Connecting Relationships for Compliance and Competitive Advantage

Semantic technology holds a future filled with many benefits, but most organizations either struggle with how to get started or are reluctant to migrate away from legacy approaches to managing enterprise data assets. As the industry faces new regulatory requirements that reluctance may be overcome. Pronouncement 239 from the Basel Committee on Banking Supervision (BCBS) is a broad-ranging and challenging mandate for which semantic technologies may provide a silver bullet. BCBS 239 will be a key case study for how to adopt semantic technologies and capture related benefits.


In industries made up of large, complex companies, systematic or transformative changes do not come about without the application of extreme pressure. These pressures can result from the following variety of events or trends: The financial services industry is faced with a myriad of thou shalts lately. The most recent financial crisis shed added light on the challenges faced by many banks with aggregating and reporting risk data. In the U.S. alone from 2008 through 2011, financial institutions spent trillions of dollars trying to comply with numerous government programs. The losses incurred during this era served as a wakeup call for financial institutions and sparked calls for enhanced risk management and their underlying data management capabilities.

The Basel Committee on Banking Supervision (BCBS) is the primary global standard-setter for the prudential regulation of banks and provides a forum for cooperation on banking supervisory matters. The BCBS does not possess any formal supranational authority as BCBS pronouncements do not have legal force. Rather, the BCBS relies on its members’ to implement and apply BCBS standards in their domestic jurisdictions within a predefined timeframe established by the Committee. In 2013, BCBS issued pronouncement 239 known as “Principles for effective risk data aggregation and risk reporting,” or BCBS 239 for short. Here the term “risk data aggregation” means defining, gathering and processing risk data according to the bank’s risk reporting requirements to enable the bank to measure its performance against its risk tolerance/appetite. This includes sorting, merging or breaking down sets of data.

BCBS 239 required the largest banks, the G-SIBs, to be compliant by January 2016. By aligning with the principles of BCBS 239, banks need to develop key capabilities across the following areas.

The Best Path Forward

While many large banks claim to be compliant or well on their way to compliance, they have not addressed their underlying infrastructures to be able to economically, systematically and continuously comply with this standard. Compliance with this standard in early 2016 has been seen to be reliant upon a brute force method as most banks have not progressed very far with their longerterm compliance strategies. Semantic technology should be more fully examined as it will both enable financial institutions to manage constantly changing compliance and regulatory requirements while also capturing operational efficiencies in the years ahead.

During the financial crisis, the inability to compile risk data effectively exasperated the impact across many financial institutions. “One of the reasons it was so difficult to understand the risks of their own portfolios and for regulators to understand the aggregate risk of the industry was because so much data was fragmented, opaque and incongruent,” said David Newman. “It didn’t roll up and align with a common definition. It was a huge problem then and is still a challenge for the financial industry today.”

Semantic technology will help address this challenge. “It is the next generation in an evolutionary progression of information management technologies,” Newman said. “It identifies objects which can be thought of as objects or concepts in the real world with other objects or concepts. It connects relationships.”

He provides a scenario of how semantic technology would work in practice. “The notion of an interest rate swap can be described semantically as any contract with one leg tied to a fixed interest rate and one leg tied to a variable interest rate,” Newman said. “That’s something we call a fixed-float interest rate swap. It’s a concept a machine can understand because we can define it semantically. Another program might use a vanilla term, interest rate swap, which might mean exactly the same thing but there is no way to break it down to confirm the alignment of meaning. With semantic technology, we can provide a Rosetta stone capability such that we align on meaning and don’t care what it’s called as long as we have a consensus that we can tie the label to the same common, consistent meaning which becomes a standard.”

To capture the benefits of semantic technology, companies will need to focus on the following:
Alignment to Standards

The Enterprise Data Management (EDM) Council is developing the Financial Industry Business Ontology (FIBO), which is a business conceptual ontology standard for the financial services industry. Newman chairs the council’s semantic technology program. “With FIBO, financial institutions can integrate and align data to a common meaning and can then perform processing against it,” he said.
Integration and harmonization against a common meaning is one of the major use cases for semantic technology and for FIBO.
The development of FIBO will allow for the efficient implementation of semantic technology by financial institutions. “Semantic solutions can be deployed in days or weeks,” Newman said. “We are not talking about months and years. There is an order of magnitude improvement in the time to deliver a semantic solution versus a conventional solution assuming we already have the ontology defined.”

The health of the entire industry stands to benefit from these solutions. “Semantic technology and FIBO can really help the industry reduce the risk of systemic failures in the future,” Newman said. “Data is now more credible, aligned, effectively integrated and linked together so that aggregations of different abstractions can occur very rapidly.”
Adopt an Incremental Approach

Newman recommends a measured approach for financial institutions that are eying the technology. “Organizations need to start with small steps with non-mission critical applications and begin to grow into the mission critical applications where semantic processing makes sense,” he said. “Over time, we will begin to identify more and more use cases where semantic technology can provide value.”

