Regulatory demand for greater insight into banks has pushed financial institutions to seek out scalable and low-cost data processing. 

The number of changes banks face as a consequence of regulation are enormous. By way of example, as of December 2, 2013, 280 rule-making deadlines have been passed for the wide-ranging Dodd-Frank Act in the US, since it was signed off in July 2010. While few jurisdictions are as advanced in their own rule making, nevertheless regulators across the globe are asking for more granular levels of information, on a more frequent basis than ever before.

Such demands are testing the limits of database technology and so banks are examining new platforms that can accommodate and process data sets that could not be run through a traditional database, so-called ‘big data’ technology.

“There are some areas where we are seeing big data being applied in its purest sense,” says Dan Higgins, financial services partner at consultancy EY. “It is early days, but banks need to change their technology model. There are some that are beginning to take the big data concept and apply it in anti-money laundering [AML] and compliance functions, using what they call meta-alerting. They combine holistic alert data, unstructured external sources and so on, to produce a better informed view of the client and the transaction.”

Opening doors

By allowing unstructured data sources to be checked in parallel, distributed across many low-cost servers, big data technology makes a faster and cheaper alternative to old models of data storage and retrieval, while opening the doors to a greater range of information sources. That can make risk calculations faster, cheaper and more flexible.

“There has been an explosion in what is possible in data and analytics,” says Alasdair Anderson, global head of IT architecture at HSBC securities services. “For 20 years there have been relational database management systems and nothing else. Then a couple of guys in Silicon Valley said, if you were able break out of that mould and could scale a process with different kinds of data and greater amounts of data, and run 'what if' analyses across it, then you have a whole business creating product out of data. That is one of the key technology trends that we are seeing.”

The extent to which banks are under pressure from new rules varies by geography. The US is admittedly ahead of the curve; Dodd-Frank is the spearhead of post-crisis capital markets reform. Europe and Japan are still hammering out the details of their equivalent market reform regulations, while other countries are discussing how to respond.

Capital adequacy rules have been more widely adopted. By April 2013, 14 of the 27 Basel Committee member countries had issued Basel III-based capital regulations, acting as a major driver for data management reform.

“Everything that has to do with Basel is a big driver, especially the Basel Committee on Banking Supervision regulations around risk aggregation reporting,” says Hyong Kim, financial services partner at EY. “There are new kinds of data capture, new data quality measures and new kinds of reporting. Clients are using that as a driver to re-optimise their data infrastructure.”

Cause for compliance

The burden of compliance with new rules so far falls upon firms operating in well-developed markets, as these countries are typically in the lead when it comes to enacting new post-crisis policy. In addition, the more expansive a firm’s operations, the more regulations it must comply with. But authorities are increasingly making examples of firms that transgress existing rules. Their enforcement of AML rules and the beefing up of anti-tax evasion programmes has led to massive fines for banks found wanting.

In 2012, HSBC was hit with a $1.9bn fine for allowing cash to cross into the US from Mexico with inadequate checks in place. The same year, Standard Chartered paid $667m for processing transactions connected to Iran during a period in which the country was under economic sanctions. ING also paid $619m for historical breaches of AML and know your customer rules in 2012.

Authorities would have the ability to ban firms from continuing to operate in the US should they be perceived as unrepentant or repeat offenders, and that is driving them to make their processes watertight. Meta-alerts would enable banks to track individuals between geographies, using a wide range of data sources to check that they are who they claim to be and monitor their activity in order to identify suspicious behaviour, minimising any chance of a criminal transaction slipping through.

“There is a philosophical change as to what constitutes relevant data to be analysed, and there is a tooling change that is required to the traditional architectures to allow for more efficient processing of very diverse, higher volume and more loosely connected data,” says Mr Higgins. “It doesn’t stop at the technology model. If you bring all of this together it requires a skill and capability change as different skills are needed to architect informational transactions as well as financial transactions.”

New model data

What will enable banks to manage this data montage is precisely a breakaway from the adaptation of financial transaction data models and associated hardware.

“Things such as AML or understanding your customer are all about spotting patterns over time,” says Professor Mark Whitehorn, chair of analytics at the school of computing at the University of Dundee.

This is not what SQL (structured query languages) relational databases, which banks use for tracking and recording financial transactions, are best at. Structuring a search in SQL requires a skilled programmer and can be time intensive. Instead, models such as Apache Hadoop, an open-source platform that was based on the Google storage and retrieval system, are being investigated, as they allow searching across much more varied sources of data, with searches constructed far more simply, just as one sees in search engines.

“The smarter banks are saying: 'We tried to do this with SQL and we cannot'," says Mr Whitehorn. “In fact, you can do anything with SQL but at some point it becomes so hard to do it with SQL and so much easier to do elsewhere, there is no good reason to continue with SQL. It's a matter of cost versus payback.”

A question of trust

As open-source technologies are typically not tried and tested to an industry standard, they are not stable enough for banks to trust with sensitive data. As a result, banks are at early stages of development with Hadoop-based technologies, exploring how they can best interact with existing database and data warehousing installations. 

“Once you have built up expertise with things such as Hadoop, you can do some interesting things,” says Mr Whitehorn. “Data from a data warehouse is assumed to be clean, audited and has an agreed meaning. Without those three, any analytics is worthless. There is a lot to be said for taking the data from a data warehouse and writing queries for it in Hadoop, because it is orders of magnitude easier than writing them in SQL.”

To effectively build these technologies into their architectures, firms need to develop a strategic model approach to data management that will allow an evolution of existing platforms, and an integration of new systems.

"Banks are seeking to deliver enterprise-wide data governance and so they are creating chief data officer functions to take responsibility for managing data standards, data policy and compliance with those policies,” says Mr Kim. “Those functions are looking at various components around the solution and are responsible for defining the data architecture across the bank. These enterprise data architectures are being used as the blueprint for pulling together information needed for each regulation and defining how each reporting obligation will be met."

PLEASE ENTER YOUR DETAILS TO WATCH THIS VIDEO

All fields are mandatory

The Banker is a service from the Financial Times. The Financial Times Ltd takes your privacy seriously.

Choose how you want us to contact you.

Invites and Offers from The Banker

Receive exclusive personalised event invitations, carefully curated offers and promotions from The Banker



For more information about how we use your data, please refer to our privacy and cookie policies.

Terms and conditions

Join our community

The Banker on Twitter