By Nigel Pickering, Founder, 1View Solutions
Designing a data warehouse or consolidated database is next to impossible in a big company, or at the very least prohibitively expensive.
The first step on the journey is defining the database that the user community wants or needs. While that sounds simple, for a large organisation it is anything but straightforward. The IT team leads the process, asking questions such as ‘what fields do you want in your database, what do we have – what indices, tables?’ and so on. The process of database design therefore becomes very subjective, relying on the skills and personal preferences of those building the database, as opposed to being driven by the users.
Compounding the problem is that when you have a team of 20 or even 100 people that are brought together to design the database, you are by its very nature ‘designing by committee’. The result of this is a very long, drawn-out process. By the time the database is built, the needs of the business will have changed and the database will be out of date. It will also be an inflexible system: if it has taken two years to design the data warehouse and then another year building it, there will be substantial resistance to any change.
The issue is that legacy design methods don’t fit today’s business models. EDM is a mature approach to technology that worked for systems development 20 years ago. Many now view EDM as an expensive and technology-driven exercise that is intensely time-consuming. For many COOs of investment banks, the impact of following the EDM route has been failed projects and massive cost over-runs. They have also rarely been successful from the user’s perspective because they take so long to complete. That approach isn’t viable now, but what is the answer?
Configuring your own data independently
One solution is the Collaborative Data Management (CDM) model. With CDM, every data owner is able to independently contribute and configure his or her own data into a consolidated model. It is built upon the simple concept that the database grows as users put more data in. Each user simply uploads the data that he has in his system and a consolidated database emerges. Users can then add more data as business needs change.
CDM also means that all users can readily view any anomalies in the data that must be fixed. The data owners benefit from highlighted exceptions, showing where their data has different meanings from other contributors.
Where the same data elements are supplied by multiple sources, irrespective of local formats, each contributor can determine whether their data is valid by transparently viewing all the values from other available sources.
How CDM can help
The outcome of CDM, apart from group-wide data consistency, is a continually maintained and consolidated data superset of all contributing sources.
Two examples of current challenges that a CDM approach could significantly help investment banks with are:
1. Regulatory reporting There are currently many regulatory reporting objectives (such as EMIR’s OTC derivatives trades or the pending EU Financial Transaction Tax). The simplistic requirement of all these regulations could be summarised as an extract list of trades in a particular category(s). If the banks are having a problem satisfying the regulatory reporting requirements, it is because they don’t have a consolidated database of all data, positions or trades. As a consequence, solutions tend to solve each regulatory project independently. This means immense duplication of effort and further reconciliation needs. CDM represents a significantly quicker, simpler and longer-term solution.
2. Legal Entity Identifier (LEI) For an investment bank with, say, 100 systems around the world each holding company information, the bank will have to merge that LEI data with all 100 systems. Conversely if those 100 systems imported their data into a CDM database, they would receive the LEI individually, thereby achieving an ongoing solution within days.
The ability for any large organisation to manage and use every piece of its operational data will differentiate the winners from the losers.
CDM is by far the quickest, simplest and most adaptable way to deliver any project that requires the consolidation of data from a number of source systems. It is a pragmatic solution and highly efficient in terms of time and cost.
EDM has served the investment banking industry very well for a long time, but it has finally run its course.