Order Management Solution Integration using Cadis EDM

Business Problem:

The Cadis EDM product was chosen to form a hub consolidating data from a major Asset Manager’s existing information systems and also from new strategic systems that were about to be introduced. The integration of an external market leading vendor’s Order Management System (OMS) was identified as the key first stage in this project and it was targeted as the repository for data from three of the Back Office systems, Bloomberg security enrichment services and Index constituent information

 

The aim of the data hub was to consolidate this scattered enterprise information to produce Master or “Gold Copies” of:

 

  • Securities
  • Positions
  • Issuers
  • Indices
  • Accounts

 

Cadis EDM and ReformIS solution:

A combined ReformIS and Cadis EDM team followed the Cadis implementation philosophy and best practices to deliver the data hub solution. This approach used focused design sessions to understand crucial areas of the implementation, to clarify the requirements and produce Cadis Design diagrams documented in Visio. Using this approach, the associated build work was able to begin in parallel, fleshing out key design decisions at an early stage.

 

Five streams of work were identified: Issuer, Security, Index, Position and Account, with each being assigned an owner. Each owner was responsible for refining their stream ensuring it would integrate with the others and take advantage of solutions already constructed.

 

Delivery of the solution to a strict schedule was critical to the client, as the Cadis Hub would affect the timelines of the majority of the wider data project. Additionally, the team were required to work with and mentor permanent staff and consultants on the use of Cadis. The team succeeded in delivering to schedule and ensured implementation according to the joint plan.

 

Key Features:

The Cadis hub was aligned to the OMS published formats and data standards and consisted of “Gold Copies” of the following key business information:

  • Securities – Held securities from the three Back Offices combined with the securities described in constituent data from various Index vendors. Unique securities would need to be identified, whilst cleansing and aligning data to a common standard and enriching gaps in the information using Bloomberg PerSecurity.

  • Indices – Index Constituent and Index Level data would be initially sourced from Rimes, an Index consolidator. However this data would need to be verified and aligned to the client’s common standards and security information extracted.

  • Issuers – For held securities from two of the Back Offices enriched with Bloomberg data

  • Positions – From all three Back Offices aligned to the gold copy securities and accounts

  • Accounts – Aligned to positions, also providing Account Market Value.

 

The initial goal of the project was to get data from the source systems/files, cleanse and align the data and ensure that the security information was extracted ready for matching with securities from the other sources based on the availability of market identifiers SEDOL, CUSIP/CCY and ISIN/CCY.

 

The matching stage would then make use of and validate these market identifiers to decide how to group them together. Each security is then assigned a unique Cadis Identifier, with securities deemed the same between differing sources getting the same id. Exceptions and matches with a low confidence are automatically routed to an inbox allowing the data management team to review the match. CADIS Identifiers are then fed back to the source data and used in all downstream exports.

 

The technical implementation itself consisted of Cadis EDM being deployed on a Windows 2003 application server. SQL Server 2005 was deployed on a separate Windows 2003 server and housed the Cadis database which contained the configuration of each component along with the Gold Copy tables. Separate environments for Development, Test and Production environments would be supported by the Cadis component promotion tools, along with traditional database scripting.