Of all the new incoming regulations, few will have a greater impact on financial institutions’ trading business than Fundamental Review of the Trading Book (FRTB).
This Basel Committee on Banking Supervision update to the current Basel 2.5 introduces stricter rules for the treatment of market risk across national jurisdictions and addresses some of the weaknesses of the current framework. In practice it could result in higher capital requirements to protect against trading book positions. These requirements were finalised in January 2016 and are now in the process of being adopted at a regional and local level.
Notable regulation changes include the following:
- A stricter separation of the trading book and the banking book is required. Banks will need to review their portfolios to determine whether existing classifications of instruments are correct.
- How models are governed has evolved. Banks that want to use internal models must pass a set of tests to bypass the standardised approach.
- The current Basel 2.5 internal models approach has been made more rigorous. Regulators can now provide approval for internal models at the individual trading desk level instead of at the entity level.
- Modeling risk for regulatory reporting has changed. FRTB mandates that core market risk calculations be based on Expected Shortfall (ES) as opposed to Value at Risk (VaR) methodology.
The need for data
Banks will need access to significant data and analytics to meet requirements. Most large banks will likely use an internal model to access three sources of external data: time-series market data, reference data and “real” price data. Large quantities of internal data are also required as part of capital calculations.
For banks using internal models, time-series data is a critical input to the capital models. FRTB mandates that banks conduct ES and Default Risk Charge calculations based on data sets that go back to 2007 and 10-years, respectively. Because current methodologies use a shorter calibration period, many banks don’t have access to time-series data stores back this far. These banks will have extra work to do in order to collect high-quality, time-series data.
Irrespective of whether a bank is using internal models or the standardised approach, accessing the right reference data is a critical requirement of its FRTB response. High-quality and complete instrument reference data allows banks to accurately classify their instruments with the regulatory framework.
Another large data challenge results from the introduction of the Non-Modellable Risk Factor (NMRF) requirements. This specifies that banks using internal models must pass the NMRF test if they want to calculate their capital based on an ES calculation. For this test, banks are required to determine whether they have continuously available “real” prices for a sufficient set of representative transactions.
The requirement to conduct the NMRF test creates two challenges for banks. Firstly, for many OTC markets, executed trade and committed quote data may not be readily available. Secondly, once the data has been sourced, it needs to be organised into a consistent set of data models to be mapped into risk factors in order to determine whether they pass.
Synergy with MiFID II
FRTB compliance will have certain synergies with the upcoming European Union legislation, the Markets in Financial Instruments Directive II (MiFID II). This regulation comes into effect in January 2018, prior to FRTB.
MiFID II mandates that trading platforms as well as buy- and sell-side firms have obligations to publish trade details. The obligations cover cash and derivatives across all asset classes, subject to an exception for spot FX markets. It is this extensive source of information that could be used as a source of “real” price data under FRTB.
Despite the significant benefits that MiFID II post-trade data should bring in determining the NMRF requirements, it is not a true solution for two reasons:
- Data models in MiFID II are not necessarily aligned with those proposed for use within FRTB, so banks seeking to use MiFID II post-trade data for their “real” price observations may need to map the data into a new data model.
- Post-trade rules require each trading firm to publish trades that they execute, and each buy- or sell-side firm to report through a new type of regulated entity called an Approved Publication Arrangement (APA). The splintered nature of OTC markets in combination with the competition between different APA providers means that the MiFID II post-trade data will be fragmented.
Changing the nature of the game
No bank has ever before had to collect the wide-ranging type of data for this type of regulatory purpose. Banks are responding to FRTB by participating in Quantitative Impact Study and industry working groups as well as undertaking proof of concepts. Some more advanced banks have started their implementation phase through either undertaking request for proposals or implementing new systems.
Over the long term, banks will require a high-quality data vendor that can provide a broad range of market data, reference data and “real” price data. Taking these necessary actions to remain compliant will help banks avoid a negative impact on their business.
Download the white paper at: bit.ly/FRTB-data-demands