Wall Street & Technology is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.


09:53 AM
Gregory E. Smith
Gregory E. Smith
Connect Directly

All That Glitters

What is involved in creating and maintaining a "golden copy" of reference data for trading and settlement, risk management or other critical applications at your firm? Should you "buy" or "build" a solution?

A Matter of Trust

The concept of the "golden copy" is the notion of creating an integrated, trusted set of reference data for all requirements of straight-through processing (STP) within one organization. The need for a trusted set is created by the fact that STP depends not on one single application but a series of applications, each of which requires access to instrument, settlement, routing, trade, counterparty and other data. Compounding the problem are the requirements driven by Basel II and anti-money laundering rules.

All That Glitters is Not Gold

The challenges in creating the golden copy are many, including (1) no single source of all the required data exists, nor can it, as data originates from both internal and external sources; (2) there are many categories of error, including outdated data definitions; (3) each organization is unique, with its own applications, policies, asset mix and relationships; (4) there are always caveats associated with any external sources; and (5) instruments, issuers and associated data are not static, so the golden copy tarnishes quickly if not continuously polished.

The solution is really a "golden process", an infrastructure, a consolidation of problem ownership and a commitment - with an eyes-open account of the complete cost of bad data. This last point is crucial, as often the full cost is spread, with some organizations bearing total cost in the hundreds of millions globally. The organization that does not know its true full cost will inaccurately size and poorly fund its effort to create a trusted golden copy.

A successful process has several cornerstones, including:

Definitional integrity. Each asset class, category, type of trading and source has its own internal definitions and logic.

Data validation routines that are dynamic, tailored to source, type and subject.

Procedures. About one-fourth of the cost of an error is finding it, whereas three-fourths is fixing it, so procedures for handling errors are critical. Prioritizing corrections against impact significantly reduces ongoing costs.

Audit trail. The process must be transparent so that the life and source of every data element can be traced. This is critical for monitoring data quality, conducting operational risk and data license audits, and sorting out failed trades with any organization that has not solved its data quality problems.

How to Get There?

Given the ramifications and long-term commitment involved in maintaining reference data, it is critical to carefully consider what will constitute a solution for your organization, including whether it is best to "buy" or "build".

Cicada's experience is that creating a successful solution requires your organization to complete the following assessment, regardless of whether the path leads you to buy or build.

1.Identify the scope of applications to be supported by the centralized reference data. Include applications to be acquired, licensed or built within the next 24 months.

2.Define the data requirements of these applications and their associated service level requirements.

3.Identify the current data sources used to feed the applications.

4.Identify content gaps.

5.Benchmark current problems and costs. Focus on total cost, including those from correcting failed trades, internal collection, reconciliation, data cleansing, etc.

6.Identify new costs associated with new data sources.

7.Identify all pending regulatory issues and the resultant requirements standards that should be accounted for in implementation.

8.Prioritize implementation with emphasis on a series of real deliverables. Focus on complete solutions based on asset classes, user domains, geographic region, business unit, or content type or application.

9.Identify available human resources for implementation and ongoing operations.

10.Recognize the solution is a process, requiring constant management for as long as your organization is buying, selling, settling or holding investment instruments.

11.Agree organizationally on the specifications of the solution and what it is worth.

Making the Decision: Have It Your Way

Once you have framed the solution, you will have several key criteria for your decision, involving: what must be done in-house, the implementation timeframe, the level of flexibility required, any unknowns, what you are willing to pay for a lasting solution, and what resources you have in place for implementation and maintenance.

With these in mind, let's consider effort and risk. The build option entails more "building" than one might first think. Building includes data management software as well as procedures for exception handling and cleansing. A build solution must address the operational infrastructure, from hosting to networks and operations staff. Data management software should address multiple data sources, validation, exception handling and distribution to end-user applications. In addition to operational staff, human resources for data exception handling, correction, source evaluation and customer care are also required.

Within the buy option are two basic models: Business Processing Outsourcing (BPO) and Application Service Provider/Software (ASP/SW). In general, BPO is the complete end-to-end, all-in solution that entails outsourcing the entire process and its requisite components. The ASP/SW model is a hybrid approach; it typically provides an outsourced or purchased solution to only part of the problem (such as software, hardware, operations and/or network issues). It does not address data domain knowledge, nor does it provide resources for data correction, research and end-user support. In the ASP/SW model, these functions must still be built in-house.

What Drives You?

What are the buy vs. build business drivers? These drivers are based on the specific solution requirements as well as the organization's business strategy and culture. A simple table can be used to assess the relative merits of the available models.

Taking all the factors into account, the Business Process Outsourcing model is compelling. Beyond its advantages with respect to true cost, it offers important attributes for flexibility, adaptability and control - while the overall organization is free to focus on its core missions.

Whether your evaluation leads your organization to buy, build, or opt for a hybrid approach, the decision to act is crucial and time-sensitive. Amidst the surging tides of change and business realities today, those firms that thoughtfully consider the problem and implement a lasting, dynamic solution will rise to the top.

Cicada provides data cleansing, management and validation services to support STP, risk analysis and other mission-critical applications. The company combines its technology for data acquisition and automated validation with the research and support services of its Data Quality Management group, to provide comprehensive solutions for ongoing data quality maintenance.

Article Written By Gregory E. Smith, President & CEO, Cicada Companies

Cicada Companies
Data Quality Management Operations
2123 N. Classen Blvd.
Oklahoma City, OK 73106
Phone: 972.385.9658

Contact: Laura Perdue
email: [email protected]

Register for Wall Street & Technology Newsletters
Exclusive: Inside the GETCO Execution Services Trading Floor
Exclusive: Inside the GETCO Execution Services Trading Floor
Advanced Trading takes you on an exclusive tour of the New York trading floor of GETCO Execution Services, the solutions arm of GETCO.