Traversing fair lending compliance with automation

By Tyler Barron

Machine learning (ML) is a hot topic in the financial industry — especially now as banks try to navigate elevated regulatory pressure, labor issues and increased credit risks. As technology advances and regulations become more stringent, utilizing machine learning in everyday bank processes is becoming a necessity. In an effort to stay competitive in a fierce financial space, financial institutions must look for new ways to solve these problems.

Banks have been focused on digital transformation and automation for years now, but recently there has been a stronger push toward leveraging machine learning in the compliance space. Institutions have realized that they must incorporate intelligent systems into their operations to reduce human error, eliminate source document verification and lower overhead costs. 

As a result of integrating these ML systems into their existing processes, banks are mitigating compliance risks and increasing efficiency by significant amounts. Implementing strong controls on how banks process data is crucial to thriving in the ever more regulated fair lending space.

Following are two important ways machine learning is helping banks traverse the current and future fair lending compliance landscape.

IMPROVE DATA INTEGRITY

Fair lending compliance is a critical risk area for banks today — with scrutiny from government regulators only intensifying. Under increasing regulatory pressure to deliver accurate data in a timely fashion, it has become imperative for lenders to improve data integrity in line with vigorous HMDA enforcement, CRA expansion and the Dodd-Frank 1071 changes just over the horizon. 

The new fair lending requirements and regulatory pressures essentially require compliance automation in order to provide banking institutions with the confidence in the data they are providing for HMDA and CRA. Automation systems, for example, can generate daily reports that reveal whether any data integrity issues exist within your current loan portfolio and directly route users to resolve these discrepancies — a task commonly referred to as “data scrubbing” that is fraught with human error and requires additional staff. 

For many Texas banks, ML has proven to be a more scalable solution by enabling deeper loan penetration for reviews and addressing hiring challenges. The one-two punch of increased regulatory pressure and a shortage of compliance professionals has created a scenario where banks must either accept more risks by reviewing less loans or spend more and staff up. Without augmenting through ML, many banks are forecasting having to double their compliance teams in order to prepare for these upcoming regulatory changes. 

Rhonda Carroll, the chief compliance officer at Houston-based Prosperity Bank, one of the largest banks in Texas, has firsthand experience leading her team through the changing, complex requirements for fair lending data reporting.

“As we were planning for new regulations on the horizon, we realized it was going to require increasing staff to implement,” says Carroll. “That’s why the bank decided to bring on automation to support the tasks associated with the data validation.”

Maintaining data integrity is a #1 priority, and ML offers the tools to speed up processes and improve outcomes. Machine learning works within your institution to mitigate risk, reduce expenses and free up your staff to provide better service to your customers — a worthwhile investment that will pay dividends in the long run.

REDUCE LABOR COSTS

ML can help banks reduce their document review time and costs — a typical bank might require a team to manually review each application that requires supporting documents to be verified. By using ML algorithms, banks could reduce this manual review time by as much as 80% by having machine learning flag errors or discrepancies. When implemented, ML algorithms facilitate over 90% straight-through processing of all reviewed documents. 

Having artificial intelligence verifying and double-checking documents does a few things: One, it drastically reduces human error and guarantees the institution will satisfy applicable regulations. Two, by utilizing a centralized, semi-automated document verification solution, financial institutions dramatically reduce labor costs, allowing the money that would have gone to staff pay to be invested through other avenues. And three, decreasing the number of monotonous tasks staff must complete enables them to serve more customers and increase revenue by speeding up the application process for new loans or credit cards. 

Intelligent document processing, when integrated into current systems, can perform a variety of time- and money-saving processes. It can quickly determine which delinquent loans should be pursued for collection and which are legally protected, saving your institution potential legal and reputational risks. ML can also automatically sort, scan and track customer mail while allowing your bank to track their correspondence from customer service inquiries, collections notices and court documents, as well as read judgements and bankruptcies to protect banks against CFPB scrutiny of loan servicing and collection practices. 

Machine learning offers a thorough and comprehensive solution to every bank’s compliance problem and works with their existing infrastructure to create a sustainable, scalable solution to rapidly approaching stringent compliance regulations. Accuracy and completeness are at the forefront of every bank director and compliance officer’s mind as they explore best practices for compliance, labor challenges and financial pressures.

As financial institutions lean more and more into automation, banks must stay ahead of the curve and start incorporating machine learning into their everyday compliance tasks. The current environment of economic uncertainty and increased regulatory pressure has created the necessity for financial institutions to examine their existing compliance systems, the scalability of those solutions and find alternatives to solving these challenges that can address both the financial and regulatory pressure that Texas banks are currently facing today.

Tyler Barron is the chief revenue officer at Encapture. Encapture is an intelligent automation platform, providing banks with documentation efficiency. Using machine learning and AI technology, Encapture helps banks reduce compliance risks associated with Fair Lending Guidelines, reduce overhead costs and improve profitability in a volatile market.