Congratulations! You finally got the data analyst job for the debt capital markets (DCM) group. You impressed all of your interviewers with your knowledge of the fixed income trading venues and your online degree in data science. Your first project is to create a database that aggregates all of the historical fixed income trades and inquiries from the banks back-office system, Bloomberg TOMS and every trading venue that is used by the DCM group. It is a tall order but you got this.
After a few weeks you finally acquire the historical trade/inquiry data sets from your operations group, MarketAxess, Bloomberg, Tradeweb and MTS. The excitement is high because you will finally be able to build the data visualization tool that everyone on the desk has been demanding. You are also excited to apply your predictive analysis tools to look at patterns of customer behavior. This is it! AI is on the way!
You can immediately see the potential of the data with thousands of inquiries with CUSIPs with prices, yields, spreads and sides. The ability to predict intent is in your grasp! But wait!
The account information isn’t matching between inquiries and trades. You have multiple identifiers for each account for your inquiries. You try and drill down to one account to understand the granularity of the problem. Account 88888 is the internal account for Leeson Fund Management in your back-office system and on Bloomberg TOMS. All of your trades have the 88888 account number. However, none of the inquiries that were missed or passed from LEESON FUND have the 88888 account number. Bloomberg ALLQ inquiries have a customer number of 1940393, MarketAxess inquiries have the account identifier as LEESON FUND, Tradeweb has BARINGS SING and MTS has 22088888.
A lack of harmonized data kills any progress with your grand plans of AI or data visualization.
You need find a KEY that will establish a relationship among the tables that you have in your relational management database. How can you tie these account identifiers for inquiries to an identifier that is a global standard?
What are the different identifiers that can be used? Looking at FIX tag 447, you have several choices for what is available in FIX:
You have heard of the LEI but your not familiar with its application. So you go into a Youtube wormhole for the Legal Entity Identifier.
The project just grew by 10 times because now you have to start a project with your Know Your Client team, risk management team, account onboarding team and operations team.
You put together a statement that highlights the following for your company: We are no longer a bank, we are a data company. Adoption of the Legal Entity Identifier will help us produce insight that will be proprietary and profitable to our firm.