"Do You Know Where Your Data Is?"

David Saul


Moving from Data Confrontation to Data Collaboration in Financial Services

My title derives from one of the most successful public service announcements of all time. A written message would appear on your television screen and a sonorous voice would announce – “It is 10 PM, Do you know where your children are?” I believe that attention-getting phrase can be adapted to the increasing data management challenges that all financial services institutions are facing. It sums up the need for effective data governance built on an appropriate technology base.

I am not speaking solely of “big data” which represents only one aspect of the situation. I tend to avoid using the term “big data” since it leaves out the most important characteristic of data – its meaning. I believe that by embracing semantic data as an industry we can increase economic value and reduce risk. The result is true “smart data”.

The concept of the semantic web was first proposed over ten years ago by Sir Tim Berners-Lee, the creator of the World Wide Web, and has since been realized in multiple implementations. Semantics is a natural evolution of earlier work on metadata, language dialects and taxonomies for regulatory compliance. Examples include the SEC’s XBRL mandate and OFR’s Legal Entity Identifier (LEI) as part of the Dodd-Frank legislation.

The increasing complexity and pace of global regulations is making it more difficult and expensive for financial services organizations to comply. At the same time, firms want to derive value from their data assets. How do they create synergy between these two seemingly divergent goals? The maturation of semantic technologies, when combined with increased acceptance of industry standards, holds out the promise of resolving those issues. Semantics and ontologies provide greater transparency and interoperability, thereby enhancing the overall
trust in the financial system. Enhanced trust benefits all constituencies who have a direct interest.

Let’s look at four constituencies that have a shared interest in restoring and improving trust in the operation of the global financial services environment. By enhancing trust in the market we increase investment and raise economic standards for everyone.

  • Financial services firms gain additional revenue from their clients while keeping risks at an acceptable level.
  • Product and services companies have clearer requirements to innovate, develop and sell.
  • Regulators and supervisors receive the information they need to meet statutory mandates and ensure that laws are complied with.
  • Standards organizations follow their mission to enable simple and effective communication among the parties.

When those four constituencies treat one another as adversaries, the financial services marketplace is less efficient leading to loss of overall trust. When the four work together in pairs or as a group they all gain value. I propose that we move to a new model in which those four groups collaborate to their mutual benefit. To accomplish this goal I posit that we need better data governance built on semantic data standards. When data moves along with its meaning based on standardized definitions it enables transparency. Transparency is at the heart of trust that benefits everyone. This approach is consistent with the Basel Committee on Banking Supervision (BCBS) 239 “Principles for effective risk data aggregation and risk reporting”.

At last year’s MIT Chief Data Officer & Information Quality Symposium, a group of us, representing a standards organization, a former regulator, a data products company and myself from a financial services institution came together to propose just such a collaborative model. Our message to the audience was that they go back to their business and information technology groups to implement and/or enhance data governance in preparation.

  • They should answer questions about their data governance, beginning with “do you know where your data is?”
  • They should catalog and monitor their current and future regulatory requirements.
  • They should understand their existing products/services solutions and identify any gaps.
  • They should get involved in and influence relevant semantic data standards.

Just one example of the latter is the Financial Industry Business Ontology (FIBO) from the Enterprise Data Management (EDM) Council and the Object Management Group (OMG). Recent publications from regulators in the US and elsewhere have endorsed the use of data standards as the only way to deal with the increase in the scope and complexity of their responsibilities. For example, in its 2014 Annual Report the US Treasury Office of Financial Research (OFR) devotes its entire section 5 to “Advancing Data Standards”.

Semantics provides additional advantages over traditional technologies in its speed and flexibility. Developing Extract, Transform and Load (ETL) processes and data warehouses cannot keep pace with changes in our business models and relevant regulations. The ability to easily create and change semantic maps of data ecosystems is being offered today by a number of vendors. The open nature of data standards like FIBO not only provides transparency but also provides assurance that these standards will be long lasting. Current academic research is showing our semantics can be a path into more leading edge technologies like machine learning and natural language.

In future months our core group intends to continue to spread the message of data collaboration through conferences, white papers and other forums. In the interim, feel free to contact me directly to show your interest and support.

David Saul,
Senior Vice President & Chief Scientist,
State Street

June 2015