Making systems Straight Through Processing (STP)

Secure an initial transaction as an electronic message,

Transform and transport it to its initial execution/processing location,

Pass it through the processing cycle with little, if any, human intervention

I started my quest for understanding of T+1 settlements and Straight Through Processing (STP) while I was working at the Strong Financial Corporation in 2000. I was very intrigued by the problem of T+1 settlements and STP solutions to get there. I approached my then group manager that I would like to take the STP/ T+1 initiative at the firm. Since I had background in Finance besides in Mathematics and Computer Science, my manager gave me full support for it with full confidence in my ability to deliver.

My first challenge was to understand the problem of ‘Trade Processing’ in an investment management house. As a part of data analysis, I developed a flow chart on trade life cycle and processing steps. I had involved not only my senior colleges but also our clients in business departments, and head of operations in this task. Within a year, I gave my presentation of possible STP architectures to the Director of trading, and the Director of securities processing operations.

I convinced my manager and Director of securities processing operation that we needed a STP/ T+1 initiative group. I was able to assemble a group of people from the securities processing business unit, the technology group, and the trading desk. This was especially exciting moment as I was successful in educating all the involved individuals, and able to form a group. The mandate given to the group was to analyze the current workflow, and propose a new workflow to meet STP and T+1 requirements.

I made a case to the enterprise Architect and my group manager that a common data dictionary is necessary for the STP so that all the systems can understand the data without having to have a tailored data format for each system. Both of them agreed that it is necessary; hence, the Architect and I work on a data dictionary in XML schema in the year 2000. We took the opportunity to develop a common language for all the systems as a part of developing a messaging system to integrate enterprise systems.

The work of creating a common data dictionary for the enterprise system was a monumental task. However, our understanding of securities and trade data helped us to achieve our goal. We received help from ISO 15022, SWIFT messaging, FIX and other industry standards. We discovered that one standard or mix of couple of standards did not meet our enterprise data dictionary requirements. Therefore, we pull together all the best pieces of standards along with our internal standards to create a new flexible XML schema.

With the development of the XML schema and the messaging system in Weblogic, we were able to integrate the systems quickly and easily. When in production, the system was solid and stable, and reduced daily exceptions by almost 60%. This was our moment of triumph over integrating systems enterprise wide and was a huge step toward achieving the STP and T+1 goal.

Leave a Reply