Report Extract: Innovation in risk models

In this extract, Paul Childerhose, Strategic Advisor to Prodago and the CRTA, considers whether RegTech has the ability to solve regulators’ problems and tells us the most exciting use cases in which he’s seen AI/ML being utilised.

Report Extract: Innovation in risk models

In this extract, Paul Childerhose, Strategic Advisor to Prodago and the CRTA, considers whether RegTech has the ability to solve regulators’ problems and tells us the most exciting use cases in which he’s seen AI/ML being utilised.

Following the success of our report, RegTech for Regulatory Change, in collaboration with Burnmark, CUBE has extracted the key interviews with RegTech experts.

In this extract, Paul Childerhose, Strategic Advisor to Prodago and the Canadian RegTech Association, considers whether RegTech has the ability to solve regulators’ problems and gives tells us the most exciting use cases in which he’s seen AI/ML being utilised.


You have worked in several roles within the regulatory ecosystem and have a view of banks, RegTechs, law firms and supervisors in the Canadian market – could you explain your current role?

Based in Toronto Canada, I provide strategic advisory services to a data governance SaaS company called Prodago. I work on product research, design and development of use-cases for performing assessments of data management risks that will have a direct impact on successful deployments of AI / machine learning initiatives within both government and the private sector. Additionally, I serve as a Strategic Advisor to the Canadian RegTech Association, where my focus is on fostering collaborative engagement with the key participants from supervisory bodies, law firms, financial institutions and tech incubators.

Regulators around the world are increasingly using technology to solve some of the challenges in the regulatory world. Do you think technology has the ability to solve regulators’ problems?

Regulators should assess nascent and emerging technology from the perspective of how they can be both informative and directly influence the design of future state regulatory structures. They are really in the phase of research and policy development mode right now.

Traditionally, there have been two primary functions within a supervisory organisation – develop policies relative to mitigating financial risks to the public and then monitor and enforce the banks’ level of adherence with the regulations. Now we need the supervisors to think like the banks and place a premium on technology as a value creator within their own organisations, and not merely an expense.

The volumes, variety and velocity of data that regulatory supervisors are consuming in both ad hoc and scheduled reports submitted by regulated entities will continue to grow exponentially, so the design and implementation of a data management strategy that employs machine learning technologies would definitely help with addressing high labour costs.

Do you think a strong collaboration model exists in CanadianF intech?

We have a strong financial services regulatory regime that is a result of the co-operation and collaboration among the Office of the Superintendent of the Financial Institutions (OSFI), the Department of Finance, the Bank of Canada, the Financial Consumer Agency of Canada (FCAC) and the Canada Deposit Insurance Corporation (CDIC).

What are some of the most exciting AI/ML use cases for you in the RegTech ecosystem?

Having spent 2+ years as Director of Data Governance for a bank’s enterprise-wide anti-money laundering program, I can say without hesitation that this area of financial crimes risk management presents the most exciting use-case for introducing AI/ML into the business operating model. It’s been well reported for many years that traditional methods for detecting unusual or suspicious activities taking place is entirely inefficient, ineffective, and manually intensive. Banks have over invested in people in this space and under invested in technology, while massive amounts of data emanating from billions of customer transactions is not being used to its full advantage.

The current approaches are limited to relatively simplistic linear rules that look at value and volume profiling and use simple static single dimension models to attempt to detect anomalies….

 


Related resources

Top 10 leading voices in financial services

Top 10 leading voices in financial services

Ever wondered who the most influential figures are in regulatory financial services from the past 8...

View More