Skip to main content

For updates and guidance related to COVID-19 / Coronavirus, click here.

Guest Blog: Digitizing Regulation

digitizing regulation

Editor’s Note: The views expressed in this blog are those of the author and do not necessarily represent the official policy or position of FINRA.

Imagine a new age of regulation in which compliance efforts were highly effective, and were also inexpensive. This seeming alchemy is coming to financial services in the form of new-generation “regtech.”

While regulation has always leveraged technology, these cutting edge solutions break from the past by leaping straight into the digital age. The same digitization that is transforming publishing and retailing and medicine and driving -- and finance -- will transform financial regulation too. Instead of starting with an existing process (which, in finance, was probably originally designed on paper) and layering automation on top of it, these new tools are “born digital.” They start fresh, gathering today’s abundant, easily-accessible data and leveraging it through today’s ubiquitous, cheap computing power and new analytic tools like artificial intelligence and machine learning. The result is as different from old systems as an iPhone is from a landline telephone...or from a camera, a paper calendar, a map, a newspaper, an encyclopedia or any of the other digitally-based tools we now use in our phones.

Governments and entrepreneurs worldwide are applying regtech to a growing list of regulatory and compliance pain points. Some are using “big data” and artificial intelligence for market surveillance to find signs of misconduct and manipulation. Some are developing “machine-readable” regulation, in which rules carry electronic tags enabling a computer to interpret coverage and requirements. Others are using API’s -- application program interface technology -- that enables real-time, full-information risk monitoring to replace sample-based analysis or periodic reporting. Work is even underway on “machine-executable” regulation, in which some guidance would be issued in the form of computer code, rather than words, to become, in effect, self-implementing. The UK Financial Conduct Authority successfully experimented with this last idea, enabling a regulatory report to be made accurately, machine to machine, in 10 seconds.

One advanced use case is anti-money laundering, or AML. The UN estimates that global financial crime is approaching $2 trillion annually and that the current system catches less than one percent of it. That meager result costs the financial industry tens of billions of dollars a year in compliance headcount and activity, and still produces high levels of failure and enforcement penalties. People are realizing that current methods can never be scaled up enough to make a major impact, and that newer technology could -- and could contain costs at the same time. Government and regtech firms are developing “digital identity” authentication to solve for the costs and failures of current Know-Your-Customer requirements in AML. They are also exploring new data encryption techniques to enable financial companies and law enforcement to share vastly more data, more easily, without compromising individuals’ privacy. Once data can be shared, machine learning can spot patterns -- typologies that signal human trafficking, or trade in illegal drugs and weapons, or terrorism -- that are now invisible to human investigators. 

Financial regulation is complex. Fully modernizing it will take time. Still, better tools are on their way.

Jo Ann Barefoot is the CEO of the Barefoot Innovation Group and member of FINRA’s FinTech Advisory Committee. She recently spoke at the FINRA RegTech Conference in New York City.