The Fundamental Review of the Trading Book (FRTB), introduced by the Basel Committee on Banking Supervision (BCBS), will transform how banks measure risk. FRTB is designed to address some fundamental weaknesses that did not get addressed in the post-2008 financial crisis regulatory reforms. In order to help make banks more resilient to drastic market changes, it will impose capital requirements that are more closely aligned with the market’s actual risk factors. The changes introduced by FRTB are substantive, which is why the regulation has been delayed several times, even before the latest delay attributed to the global pandemic. At a glance, here are the primary changes FRTB introduces:
- A shift from value-at-risk (VaR) to an expected shortfall (ES) measure of risk under stress. This is likely to increase capital requirements by 22%, on average, according to BCBS.
- A revised boundary between the trading book and banking book. This will increase scrutiny of the trading desk(s) at a more granular level.
- A revised internal models approach (IMA) combined with an overhaul of the standardized approach (SA). For instance, banks are now required to conduct quantitative impact studies to compare the two methods for their trading business overall. This must then be repeated for each individual trading desk portfolio.
- A means of incorporating the risk of market illiquidity, including liquidity horizons that range from 10 to 250 days. Continuous monitoring will be required, and banks will need to conduct back-testing to ensure accuracy.
The new capital requirements are currently due to take effect in January 2023.
Operational Impact of FRTB
From a data management point of view, FRTB’s requirements will require greatly increased quantities of historical data, along with an increased need for analysis and intensive computation against this data.
FRTB also puts a spotlight on all the data sources that a company may be utilizing and on the management and governance of the data used in complex models.
Banks and insurers facing the FRTB transition need to ready themselves to manage an increased variety of data, including new sources, both real-time and batched. For example, banks may need data from external sources like Bloomberg to supplement trading data they already have on hand — and these external sources will likely not conform to the same data structures as the internal data. There will be an increased volume of data storage required, due to the longer history needed by the ES approach to risk measurement. And there will be expansions on the requirements for managing and monitoring both data lineage and data security. These requirements will stretch the data collection and modeling requirements beyond previous regulations seen
Here’s how we estimate FRTB will impact computational, storage, and data management:
- 24x increase in historical data storage.
- 30x increase in computational requirements.
- Expanded requirements for a centralized and secure single view of risk data.
- Complete audit trail(s) for data, models, and risk calculations.
- A machine learning ops framework that supports regular backtesting and P&L on attribution testing.
FRTB Demands a Streamlined Architecture
In order to support a transition at this scale, banks will need to establish internal team alignment first. A single, enterprise-wide platform or coordinated approach for storing and maintaining data can facilitate alignment between the front office, risk management, and finance, setting the stage for a more seamless transition.
Banks will also need a standardized workflow for the production of P&L and Independent Price Verification (IPV) metrics on a daily and monthly basis, as these will drive much of the information gathering that powers FRTB models.
Universal adoption of a single pricing library will be required.
And organizations should enforce full adoption of a complete, “golden” source of static data.
Ideally Suited for the Hybrid Cloud
FRTB’s data and compute requirements are ideally suited for deployment in hybrid cloud environments. The flexibility to scale up and down as needed can help facilitate efficient modelling. Container technology enables the clean separation of storage and compute resources, making it simpler to manage the increased requirements of FRTB and to scale gracefully as needed The cloud also offers flexible options for comparing and assessing models.
The Cloudera Data Platform (CDP) offers financial institutions a unified data and analytics platform to manage and model risk exposures to drive improved risk management and enable risk digital transformation. CDP supports big data, machine learning, and predictive analytics to enable improved risk modeling and compliance support for FRTB and across the enterprise to support CCAR/EBA Stress Testing, BCBS-239, MiFID II, IFRS-9, IFRS-17, etc.
To learn more about how Cloudera helps to support the FRTB requirements, read our FRTB solution brief.