A single view of risk

Enterprise Risk Management enables banks to measure and manage risk across the entire company in a centralised, less departmentalised manner, saving your business time and money while minimising the likelihood of any nasty surprises.
Jim Banks finds out how from Simon Doherty, Teradata.

In business, as in life, risk is a constant factor. Effective risk management requires, above all else, a solid grasp of what is going on across your organisation and the ability to successfully interpret it. The financial sector has always taken this issue seriously, but competitive pressure and a series of regulatory initiatives have made streamlining existing systems a growing concern. Banks have traditionally managed systems that are fragmented along operational and product lines. Now, the guidelines laid down by Basel II, as well as other initiatives such as Sarbanes Oxley and equivalent legislation, are making such operating systems obsolete. The industry is required to measure, mitigate, manage and report on risk in ever more sophisticated and proactive ways, faced with the need to be able to produce information on request, both for regulators and internal purposes, in a shorter timescale than ever before. One cannot afford to be anything but proactive.

An appropriate infrastructure that enables the straightforward use of all data in an effective and efficient way is fast becoming a must. The risk data architectures that have evolved in previous years can make this seem rather a daunting task. Each new application or set of business requirements often results in an additional data mart being put in place which, in turn, leads to spiraling costs and increased complexity. The fallout from Basel II is a case in point, with the costs for many banks to install new Basel II data marts proving to be immense. This is despite the fact that the vast majority of required data will have already existed on other data marts contained within the organisation.

Integrated warehousing

So, how best to proceed? In the opinion of Simon Doherty, head of risk practice for Teradata in Europe, Middle East and Africa, an integrated, centralised, forward-looking approach is essential. ‘The industry needs to move away from the complex and costly approach of maintaining different data marts for different functions and look for a sustainable, integrated data warehousing system – a single view of risk,’ he explains.

Teradata, a division of NCR, is a leading data warehousing provider. The company has a history of helping companies make faster, better decisions that drive profitable growth. It has now developed a dynamic and scalable platform for integrating, managing and analysing data, serving the ERM needs of major financial services organisations. The platform boasts unprecedented decision support capability for addressing the short and long-term health of your business and Teradata’s active data warehouse puts right-time, actionable information in the hands of decision-makers on demand. It also supports event-based management, enabling the establishment of thresholds and criteria for triggering reports as information comes in from business lines across the enterprise.

‘Our approach to enterprise risk management focuses on the practical data and analytical issues which are challenging banks today,’ says Doherty. One of the greatest challenges in constructing an effective ERM system is identifying and gathering relevant information from various departments across the organisation. In many cases, data may reside in various formats, under separate owners, in separate systems across the enterprise. Teradata’s platform eases this process, but Doherty knows that the gathering of data is only half the battle – information means nothing if it is not properly correlated.

‘Of equal importance is the logical data model [LDM] which is used to structure the data,’ he explains, ‘A comprehensive and well structured data model that reflects the underlying business structure is essential.’ The adjacent figure suggests one such model.

Business-oriented model

Teradata’s already comprehensive financial services logical data model (FS-LDM) has been extended
to include new areas such as financial management and Basel II, and the progressive nature of the format does not stop there. A major benefit of the platform is its ability to physically implement a business-oriented LDM. ‘Many platforms have to compromise in this respect to meet performance constraints,’ says Doherty, ‘This has a consequent detrimental impact on business users.’ Doherty highlights four major areas of benefit in ERM approach: a single view of risk; more timely risk reporting and analysis; lower total cost of ownership (TCO) compared with a fragmented data mart approach; less reconciliation and data validation.

The consolidation of data into a single view is one of the key drivers for an ERM approach and Doherty believes that the importance of this cannot be overstated. ‘Through ERM, analysis and reporting can be undertaken consistently and efficiently without all of the manual data collection processes,’ he explains, ‘This single view has serious benefits for the bank’s regular analysis and reporting requirements.’ One of these ‘serious benefits’ is the ability to answer new questions through ad hoc analysis a lot more quickly. The new regulations coming into effect can leave little time for analysis gathering; answers must now be supplied on demand. Once data is available from operational systems, it is loaded directly into the Teradata enterprise risk data warehouse where it is becomes available immediately. ‘This contrasts with a data mart approach, where data often passes through a web of different marts and transformation processes before it can be used,’ explains Doherty.

One validation process

Data validation and reconciliation can be a massive hidden cost in many organisations and the Teradata platform seeks to rectify this. ‘The same data is often checked and reconciled multiple times, and to different quality standards across different data marts,’ says Doherty. ‘An ERM approach enables data to be validated and reconciled once, which saves time and effort, improves confidence in the quality of the data and enables the data to be available to users more quickly. Perhaps the easiest way of quantifying the importance of ERM is not in how much money it will make you, but the amount it can potentially save.


The reduction in hardware platforms, data duplication transformation processes and support costs are often substantial. ‘If compared on a like-for like basis,’ says Doherty, ‘the costs of operating a single enterprise risk data warehouse will always be less than a fragmented data mart approach.’ Your business will also find itself far better prepared for any bolts from the blue. Historically, the field of risk management has been dominated by theoretical discussions and indecipherable algorithms. ERM and Teradata’s platform simplify matters, making the process a lot more comprehensible and focused. ‘We concentrate on the practical data and analytical issues which are challenging banks today,’ explains Doherty. ‘You can’t move from a fragmented data mart platform to an ERM approach overnight, particularly if a very large number of data marts have built up within the organisation. However, it’s vitally important that banks start to move in the right direction, otherwise they will certainly find themselves at a serious disadvantage to their competitors.’ Now that does not sound like a risk worth taking. FBA

Simon Doherty

Simon Doherty is Teradata’s head of risk practice (Europe, Middle East and Africa). Prior to joining Teradata, he worked for 18 years for NatWest, a large UK clearing bank now part of the Royal Bank of Scotland Group. His final role at NatWest was as finance director and chief operating officer of the bank’s international trade and banking services division. During his time at NatWest, he was responsible for developing and supporting a wide range of decision support systems and data warehouses across a number of business areas including the retail, corporate and international divisions.