The start of the new millennium saw an explosion in the number of XML standards to support the needs of the financial sector. Though a few lacking the necessary momentum have fallen by the wayside, the sector now has a comprehensive set of standards able to meet the needs of the trading environment.
|Financial XML Standards||Scope|
|MDDL (Market Data Definition Language)||Market data, reference data and corporate actions|
|FIXml (Financial Information eXchange protocol markup language)||The pre-trade, trade and post-trade processes.|
|FpML (Financial products Markup Language)||Derivatives trading|
|TWIST (Treasury Workstation Integration Standards Team)||FX and money markets|
|RIXML (Research Information eXchange Markup Language)||Company research|
|SWIFTStandards XML||The settlement process|
|NewsML (News Markup Language)||Market announcements|
|XBRL (eXtensible Business Reporting Language)||Corporate business reporting|
Due to the need to develop at Internet speed each standard has kept its scope narrow and therefore overlap has been kept to a minimum. Where overlaps have been identified the bodies concerned are actively addressing them.Since the beginning of the year (2004) there has been a great deal of cooperation between the standards organisations. In January, ISDA (International Swaps and Derivatives Association) maintainer of FpML had its ISO liaison status approved, meaning that it will be able to actively participate in the ISO process. The FISD (Financial Information Services Division) of the SIIA (Software and Information Industry Association) also agreed to contribute the MDDL terms and definitions to the working group, ISO 19312 (the Securities Data model). This was followed in February by the announcement that FIX and FpML would create a joint working group, which tied up nicely with the previous collaboration agreement between FIX and ISO 15022 Second Edition (TC68/SC4/WG10).
The biggest challenge to trade automation is consistent reference data, without which incorrectly identified transactions will continue to need repair and hence expensive manual intervention. With the release of MDDL version 2 the FISD started to address the reference data challenge. In 2002 the FISD founded REDAC (the Reference Data Coalition) to provide business oversight into reference data standardisation. To ensure global coverage, REDAC has been working closely with RDUG (the UK/European-based Reference Data User Group) and in February 2004 representatives from REDAC, RDUG and ISITC (International Securities association for Institutional Trade Communication) formed a joint working group to document and identify solutions to securities processing errors caused by instrument identification issues.
In light of this convergence and to avoid confusion in the securities industry between the ISO initiatives 15022 and 15022 Second Edition it was decided to rename the 15022 Second Edition the "UNIversal Financial Industry message scheme" (UNIFI) and allocated it a new ISO classification number, 20022. This provides a clear message to the industry that both 15022 and 20022 can co-exist. Thus differentiating between the remit of ISO 15022, a standard for electronic settlement messaging and ISO 20022, which extends this to create a single standard for financial instruments throughout the trade lifecycle. The initiative is built around a repository where processes, messages and the data dictionary are modelled together using the Unified Modelling Language (UML).
One must have realistic expectations for the initiative, as it's likely to be several years before the full benefits are realised. In the meantime this should not stop an organisation deploying XML standards within the enterprise. Given that standardisation can result in the reduction of application development times, and the number and complexity of interfaces, which in turn leads to reduced maintenance and support costs, Heads of Departments should be asking why are we not using XML standards?
The XML representation of data is often several times larger than the text equivalent and up to ten times larger than its binary form. This has resulted in concern being expressed about the size of XML documents, "Given its verbosity, can XML cut it in a time critical trading environment?" When one examines the daily lifecycle of data at any financial institution, it becomes apparent that the deployment of XML is not automatically an issue. Where it may become a concern is when transferring large volumes of information, such as the daily data downloads or large time and sales files. Even so by using data compression techniques this is not an insurmountable obstacle.
The best results have been achieved by solutions based around MPEG-7. This standard was originally developed for streaming audio and video data with embedded XML information. Using MDDL documents we have achieved compression to less than 15% the size of the original data; with some documents 5% is routinely achievable.
The FISD has recently entered the fray by defining a compaction scheme and messaging protocol for efficiently conveying XML content. Known as the fisdMessage, it is an open source offering made freely available to its membership. Since its release some confusion has arisen about the relationship between fisdMessage and MDDL. Even though the documentation for both reside on the MDDL website, the two are not intrinsically linked. It is possible to use one without the other, and according to the FISD there is no reason why fisdMessage cannot be used for streaming other XML documents.
A well-designed schema can also help to reduce bandwidth usage. MDDL makes use of inheritance and deep structural hierarchies supported by XML. FIXml has opted for keeping the number of elements to a minimum, concentrating on the use of attributes. With FIXml version 4.4 the tag and attribute names have also been reduced in size, for example, "price" has become "px". This raises the dilemma faced by all schema designers, balancing the trade off between readability and the need to reduce document size.
Given the air of cooperation between the standards groups, the outlook is very positive, though it is worth being realistic about the potential outcomes. As most groups are run on a voluntary basis, timescales tend to be longer than in a commercial environment. Therefore it's likely to be some time before we see the full benefit of the UNIFI initiative. In the interim, as solutions are available to overcome XML's verbosity, companies should take a serious look at using XML standards to solve part or all of their data management needs, especially as by deploying industry standards can result in measurable financial benefits, even in the short-term.
Martin C. Sexton
London Market Systems Limited
33 Throgmorton Street
London EC2N 2BR
Tel: 020 7397 3350