data that is required in order to consider it of high enough quality to be usable. We will talk about this in more detail in a separate post. However the challenge will be getting good enough data for all risk factors – especially for the stress period calibration which will need banks to go all the way back to 2007. There is of course the ability to use a reduced risk factor set (if this explains 75% of the risk under full data) but there will still be considerable challenges across market data teams to source adequate quality data that can ensure that instruments are not pushed into the non modellable risk factor bucket and thereby attract a higher capital charge.
  10. Standardised Approach – Currently not every bank, especially the larger Tier One institutions, has a standardised model that they run. Under FRTB every institution will have to run the sensitivities based approach. This will be something that needs to be built out and will need to be calculated at overall level and also at desk level. The regulator views this as a benchmark which allows them to compare institutions using one standard framework and benchmark them accordingly.
%d bloggers like this: