Complying with financial regulations is not an easy task, but it is one that is made much easier if you already have control of your business-critical information.
In most cases, the quality and supporting infrastructure of an institution’s data is fundamental to achieving compliance and the Fundamental Review of the Trading Book (FRTB) is no different.
The first thing is to ensure you understand your data
FRTB, part of the Basel III rules aimed at significantly transforming the way banks manage their capital requirements and how they are structured and managed internally, necessitates a number of specific data challenges in terms of understanding the data, the volumes of data required, the quality of that data and most importantly, tracking the flow of the data. One example relating to understanding the data is the distinction of the boundary between banking and trading books. Transactional data needs to be strictly classified into one of these two categories, with robust governance to ensure compliance with the rule set making that distinction.
Watch out for data held in spreadsheets
Another concern is the real-time reporting requirement for intraday risk and the comparison between risk management and pricing models. In addition, the management of reference data in terms of client, book, instrument and market remains a challenge with respect to maintenance, especially as the volumes of data increase. As with BCBS 239 and other banking regulations, reliable golden sources of data that can be independently validated are crucial to achieving compliance.
Spreadsheets are often used to collect front office prices/marks – and this is where problems may arise. Data is typically passed to internal pricing services where risk factors are generated. These marks are then passed to the enterprise data repository where additional validation is applied and the EOC, or golden copy, is compiled and published to downstream applications for their close of business processing and reporting. Under the FRTB regulations, organisations will need to demonstrate lineage between the trader marks, the risk factors generated, and real price observations seen in the marketplace, so the accuracy of trader marks (which may originate in spreadsheets) and an ability to demonstrate alignment, is essential.
So how best to achieve this?
At the root of this ultimately very sensitive data, are the spreadsheets that the front office uses to collate and publish their view of the products they trade. Of course, there are checks and balances both in the pricing service and the enterprise data repository, but the marks also need to prove they are representative of the prices in the market (which could be very illiquid). The fundamental accuracy of the marks and the maintenance of the lineage of those marks is therefore of paramount importance.