XML in the City

Financial institutions and technology suppliers are working on a way to streamline the huge quantities of research reports...

Financial institutions and technology suppliers are working on a way to streamline the huge quantities of research reports hitting City desks each week. XML could have the answers

Investment managers want facts not documents. This simple statement cuts to the heart of the problems surrounding the generation and distribution of research analysis.

Easy access to data and clear recommendations is paramount, yet even with advances in electronic formats and the Web all that has changed so far is the partial replacement of paper with PDF (a file format that can be viewed on multiple platforms). Nothing fundamental has improved. Consumers of research have been deluged with ever larger volumes of raw output. True, print costs for research houses have been brought under control but this is still the same business process at work. Highly paid and respected analysts produce worthy but difficult to consume tomes for their desired audience who wish instead to easily find, compare and contrast the central analysis, forecasts and trading ideas.

All this will change with the advent of research componentisation.

Analysts can continue to author content in traditional ways assisted by a range of information provision tools, but the result will no longer be a simple document. Instead, componentisation tools disassemble the constituent consensus data, forecasts and ratings together with the commentary, charts and other content. Thereafter, these component items can be separately revised and manipulated, and seamlessly re-assembled.

The most suitable target formats for particular distribution channels, clients and regulatory environments can all be automated. Hence clients wanting a short format automatic email or those who wish to search and extract specific items from an online research aggregator will all be serviced from a single point of entry under common compliance controls.

To bring about this change and generate real value requires that industry, process and technology forces all converge.

From the investment management industry perspective, especially given the doldrums in the equity markets, buy side decision makers are demanding clear and independent analysis. In the UK this is further reinforced by the Myner's Report proposals, requiring fund managers to provide well-documented evidence for their investment choices.

For the sell side suffering a dearth of attractive new share issues to leverage business, their selection as brokerage will increasingly rely on the strength of their research and relationship offerings. To justify the considerable cost of maintaining highly rated analysts their output must be seen to win the hearts and minds of clients and their resulting trades.

Analysts' reports must be delivered in the most effective and appealing manner for their brokerage's brand to surface in the sea of content available through Web-based portals and distribution services.

The major players in all stages of the research lifecycle have been active in acknowledging the need for change, and collaborating to improve their mutual processes. Consortia of buy and sell-side institutions, together with information providers and tools vendors, have launched two significant initiatives in this area.

Research Information Exchange Markup Language (RIXML) was formed as an open standard for investment and financial research. It is a unified system which aims to makes it easier to categorise, aggregate, compare, sort and distribute research.

Investment Research Markup Language, (IRML) launched by research portal Multex, has a similar mission statement for an 'open-architecture standard which would benefit all facets of financial information technology', and has attracted many industry supporters.

Unsurprisingly both standards have XML as the underlying format to take full advantage of the universal adoption of tools to support this technology. Despite this common underpinning, and the substantial collaboration between the two initiatives, they do differ fundamentally in their approach.

RIXML aims to highlight key items from the document, without replicating all the original content of the typical publication - with IRML taking the contrasting, comprehensive approach to holding all items in XML.

RIXML is also initially focused on equities, whereas IRML will address fixed income, commodities, foreign exchange and derivatives.

Each authoring operation will evolve internal XML-based formats to best reflect its unique working practices and product coverage. This will be a superset of the information required by either industry standard, to be transformed as and when needed into the chosen target for delivery to a portal or end-client.

The componentisation process does not stop at the sell side. Investment managers also have to collate and publish reports, strategies and decision documentation, which in turn are partially sourced from brokerage research. Having the technology in-house to receive and manipulate material in component form will have a huge impact on the quality and cost of this work.

In future, it is likely that the entire research process will be integrated with coverage systems, client relationship management and self-service portals. For the first time, analysts, sales forces and management will have clear and timely feedback on their collective impact on clients and orders.


Richard Hall is chief technology strategist of financial industry consultancy, Evolution

Research the XML way
The quality of research has always been a key differentiator for investment banks and a primary tool for winning clients. A new approach to authoring called 'research componentisation' offers a breakthrough in the online presentation and distribution of research. Research componentisation uses XML (eXtensible Markup Language) to build tailored reports from research elements, but without the need to force rigid formulae and irritating templates onto analysts.

Read more on Operating systems software