Compliance with BCBS 239 is one such use case where Newman already sees potential for semantic technology paying dividends. “There is hardly a better use case then to apply semantic technology to help organizations comply with BCBS 239,” Newman said. “The real spirit of BCBS 239 can be strategically fulfilled using ontologies because of the degree of expressivity that leads us to link and harmonize information much more effectively than the use of conventional approaches. The level of credibility empirically has to be vastly higher and therefore the element of risk is significantly lower when you take a semantic approach to support BCBS 239.”

As financial institutions consider strategies that will ensure compliance with changing regulatory requirements, Wisnosky has seen a trend unfolding that involves semantic technology. “For new requirements, banks will build their new data storage with Resource Description Framework (RDF) triplestores – a World Wide Consortium (W3C) standard,” he said. “For their existing data, they can do mapping using other W3C and Object Management Group (OMG) Standards.”

Data mapping allows financial institutions to retrieve data from legacy databases without a major infrastructure overhaul, Wisnosky explains. “There are other standards that enable them to map their relational data stores to RDF triplestores in real time in some cases,” he said.

Wisnosky believes that financial institutions will continue to transition to triplestores, but the transition will be gradual. “The infrastructure of banks will change to become much more efficient using RDF triplestores, but it will change over a period of time,” he said. “There are banks and other financial institutions that are far along the learning curve path of adopting RDF triplestores internally. There are banks that are already using FIBO as the beginning of their Rosetta stone as they bring on new applications or look at how they’re going to comply with regulations.”

In addition to more reliable data, Newman sees semantic technology translating into financial gains. “Globally there is a huge bottom line loss in the way we process information today,” he said. “Semantic technology is going to make information processing more accurate, credible and reliable and will reduce the overhead of writing a lot of unnecessary logic and application programs.”
Instill Data With Meaning

Writing code has long been the way for many financial institutions to obtain risk data. “The existing IT infrastructure of many banks is about 40 years old,” said Dennis Wisnosky, a senior advisor for the EDM Council. “Banks have hundreds to thousands of relational data tables. The tables have been built over the years with more and more code on top of code. This approach has been required to put meaning into the data by writing the meaning into the code that reads the data.”
By instilling data with meaning, financial institutions will be able to improve flexibility of their IT infrastructure and gains a true competitive advantage.
There are standards that have been developed to support semantic technology in an unambiguous manner. “These standards give meaning to data, which computers can execute. With semantic technology, the meaning of the data is in the data itself. A simple query produces an unambiguous answer,” Wisnosky said of FIBO.

Wisnosky notes the development of FIBO has been a collaborative effort that has included the EDM Council’s approximately 140 member banks. “We have proof-ofconcept teams,” he said. “A bank can decide to become a part of a proof-of-concept team or create its own for an area it’s interested in. Each time a proof-of-concept team is formed, the council assigns one of its world-class ontologists to work with the team.”
A reduced reliance on hard coding is one major benefit that Wisnosky sees for financial institutions that choose semantic technology.
“Software engineers and coders are very expensive,” he said. “When you put data in a RDF triplestore, the meaning of the data is in the data and not in the software that has to look for that meaning ̶ it’s there. There’s no programming involved in making changes. Data stores can either be extended or linked to other data stores that obey standards such as RDF and the Web Ontology Language (OWL).”

Semantic technology can also result in significantly lower software and data maintenance costs. “Others have found that the cost to discover and to maintain data using this technology can be less than half of the cost of the traditional way,” Wisnosky said. “The hardware that stores the data is still there, but they find that the number of data stores shrinks because data is not duplicated in multiple data stores and data from external sources can be easily linked with simple queries using the SPARQL Protocol and RDF Query Language, another W3C standard.”

Wisnosky says that there may be some resistance among financial institutions to adopt semantic technology, but that it will likely subside over time. “Just like with relational data technology, there were people who were saying semantic technology would never work, but the Department of Defense which invented this technology, and others and others have put it to wide use,” he said. “The upfront costs of learning something new are quickly put behind you because of the great efficiencies gained.”


If organizations stick with existing technology archetypes, BCBS 239 is a monumental challenge for the financial services industry. The use of semantic technologies can not only help the financial services industry deal with the specific requirements of BCBS 239, but it will also provide a truly flexible data management platform by connecting relationships between data points. This flexibility will provide organizations with the ability to easily digest new requirements whether they are from the business, regulatory or other mandates. This flexibility is a true competitive advantage but the organization must have the courage to move past the legacy technology approaches in order to reach this level of performance.
Share this page: