FISD Launches XML for Market Data Initiative

January 23, 2001

Executive Summary of Meeting

FISD members launched a formal activity designed to produce an open XML standard for market data. Tentatively labeled MDDL (market data definition language), the primary focus is the development of a market data DTD/schema on the fields needed to describe a security and its price. And while good progress was made at the initial meeting, the full parameters of MDDL have not yet been completely defined/agreed. Refinement and agreement on the scope of the domain to be defined by MDDL will be one of the first objectives of the working group.

Rationale: At the outset of the meeting, FISD members quickly established that while there are a lot of groups working on a number of niche areas within the financial arena, no one is dealing with the product/price aspects of market data. And since a lot of what is passed around within firms is product/price data, a common interchange format among multiple sources of market data is both important and useful. And, of course, interoperability is the goal so coordination with the other XML groups in the financial domain was characterized as essential.

Strategy: The initial target is a workable, practical and realistic MDDL standard (of the most useful fields) in 6 months that can be extended as appropriate. Three working groups were created: (1) Steering Committee (responsible for overall strategy); (2) Technical Committee (responsible for structural design and the DTD architecture); and (3) Vocabulary Committee (responsible for definition of the data elements to be included, semantics and terminology). Participation on the Technical and Vocabulary Committees is voluntary and open to all interested parties. FISD membership is not a requirement. Jeremy Sanders (Merrill) is the Chair of the Steering Committee. Steering Committee members have not been finalized. Michael Atkin (FISD) is the staff coordinator (e) matkin@siia.net (p) 202-452-1600 x 361.

Next Meeting: The first meeting of the Steering Committee is scheduled for February 5 at Merrill Lynch in NYC. Additional information to be provided shortly.

Why MDDL?

The rationale for MDDL is threefold -- (1) facilitation of straight through processing, (2) ease of mapping between market data applications, and (3) enhanced market data product functionality.

Straight Through Processing: T+1 and global straight through processing (GSTP) are efficiency problems. The global financial pipeline and network is under tremendous strain and is undergoing whole scale infrastructure and format design (hence the number of settlement and re-engineering initiatives dealing with post execution and settlement). The message syntax for GSTP is XML.

Applications: User firms spend a lot of time and money translating market data formats and modifying applications for internal communication. The goal is to integrate data from multiple sources into disparate systems throughout the global enterprise without having to understand how information providers format and package data internally. The payoff is to shift the focus of internal resources from issues associated with the "formatting of data" to those associated with the "quality of data."

Product Functionality: Market data vendors have spent decades building competitive advantage based (in part) on data delivery. Why would they voluntarily give that up and participate in MDDL? The honest answer is twofold. (1) In the age of efficiency, customers will increasingly demand standard data formats and interoperability. (2) XML can help vendors add value by making the data work for the user rather than requiring the user to work on the data to derive useful information. For example, the "meta data" nature of XML will allow a vendor to personalize information by embedding descriptive data into their products to allow for seamless browsing and intelligent monitoring/filtering across related data -- such as prices, news, company information, and economic data. Ultimately, the long-term benefits of interoperability and product enhancement appear to outweigh the short-term competitive benefits of proprietary data formats.

Parameters of MDDL

The definition of "market data" is potentially huge and there are currently a number of independent groups working on niche areas in the financial domain (i.e. XBRL, FIXML, IRML, and newsML). As stated, there is no current initiative dealing with the product/price dimension of market data -- which accounts for a significant portion of data passed around within firms. This is the domain of MDDL.

FISD members repeatedly emphasized their interest in the creation of a practical, achievable, and realistic market data interchange meta language. The up-front approach is to make MDDL evolutionary, not revolutionary -- and to focus on data fields where there is a high degree of congruence. MDDL will likely be flat, simple to map and relatively non-controversial (i.e. minimal granularity). The initial focus is on end-of-day and intra-day snapshot applications that can be extended into real-time and into more interpretative data models if appropriate.

There are likely to be three stages of "vocabulary" work -- (1) identification of the broad domain (i.e. equities, currencies), (2) identification of the types within that domain (i.e. common of preferred for stocks, commercial or floating rate for currencies), and (3) identification of the specific data points with the domains and types (i.e. open/high/low for equities, unit per U.S. $ for currencies) and the detail to be included at the data dictionary level.

Other Issues

Intellectual Property: The intent is to make MDDL an open standard. SIIA/FISD will own the standard solely for data quality purposes (i.e. to prevent modification by unauthorized agents). Members issued a cautionary note -- indicating that MDDL should not incorporate proprietary standards or standards that do not meet open licensing requirements into the MDDL DTD.

Decision-Making: MDDL will be managed on a consensus basis with final decision making authority resting on the Steering Committee.

Exchange Participation: FISD members are encouraging exchanges and market centers to be involved in MDDL. Exchange interest is likely to be threefold: (1) some of the vocabulary of MDDL will clearly relate to exchange products; (2) there is a permissioning component that needs to be incorporated; and (3) exchanges are becoming vendors in their own right. Vendor and user participants just want to ensure the exchange perspective is incorporated into the development at the earliest stages.

Meeting Participants

Michael Atkin (FISD), Sudhakar Bandu (Lehman), Mike Benveniste (Fidelity), Dan Berndt (Morgan stanley), John Bottega (Merrill Lynch), David Breau (Jordan & Jordan), Jing Chi (Associated Press), Ike Copperman (The Morris Group), Karin Deridder (SWIFT), Gary Dreiscoll (Lipper Analytics), Avner Gelb (SIAC), Mayauk Gupta (Morgan Stanley), James Hartley (Bridge Information Systems), Wai-Yee Helfrich (Dow Jones), Daniel Hong (Bloomberg), Mark Hunt (Reuters), Sumedh Kapoor (S&P Comstock), Doug Kemp (Bloomberg), Sheila Morrissey (Dow Jones), Stuart Myles (Dow Jones), Daniel Perez (Dow Jones), Hung Phas (Associated Press), Denise Pollock (Fidelity), Kevin Roche (Dow Jones), Linda Rolufs (Merrill Lynch Research), Frances Sames (Bloomberg), Jeremy Sanders (Merrill Lynch), Kim Sever (Jordan & Jordan), Nat Sey (Financial Times), Craig Shumate (The Morris Group), Pamela Trendell (Sungard Trading & Risk Systems), Brian Yeah (Jordan & Jordan), Richard Zellnier (Jordan & Jordan), Tony Zhang (FinPortfolio, Inc)