Weekly Writeup for AC233 Database Communication and Management

Data Marts Valued by Financial Institutions

(Adopted from Business Times - my personal archive)

Financial institutions typically run a myriad of disparate application systems to support their day-to-day operations. A major reason for this situation is the sophistication and time involved in developing a full banking application that supports all banking units and covers all products.

As a result, most banking institutions purchase different off-the-shelf packages for different business needs. Other reasons include geographic disparity and regulatory requirements on financial data within each country.

But these disparate application systems make it difficult for decision-makers to obtain integrated information that is needed for strategic and tactical decision making.

A solution to this problem is to develop a data warehouse that serves as the foundation of all decision-making needs. But before developing this data warehouse, the financial institution must put in place its enterprise data warehouse strategy.

This is important as organisations may run the risk of acquiring redundant data, wasting host resources and increasing warehouse maintenance overheads.

Enterprise architecture

This is because most organisations will find large warehouses too complex and time-consuming to build. It often makes sense for organisations to resort to building data marts.

A data mart is defined as a departmental data warehouse that usually supports a specific business need, for example risk management.

The benefits of adopting the data mart approach is its speedy implementation, responsiveness to changes in business needs and flexibility in structuring data models according to a specific business requirement.

Without an enterprise architecture, these individual data marts may still result in islands of information -- just the type of situation that a financial institution would want to avert from the beginning.

The enterprise architecture serves as a plan and a road map for managing the data warehouse environment over time.

It addresses issues arising from increased usage, growth, and functionality, and ensures the coordination of data acquisition and warehouse processes. Only with this structure in place, would the decision-maker have the integrated information to make strategic decisions quickly.