The second Spotlight Review published by the FICC Markets Standards Board (FMSB) highlights the important role of data in financial markets and stresses that if markets are to remain stable and trusted, the rapid growth of new technology and of data science must be balanced with ever more effective governance and control.
The latest paper – the first looked at algo model risk – highlights seven sources of data risk; namely, business continuity and operational risk, security, trading, exposure, regulatory, ownership and rights, and conduct. “Failure to exercise appropriate controls over data increases the significant risks to organisations and market functioning, from both a reputational and financial perspective,” the paper states, adding, “…the need of firms for secure, accurate, timely, complete and consistent datasets has never been greater.”
To mitigate these risks the paper highlights an increasing focus on the importance of developing standards for data and on strengthening governance around its use. While it acknowledges that data has increased transparency and improved overall efficiency of FICC markets, it also warns of remaining issues with the clarity, consistency and ease of its use.
“A standardised approach to data governance can be effective in controlling the risks associated with sourcing, storing, processing, transmitting, selling, and using data,” the paper states. “Different industry initiatives such as the Financial Information eXchange (FIX) Protocol and Legal Entity Identifiers (LEIs) have been successful in standardising the way in which financial data is transferred and counterparties and clients referenced from one system to another reducing risks arising from the data being ‘lost in translation’. But more is needed to manage the risks arising from the ever-increasing role of data in financial markets.”
The paper says that the post-financial crisis regulatory response has been good, however it also notes that much of the new data eco-system that resulted has not been optimised because of a lack of data quality and standardisation. “This is particularly evident in the context of Markets in Financial Instruments Directive (MiFID) II where the lack of specificity in post-trade transparency requirements has led to inconsistencies in the disclosures provided by market participants, which in turn has impacted their usefulness in promoting fair and effective markets,” the paper states.
It further argues that the increase in data required by multiple regulators has crystallised the need for stronger data governance and standardisation of how it is handled within firms. As the paper explains, there are disparate models for data reporting between regulators and government agencies in any one jurisdiction, as well as significant variance between countries or regions, which lead to inconsistencies, increased costs and requiring detailed mapping. For example, in the context of swaps, derivatives exposure is measured on a notional or risk-adjusted basis. “There would be significant improvements in risk monitoring and efficiencies in terms of cost, for both firms and regulators, from more streamlined and rationalised regulatory reporting,” it suggests.
While laying out the different initiatives underway to build data harmonisation, the paper does acknowledge that no one size fits all, however, it argues that firms can gain long-term benefits from a focus on the important building blocks and areas of developing best practice in the design, construction, and maintenance of data architectures.
“One of the first considerations for any market participant is the degree to which it should pursue a centralised approach to enterprise data architecture for coordinating the movement, enhancement, integration, quality, and availability of data,” the paper says. “In principle a centralised approach would integrate fragmented data and streamline the data architecture in the most efficient way.
“However,” it continues. “For large firms full centralisation can also be very expensive as well as unnecessarily cumbersome at the individual business unit level. An alternative ‘hybrid’ approach may be a preferable strategy. For instance, a consistent centralised data architecture is important across overlapping businesses that have common clients, counterparties and data needs.”
Such an architecture may be less relevant when looking at very different business lines with no overlapping client base, the FMSB accepts, but “It is also important for firms to build a data governance strategy that can deal with inconsistencies across an ever-wider array of external data sources including other firms, vendors and regulators.”
A data governance theme of increasing cross-industry importance is how to control appropriate internal and external use and sharing of data, it adds, arguing, “Data governance must ensure the consistency, timeliness, security and delivery of data. The paper delivers eight key components to promote effective data governance and standardisation. They surround the creation of a data lifecycle, data policies, taxonomies, mapping, movement and lineage, classification, leakage protection, and data quality.
“There is a clear need for robust data governance and management strategies across firms who are active in rapidly changing wholesale FICC markets,” the paper concludes. “Market participants are increasingly moving to more centralised data strategies, with significant scope for efficiency gains in terms of longer-term cost reduction, risk reduction, and allowing better use of data to drive value for commercial purposes.
“At the same time, in the near term this can be a costly and complex exercise with centralisation creating most value in overlapping parts of a group in terms of common customers, counterparties and business models,” it continues. “Therefore, the magnitude of centralisation and the specific structures required will vary from firm to firm depending on size, complexity and business models.
“The seven sources of data risk outlined in this Spotlight Review do not impact all firms equally, so firms need to undertake careful analysis of their own organisations, systems and processes in order to consider how best to combine the eight key components to promote effective data governance and standardisation discussed above,” it adds. “Moreover, there are some areas that the market can address collectively to avoid duplication and inconsistency of approach. This would help to build transparency and consistency – regardless of organisational and international boundaries – which are so important to ensuring the fairness and effectiveness of global FICC markets.
“Data standardisation involves complex problems and potentially significant changes to infrastructure that will not be resolved quickly. As an industry, we must ensure that efforts at standardisation deliver the intended benefits, improving efficiency and effectiveness for market participants and end-users in global wholesale FICC markets.”
The paper has been welcomed by Gareth Ramsay, executive director, data, at the Bank of England, who says, “The rapid pace of systems development and the increasing dependence of the financial system on data is a key strategic focus for the Bank. Therefore, the FMSB’s Spotlight Review is a welcome exploration of the critical role of data management for the financial industry. The key themes in the review, particularly the importance of data governance and standardisation, complement the Bank’s ongoing data collection review, which aims to make data collection more efficient for the Bank and for firms, and to improve our ability to use that data effectively.”