Morgan Stanley (MSCI Barra)

During the spring of 2007 I was attending San Francisco State University working toward my Master's degree in Business Administration. One of the classes I was taking was Risk Management, thaught by the head of the finance department. One day during our weekly socials after class my professor asked me if I was interested in doing an Internship at Morgan Stanley (MSCI Barra) in Berkeley, CA. Apparently one of his former students who became a department manager there was recruiting interns for the summer. He explained that he thought my software development background and understanding of financial theories qualified me more than anyone else that he could think of and that he was going to recommend me if I was interested. When I found out that it was going to be a paid internship the answer of course was yes. After a couple of informal interviews with the department manager and MSCI Barra I was called in for a formal group interview where I was interviewed by a handfull of people from various departments of the company. A couple days later I received a call with the news that I got the OK to come on board beginning June 4th, 2007.

My job at MSCI Barra can be best described as a business systems analyst in the Data Concent Services (DCS) Global Fixed Income (GFI) department. The DCS GFI department was responsible for aggregating all fixed income security information, transforming the data, and pushing it out to clients. The scope of my work included evaluating and documenting the state of the current system (database and applications), defining departmental needs, identifying the limitations of the system, and coming up with a proposed solution to remedy the shortcomings of the system. Since I was considered the expert in the field the method I was going to use was pretty much up to me.

For me the first step was to access and learn about the various data processing steps and the systems associated with them. In order to do this I documented all the asset processing procedures, and application functionality (output formats, data locations, etc.) Next I reviewed the application logs to find processing errors, identified the causes of failure, and evaluated processing logic. During this time I also tested and troubleshooted some newly developed data processing applications. The last step was to write a prototype reporting application in Perl for Linux called Decision Tree Reporter. This application extracted and reported the various asset processing failures from the logs on a daily basis, simillarly to how I did that at first manually, and published the reports to the Intranet in an HTML (web browser readable) format. In addition to the Decision Tree Reporter application prototype I also made detailed recommendations on how and where to improve logging functionality in order to make the reports more meaningful.