"Without the graphical modeling approach, we would have to hire different personnel in terms of abilities."

Daniel Hartmann,
Head of Market Operations & Functional Design

Rule-Based Master Data Management at Eurex Repo

In 2015, Eurex Repo, one of the leading electronic vendors for secured financing transactions, introduced a new trading platform which relies on a central, rule-based master data management system. With the support of STABILIT, a specialist in the area of rule-based data maintenance, Eurex Repo largely automated its master data management, improved data quality significantly and achieved maximum flexibility in the dynamic financial market.

The interview was conducted by ACTICO with Daniel Hartmann, Head of Market Operations & Functional Design at Eurex Repo, and Joël Traber, Partner & Project Manager at STABILIT.

ACTICO: Mr. Hartmann, can you give us a brief insight into your business?

Hartmann: Eurex Repo is a leading provider of secured transactions in the area of funding and financing. Our offer is aimed not only at the repo market, but also increasingly at the market for securities lending. In addition to our "classic" customers on the sell side (banks, intermediaries, financial agencies and supranational organizations), we also offer our trading markets on the buy side (corporations, pension funds, asset managers and insurance companies). At Eurex Repo, we offer these markets an electronic marketplace, covering the entire value chain as an integral part. This means that our solution portfolio extends from trading and clearing to settlement.

ACTICO: As Head of Market Operations & Functional Design at Eurex Repo, you know what moves the market. What challenges do you face?

Hartmann: The financial sector is changing and we have to face these challenges. On the one hand, there are the economic challenges. The volumes in the repo market have declined, which is mainly due to the borrowing program of the European Central Bank. Similarly, the negative interest rates of the central banks are not conducive to our business. With the expansion of our business area to securities lending and the buy side, as mentioned in the previous question, we are responding to the challenges of ensuring the continuous growth of our market place. On the other hand, we are currently working intensively on the examination of regulatory requirements (including MiFID II) and their impact on our business, as well as the practical implementation. Our new trading platform has also helped us meet business challenges because we have become much more costeffective internally and have also allowed our customers to benefit from cost savings since, with the web-based solution, installation and updates on the part of the customer are no longer required.

ACTICO: You talk about the new "F7" trading platform of Eurex Repo. In your own words, can you describe the trading platform, what happens on it and who ultimately uses it?

Hartmann: F7 is our web-based trading platform. It is used by about 150 international market participants and currently has about 1,500 users. It supports the trading of secured funding and financing transactions in various markets and segments, with our most important market being the so-called GC Pooling® market. These are money-driven, collateralized repo transactions in which a market participant lends money for a certain period and provides the counterparty with appropriate collateral for the duration. These market participants (in other words, our customers) place offers on the trading platform, on both the buy and the sell side. If an offer is accepted by a
market participant – in this case, we are talking about a "hit" – the actual trading takes place. The transaction is sent to the central counterparty, the clearing house, and settlement takes place on the settlement day. Eurex Repo provides the trading infrastructure and, thus, is the business intermediary.

ACTICO: Can you give us some facts and figures regarding the trading volume of your platform?

Hartmann: At the moment, we manage four markets with several subsegments on the trading platform, 94 Baskets and about 35,0000 securities. The volume outstanding is at a peak of between EUR 140 billion and EUR 150 billion, with a record high of EUR 250 billion in June 2014.

ACTICO: The F7 trading platform basically consists of two parts: the upstream, rule-based master data management system and the actual trading platform, the trading system. What role does master data management have?

Hartmann: Master data management is very important: If this system does not provide data, the trading system basically has no data basis. Master data management is therefore a central system for us: not only are securities managed here, but also the users, legal entities as well as all contract types. All basic master data that we need for trading come from the master data management system. They are sent to the trading system every trading day, so that the latest data is available to our customers the next day.

Traber: To expand on that: The metaphor of a "cockpit" seems appropriate for the master data management system. To a large extent, it has a control function and serves the preparation of all static data, such as securities, market participants or markets. However, the configuration of a
market model is also possible to a certain extent.

Hartmann: The aim of the project was to develop a flexible system that allows us to relatively quickly set up a new segment, compile new baskets, introduce new currencies or create new customer groups – all this should be possible without complex software releases. We have achieved this, as well.

ACTICO: What does the basic preparation process in master data management generally look like?

Traber: Basically, the data is subject to a continuous preparation process. It runs in a rule-based manner and is triggered as soon as we process an external data source (ECB3 and SNB4) and receive it in the system. In the preparation process, the data passes through a maturity process over various stages: initial, valid, and released. There can also be more stages, though. As a result, new data are created, existing data are enriched, or entire data sets are deleted. The goal is to get as much data as possible into the released state and then deploy it to the trading system.

Process-overview: rule-based master data management

 You mentioned three statuses: initial, valid, and released. Valid means that a content-related validation has already been performed, such as the validation of mandatory attributes or the verification of the data type. Is this also done with business rules?

Traber: Yes, the entire staging, i.e., the preparation process, is rulebased. Consequently, it is not just a question of examining individual data sets, but rather all relationships in the entire data model. For example, a financial instrument that users do not have access to for legal reasons does not have to be passed on to the trading system or transferred to the released status.

ACTICO: STABILIT is a specialist in the area of master data management. You implemented the project at Eurex Repo and offer a software solution for rule-based master data management with your own Data Care Manager. What are typically the requirements for such a system and why do you rely on a rule-based approach?

"The system has to deal with poor data quality, but only the highest-quality data can ultimately be provided."

- Joël Traber, Partner & Project Manager, STABILIT -