Hu Yoshida

A Break Through in Turning Regulations into Machine Executable Code!

Blog Post created by Hu Yoshida Employee on Mar 26, 2018

Nowhere are regulations more complex and difficult to comply with, than in the Financial Services sector where levels of regulations and the focus on data and reporting constantly increases. New business models, driven by FinTechs, technology driven financial companies, are disrupting the regulatory practice with innovative new ways of doing things. This disruption to regulatory practices will require the acquisition of more data, involve the use of real-time information, and incorporation of new algorithms and analytics and the re-evaluation of older regulations in terms of implementations of technology. Financial institutions are turning to RegTech companies that use new technologies to facilitate the delivery of regulatory solutions that are configurable, easy to integrate, reliable, secure and cost effective.

 

However, even with the help of RegTechs, the current regulatory system is full of inefficiencies and ambiguities, requiring significant interpretation on the part of legal and compliance personnel who are responsible for ensuring that their organization is compliant. The figure below illustrates where the pain points occur.

MDMERR .png

While RegTechs are helpful on the backend helping organizations interpret and implement regulatory solutions, it would be much more efficient if the regulations could be implemented in a model -driven, machine executable form. Although not suitable for all types of regulation (particularly those that govern conduct or relate to higher level principles, rather than detailed rules), machine executable regulation could be applied to a meaningful amount of regulation and could therefore facilitate a significant degree of ‘straight through processing’ for regulated organizations.

 

Creating machine executable regulation starts with the regulators. The UK’s financial regulators, the Bank of England (BoE) and the Financial Conduct Authority (FCA), have proposed a Model-Driven, Machine Executable Regulatory Reporting (MDMERR) approach. This could allow for complete disambiguation as it will push regulators to work out the precise meaning of rules in advance rather than in retrospect during the enforcement of the organizations proposed compliance. Benefits of the MDMERR approach could include:

 

Clarity: Machine-executable regulation will not be as ambiguous as natural language regulations and could be executed with little or no human supervision. The long term benefits of reduced regulatory risk may enable increased investment in financial innovation.

 

Time Efficiencies: An MDMERR could greatly reduce the time to create, update and implement regulatory policy by allowing regulators (and regulated firms) to test proposals in near-real time. MDMERR could also allow for more cost efficient real-time monitoring of compliance by the regulator and the firms themselves, so perhaps reducing the many meetings, memoranda and other delays in the traditional enforcement process.

 

Cost Efficiencies: From the private sector perspective it would decrease the need for the time-consuming effort to remove uncertainty and doubt in the interpretation of regulations.

 

Change Management: The MDMERR model would enable regulators to more efficiently distribute changes and reduce the long lead times for the financial institutions to adapt to some of these changes. Since the changes would be released in machine readable formats, the financial institutions would update their systems (and controls) without the need for additional interpretations by the attorneys.

 

This all sounds very good but is this practical given the complexity of the regulatory and compliance process? In order to test the feasibility of this approach the FCA and BoE enlisted experts within the public and in the private sector to create a proof of concept model and test it against real data. In November of 2017,  Hitachi, led by Nirvana Fardhi, our Hitachi Vantara Global Head of Financial Services RegTech, joined with FCA and BoE to sponsor a two week TechSprint to develop a “proof of concept”  which could make regulatory requirements machine readable and executable.

TechSprint.png

The TechSprint successfully proved that a regulatory requirement contained in the FCA Handbook could be turned into language that machines could understand and use to execute a regulatory requirement. For a brief 2:16 min video of the highlights from this TechSprint please click here.

 

While this “proof of concept” was ground breaking, there is still a lot of work to be done to transition to a market-ready solution. For example, how to: address the risks of the accuracy of machine-executable interpretations, any errors in the code base, lack of flexibility, increasing complexity in the code base as the technical infrastructure increases, versioning challenges, opportunities for abuse, and security.

 

Follow-on work to the MDMERR TechSprint presents an interesting opportunity to test a data-based approach to representing law as MDMERR. Such a data driven approach, including the use of machine learning to translate and ensure consistency across rules in the code base, could help in the translation of text and identify potential errors along with many other uses. The FCA published a Call for Input in February to ask for views on how these ideas could be developed further and will publish its findings and next steps in the summer of 2018. In addition, the FCA is working with several banks to take this forward into a production environment within the next six to nine months

 

A major step forward was taken this month when the UK Chancellor of the Exchequer, the Rt. Hon. Phillip Hammond MP, published the Fintech Sector Strategy and publicly endorsed MDMERR giving this project official UK Government support. Nirvana and her team continue to engage actively with the FCA and Bank of England and Hitachi Vantara looks forward to taking an active part in this initiative.

 

"The announcement by Chancellor Phillip Hammond heralds the next stage of machine executable regulatory reporting. Hitachi Vantara is not only an early innovator in this process but is committed to helping drive forward such revolutionary change in financial services and beyond. “

 

            Nirvana Farhadi

Hitachi Vantara

  Global Head - Financial Services RegTech

Outcomes