Sean Park Portrait
Quote of The Day Title
The real source of all growth is human creativity and entrepreneurship, which always comes as a surprise to us, especial ... [hover]
George Gilder

Semantic, shemantic…rich, open data is what we want.

A few weeks ago, I issued a call to action with respect to creating a W3C working group focused on advancing the implementation of semantic web technologies and approaches in the financial services domain. Chris kindly responded, and in particular took issue with the usefulness/appropriateness of using (the existing) semantic web toolkit (RDF triples, OWL, etc.) due to their innate complexity. He writes:

At the moment though it seems that you have a problem if you need somebody that understands derivatives OR XML schema, and a real headache if you want somebody that understands derivatives AND XML schema.

Two thoughts. Firstly, I’m not a developer and so my enthusiasm for any particular solution or outcome with respect to software and code is necessarily (due to my lack of knowledge) more conceptual than practical: ie it is hard for me to have a robust opinion on the underlying path taken to achieve a certain result. What excites me and I think is important, is to build on the technologies of the web and ultra-cheap storage and bandwidth, to create rich, linked, open data and metadata sets in finance. Much of course already exists as Chris correctly points out, but so much more remains to be done. Secondly, it might be ‘a real headache’ but what 21st century finance needs is exactly people that understand derivatives and XML schemas.* I’m sorry but it isn’t that hard to develop these kind of people, and in a nutshell encapsulated the vision we had for Digital Markets at DrKW several years ago. I suspect that many Digital Generation finance professionals already (or could easily) fit this criterea. (The barriers however are cultural. When we built Digital Markets, while I knew there would be many challenges, I completely underestimated just how threatening such a vision was to the status quo: in particular, the very idea of calling into question the distinction between front and back office staff, even if just a subtle blurring of the line for a few dozen employees, caused the corporate anti-bodies to go on full alert. Removing the distinction between star-belly and plain-belly Sneetches was not something the organization was ready to condone.)

Indeed one of the most important and valuable objectives of setting up such a working group (whether or not it is under the auspices of W3C, although I lean toward not reinventing the wheel and building on the existing infrastructure of such a collaborative industry forum and think there is a better cultural fit with the objectives of such a project at W3C than say at any financial sector industry association…) is to create a focal point – not a gatekeeper (!) – for the community to innovate around a common theme and purpose. Another is to cultivate a shared respect for and understanding of the value of open standards, something that is taken for granted in many other industries but is still anathema on Wall Street and in the City. Of course there are glimmers of light to be seen in things like the FIX Protocol, but even here the underlying cultural mindset was more Microsoft than Unix… Way back in 2003-ish (?), when I was running syndicate at DrKW, we published (on the web) an XML schema describing a new bond issue, with the goal being to help others create e-bookbuilding platforms that would be able to communicate with ours. (I tried to find the link but was unable, any current DKIB folks know if it is still live?) Pretty tame stuff right? Well suffice to say the reaction of our/my peers was various combinations of:

  1. what the hell is XML and what are you guys on about?
  2. you guys have the best e-bookbuilding platform, why on earth would you give away your data structure???
  3. is this some kind of trojan horse? what are you trying to pull?

I no longer work day to day in a big institutional banking environment, so it’s hard for me to judge how much, if at all, these attitudes have evolved over the past couple years. I may be naive but I don’t see why we assume finance and derivative professionals can understand (and apply) concepts like convexity, but balk at expecting them to understand ontology and its implications. I thought these folks were supposed to be clever. In my view it’s about leadership. If the folks in the corner office think ontology is important, so will the rank and file.

So maybe semantic web tools aren’t the only – or even the most important – path to enabling my vision; I’d still think it would be useful to catalyze a more formal community of interest around creating a truly rich set of linked data in financial services and markets. And I hope Chris, and others like him, would be keen to get involved.

* If a few more of these kind of people had populated the top of securitization groups of the last several years, we may have avoided some of the worst excesses; securitization is nothing but managing vast and complex sets of (inter-related) data. Data quality is more important than credit quality: garbage in, garbage out…

Related articles by Zemanta
Reblog this post [with Zemanta]

Post to Twitter Tweet This Post

Permalink||Comments|Filed under 'Data, Digital Generation, Tools'
Citi to re-arrange deck chairs, market applauds? FX 2.0
  • Thanks for the comments guys. Not enough hours in the day! I need to ease up working on ideas (we have plenty of those) and start working harder on raising capital (so we could put more of these ideas to work!)

    And Alexander, re: Michael O'Leary although he seems a bit scary (via the press) his disdain for 'incumbent' thinking is obviously well known and one could say the foundation of his success. Would certainly be an interesting meeting I imagine!
  • Alexander,

    I'm not sure that XBRL helps with the full scope of the problem that Sean's trying to address, though it certainly improves things at the edge (particularly in facilitating better automated analysis of reports). I know some equity analysts who are very excited about what they’ll be able to do with this, though the cynic in me still wonders what happens to the footnotes (where all the real action is these days)?

    Hopefully one of the things that XBRL will do is drive broader familiarity with XML formats and the tradeoffs needed in getting data in and out of this shape. More at http://thestateofme.wordpress.com/2009/03/18/im...
  • Wow! XBRL is such a controversial thing it seems re. RDF. Please note the following:
    1. http://groups.google.com/group/xbrl-ontology-sp... - XBRL Ontology
    2. http://linkeddata.uriburner.com/about/html/http...


    I am actually looking for collaborators to do some work with me re. finessing the current alpha work we've done.

    Kingsley
  • Chris has some good points.

    Though, I think you guys should take a good look at XBRL which is based on XML. I have been looking into XBRL for another use case unrelated to finance.

    XBRL is important as it has gained traction as a Financial Reporting taxonomy behind the regulatory drive to standardise company reporting protocols in the US.

    For example, the Federal Financial Institutions Examination Council (FFIEC) who are responsible for monitoring some 8,200 banks in the US and three control agencies, the Federal Deposit Insurance Corporation (FDIC), Federal Reserve System (FRS), and Office of the Comptroller of the Currency (OCC) have been using quarterly reports (Call Reports) submitted in XBRL since Oct 2005 to gather information from banks and track their financial health. Fat lot of good that has done, right?!
    You can download the FFIEC case study in PDF here: http://www.xbrl.org/CaseStudies/FFEIC_XBRL_06.pdf

    Right, what is XBRL? Here's a quick copy & paste:

    "XBRL stands for eXtensible Business Reporting Language. It is one of a family of "XML" languages which is becoming a standard means of communicating information between businesses and on the internet."

    "XBRL is being developed by an international non-profit consortium of approximately 450 major companies, organisations and government agencies. It is an open standard, free of licence fees. It is already being put to practical use in a number of countries and implementations of XBRL are growing rapidly around the world."

    See links:
    What is: http://www.xbrl.org/WhatIsXBRL/
    How it works: http://www.xbrl.org/HowXBRLWorks/
    Benefits&Uses: http://www.xbrl.org/BenefitsAndUses/ - http://www.xbrl.org/XBRLandBusiness/

    For your readers who can't be bothered clicking through to the XBRL site, here is the rest of the explanation:

    A Simple Explanation

    "The idea behind XBRL, eXtensible Business Reporting Language, is simple. Instead of treating financial information as a block of text - as in a standard internet page or a printed document - it provides an identifying tag for each individual item of data. This is computer readable. For example, company net profit has its own unique tag."

    "The introduction of XBRL tags enables automated processing of business information by computer software, cutting out laborious and costly processes of manual re-entry and comparison. Computers can treat XBRL data "intelligently": they can recognise the information in a XBRL document, select it, analyse it, store it, exchange it with other computers and present it automatically in a variety of ways for users. XBRL greatly increases the speed of handling of financial data, reduces the chance of error and permits automatic checking of information."

    "Companies can use XBRL to save costs and streamline their processes for collecting and reporting financial information. Consumers of financial data, including investors, analysts, financial institutions and regulators, can receive, find, compare and analyse data much more rapidly and efficiently if it is in XBRL format."

    "XBRL can handle data in different languages and accounting standards. It can flexibly be adapted to meet different requirements and uses. Data can be transformed into XBRL by suitable mapping tools or it can be generated in XBRL by appropriate software."

    "XBRL International is a not-for-profit consortium of approximately 550 companies and agencies worldwide working together to build the XBRL language and promote and support its adoption"


    Like yourself, I am not a developer but consider this fact an asset and a distinct advantage. Not being a developer has the effect of "freeing thought" allowing one to approach an idea unencumbered by legacy luggage. Developers tend to favor the method(s) & language(s) that are most familiar to them - narrowing the possibilities to innovate or hybridize.

    Having closely read your recent posts on the new paradigm you would like to see emerge in financial services, XBRL may just provide you with an initial, extensible, framework. Its Open Source to boot. In my layman's view, you could build a bespoke XBRL taxonomy to fit your use case and then slap on the semantic and contextual bits to enhance the value of the aggregate data. Stick the lot on a Hardoop cloud like the new one from www.cloudera.com (to help sort out the pain of having to deal with data sharding etc. as your data set grows big) and most importantly, don't forget to send my 10% cut of the profits to charity (list to follow!)!

    BTW, the one person I would like to see on the Board of Advisors of your New Bank is Michael O'Leary (Ryanair). Might be able to help you with this - he is a friend of an old friend of mine. I think O'Leary could make a valuable, if radical, contribution to the diversity of thought you are looking to harness.
blog comments powered by Disqus