Lord Abbett

We helped to develop a monthly process, to summarize and disseminate portfolio statistics, such as sector and credit quality breakdowns, average durations, and holdings concentrations. The system handles mutual funds, separately managed accounts, institutional client accounts, and third party indices and blends. As a part of the client’s team, we created a consistent process to:

• load and normalize position and reference data from a myriad of sources,
• assign each security a unique ID,
• run analytics through Yield Book,
• classify securities according to an asset class, industry, geographic region, and more.

For all this, the system supports different values in different contexts, such as when comparing a fund to different benchmarks. It handles funds-of-funds, differing policies for split credit ratings, and offers various approaches for handling derivatives and defining cash.
Each month, the system processes more than 20,000 instruments, and generates more than 5,000 reports for use on lordabbett.com, in the client materials, and in the reports for Lord Abbett’s board of directors.

The system uses metadata to define how to extract the holdings of various portfolio data sources, convert them to a standard form, normalize their values, and match their positions to securities that may already exist in the database. It provides for human override and complete history tracking of all values.

In the course of this engagement we introduced additional agile development practices. We set up an automated testing environment on TeamCity, and configured it to run the entire process with a limited data after each code change, and full overnight run on a clone of production data. Also, we helped introduce automated testing, continuous delivery, frequent deployments, and shared code ownership. These practices are now spreading throughout the firm.