• James Hunt, Senior Finance Industry Consultant at Teradata


• 17.12.2015 09:31 am

The torrent of regulations designed to manage risk in the international banking system shows no sight of slowing down, and threatens to overwhelm banks and regulators alike.

For a start, it is clear that many of the world’s largest banks are struggling to comply with BCBS 239 – the principles-based regulation drawn up by the Basel Committee. Unlike many other banking regulations (which focus on specific calculations, reports or alerts) BCBS 239 deals with the quality of the bank’s governance, data architecture, IT infrastructure, analytical and reporting capabilities which are used to support all of their risk and regulatory reporting.

More than 50 per cent of banks deemed to be globally important by the Basel Committee have told the regulators that as things stand, they are materially non-compliant. A third said they did not even expect to meet the January 2016 deadline set to meet BCBS 239 requirements.

Fragmented systems cannot cope

Given the fragmented IT set-ups in many banks, it is hardly surprising that they are finding it difficult. They are often dealing with hundreds of data marts and interfaces related to risk and regulation – a complex and costly legacy from years of implementing each new piece of regulation on a siloed basis.

Of course, most of them are now putting in place centralised data warehousing and analytics capabilities, however reversing the history of fragmentation is not an easy task.

Anyone who can remember the City of ten or 15 years ago will know how regulatory compliance has changed in complexity and quantity.

Looking back with rose-tinted glasses to those pre-Basel II days, regulatory reporting consisted of, well, reports; pieces of paper (or at best, electronic versions of paper) with relatively simple summary-level metrics. Today, the FINREP and COREP reporting for CRD IV consists of tens of thousands of data points – the result of millions of complex risk and capital calculations – reported through a complex XBRL taxonomy.

And the future points to even more detail - with new requirements from the European Central Bank such as Anacredit indicating that the regulators will expect to receive data on individual exposures (up to 175 attributes on all exposures above a threshold of €25,000).

Whilst Anacredit is initially a Eurozone regulation, some non-Eurozone countries such as Sweden and Denmark have already indicated their intention to participate, and Anacredit clearly signals that future regulatory reporting will be based around the supply of large and granular data sets on a monthly basis.

As well as more detail, organisations also have to deal with greater complexity and more rapid change. For example, banks which thought they had dealt with FATCA (the U.S. anti-tax-dodging regulation) have been disappointed - although perhaps not surprised - to find that they now have to meet similar requirements spanning more than 70 countries under the upcoming ‘Common Reporting Standard’. Organisations which implemented FATCA on a siloed and non-scalable basis are likely to face difficult decisions around how to accommodate these changes.

Integrated architecture – lower cost

So, given the plethora of sometimes overlapping requirements across regulation, compliance, accounting, financial and management reporting, how can organisations meet these challenges in an efficient and cost-effective way?

Most forward-thinking banks have recognised the need to simplify their data architectures to make reporting easier and cheaper, implementing one repository, an Integrated Data Warehouse.

Of course, the regulations too, have been pushing banks in this direction. It can be argued strongly that some of the key principles in BCBS 239 are almost impossible to achieve in an environment based around fragmented and siloed data marts.

Banks such as UniCredit, based in Italy or Raiffeisen, in Austria - both with extensive multi-country operations across central and Eastern Europe - have implemented integrated data warehouses so they can meet the regulations in a much more agile and low-cost way from a single source of data. This is also benefiting them in terms of monitoring internal risk and financial management.

Achieving this is no small undertaking, but UniCredit and Raiffeisen are already reaping the benefits of this approach. Raiffeisen recently described how their integrated data warehouse has reduced the cost of meeting one major new piece of regulation by over 70%, as well as avoiding long-term costs by removing the need to maintain 17 data feeds, and also speeding the time to delivery.

Regulators face the same challenges

The other side of all this regulation is that the regulators themselves need to be able to accommodate all of this new data which is being sent to them.

Organisations such as the Bank of England and the Financial Conduct Authority currently collect FINREP, COREP and other regulatory and compliance data, however Anacredit (assuming UK participation at some point) will represent a step-change in the data volumes which they need to manage.

The regulators will be facing the very same challenges which the banks themselves have been grappling with over recent years. i.e. how to deal with hundreds of millions or billions of rows of data; how to integrate this with structured data from many other sources; and how to perform insightful analytics across all of this, as well as other multi-structured data (such as economic reports or Reuters feeds).

Powerful analytics

To gain a better understanding of all this data the regulators will have to build a similar analytical capability to the leading banks, applying the same principles of timeliness, accuracy, quality and flexibility which are espoused in BCBS 239. They too will require scalability and capacity to run different types of analytics on disparate kinds of data, giving them a truly detailed picture rather than a lofty overview.

And, if a crisis did loom, they must be capable of quickly running deep analytics to identify which banks need their attention in the near term and further down the line.

The most obvious solution to the difficulty of handling and analysing all this data is for banks and regulators alike to build integrated repositories where information can be stored and analysed quickly and cost-effectively.

Such platforms can adapt to huge increases in volume and achieve high levels of performance while handling mixed workloads- for example running standard reports at the same time as executing heavy-duty risk models in-database and also answering rapid-response tactical queries.

Experience counts

However, this integrated data architecture needs to be designed and installed by practitioners experienced in building integrated data warehouse and analytical infrastructures. Yes, the technology is important – however aspects of good design, governance and organisational change do not come ‘out of a box’ and are just as vital.

Using such expertise, a phased programme of work will start to deliver value very quickly from the first drop of data, and ultimately deliver an efficient and flexible architecture which results in data being captured, validated and stored once, whilst supporting  a whole range of flexible reporting and advanced analytics downstream.

The continued change and evolution in world-wide regulation means that if a bank, or indeed a regulatory institution, is not building with an eye to the future, then it is going to struggle to meet these future challenges, and is effectively throwing money away.

BCBS 239 may represent a major challenge, however making the changes to data and analytical architectures which are necessary offers a rare win: win for organisations. Better quality, flexibility and agility to meet future changes, lower costs, and the opportunity to re-use the same infrastructure and data for many other value-generating activities.