Could a New Type of Fintech Alleviate Regulation Compliance?
September 2018 marks a decade since the global financial crisis.
Left in its wake: Hawk-like regulatory oversight that forced banks once considered “too big to fail” to spend roughly $321 billion (as of 2016) in fines, and compliance costs reaching an estimated $100 billion per year across the industry.
“There are so many regulations, so much mayhem and so many fines,” says Mary Kopczynski, CEO of regulatory technology (regtech) company 8of9. “[Financial institutions] have not yet reached the point where they can be strategic about regulation and thoughtful about how to address it holistically.”
One way to help the financial industry get more strategic? Technology that cuts down on the issues regulations create—namely, expensive fines and a lot of man-hours.
Regtech is a new class of financial technology (fintech) that promises to ease the regulatory burden. And though it’s still in the early stages, it has the ability to cut compliance costs by making massive laws machine-readable and giving financial institutions the tools to comply with the laws accurately—and on time.
Even the Most Powerful Regtech Requires Human Touch
Every year, large banks funnel 15 to 20 percent of their operating budget toward compliance oversight. To appreciate why compliance is so costly, it helps to step back and consider the data. As 8of9 COO Aaron Heisler points out, a regulatory text is basically a huge block of unstructured data.
Some regtech tools are able to “read” these texts to cut down on time and potential oversight. They process them, line by line, extracting metadata and breaking them down into clear rules. The metadata indicates which rules apply to which financial products, which helps compliance teams inform the right departments of changes to reporting requirements.
Take, for example, the approximately 7,000 page MiFID II, a regulation document that protects investors across the European Union. A recent regtech pilot run by a major Australian bank deployed a tool that used natural language processing to parse MiFID II. Without the tool, it took some compliance officers six months to read the text, consult lawyers and compile a list of rules. With it, the project took two weeks.
The results suggest that the bank could save over 85 percent of its time by using the tool—but that’s assuming full accuracy. A legal review found the AI-digested regulation was only 95 percent accurate. Regulatory complexity, resulting from both the quantity of text and the ambiguity of some of its language, means that humans will play an important role in the adoption of regtech products. As Matthias Memminger, a regulatory expert in Bain & Co.’s financial services division, puts it, compliance experts are “the heart of regtech.”
A Solve for Identity Theft?
Other regtech tools help scrape data so financial institutions can verify customer identity (an important piece of anti-money laundering regulations). Know-your-customer (KYC) regtech tools can comb the internet to verify the identity of customers, often mining up to the ten thousandth result of a search query. According to Memminger, a global retail bank keeps track of around 50 data points per customer. So, for a bank with 10 million customers, that means the KYC tool would compare online search results against an internal database of half a billion customer data points.
KYC tools are valuable for institutions, says Memminger, because the regulations that relate to KYC inflict some of the biggest costs and carry substantial risk. Still, they require some human participation. Put simply, without human judgment, the current regtech tools wouldn’t be effective; they’d be imposing misinterpretations of the law or rejecting customers just because they happen to share a name with a suspected money launderer.
The Long Road to Industry Adoption
As with any new technology, industry-wide adoption will take time. But regtech tools, in tandem with the compliance officers who use them, are showing promise in their ability to parse vast unstructured datasets and help ensure the accuracy of financial institutions’ internal data.
“For regtech to keep improving it’s important to encourage dialogue between banks and developers. Too often, fintechs barge into the decades-old financial space without appreciating the complexity of the global regulatory scenario,” says Kopczynski.
That’s why many fintech and regtech companies are working to foster collaboration between developers and financial institutions. In other collaborations, the regulator has taken the lead.
Beyond collaboration, there’s the challenge of aligning a sleek piece of regtech software with a hulking legacy system that’s been growing haphazardly for thirty years, what Kopczynski refers to as “a Frankenstein brainchild of hundreds of thousands of people.”
While there is a long road ahead before regtech is implemented industry-wide, the technology is already providing value. As the datasets related to finances continue to multiply and new laws are passed, there is a clear need for these tools in order to keep institutions current and compliant.
This content is produced by WIRED Brand Lab in collaboration with Western Digital Corporation.