Guest commentary: Compliance risks in indirect auto lending

Automotive lending has changed dramatically in the last decade, with financial institutions relying more heavily on indirect lenders to collect, manage and report data in accordance with regulatory requirements. However, with heightened Federal Deposit Insurance Corp. and Consumer Financial Protection Bureau oversight, banks face significant compliance risks; indirect auto lending is a growing market, and banks must provide greater oversight and establish effective controls to ensure quality data.

As competition in this space becomes fiercer, lenders must juggle the equally important metrics of speed and compliance. Banks report profits being squeezed as labor costs rise and compliance rules create more hoops to jump through. Fiduciary duty ultimately falls to the bank, so all partners and dealerships must be expertly trained to recognize inconsistencies, provide documentation and heed stipulations.

But how can financial institutions ensure these third parties are tracking the right disclosures in an efficient way so they can continue to scale this line of business? It's easier said than done.

Manual verification, their primary method of compliance verification, is prone to human error and slows the document review process significantly. Scrutiny is rising, with originations down, interest rates up and assets depreciating more quickly than ever. Additionally, as volumes go down, auto lenders are faced with increased credit risk and challenges around origination and delinquency ratios.

This creates a ticking time bomb of compliance issues that may well sink an institution if not taken seriously. Leaders are looking for an effective, scalable solution, and machine learning is the natural next step from manual data-scrubbing.

Compliance with regulatory requirements is crucial. Banks and lenders must create mutually beneficial relationships with dealerships to enhance the overall experience while putting controls and triggers in place to improve data integrity. However, as banks cut costs to maintain margins, executives are worried about getting pinged for compliance issues.

The largest cost and time drain lies with manual data-scrubbing. This human-centric method of validation, in conjunction with the growing volume of electronic files, often causes funding delays and requires costly additional staffing. The unpredictability in hiring needs further squeezes margins and greatly hinders bank leaders' long-term planning capabilities.

In terms of scalability and impact, manual data verification is untenable because it does not allow a financial institution to gain market share, as dealership onboarding is plagued with delays. And as decisions are delayed, the asset has already walked off, costing more time and money in the long run. Despite the growing list of concerns, most banks are accepting the risk and are content to manually review all data.

Compliance issues are entirely avoidable with automation technology. By utilizing machine learning integrated into existing digital onboarding experiences, contract processing time can be cut by almost 85 percent, with incoming documents from dealerships and borrowers being quickly and efficiently indexed into a loan origination system.

Machine learning collects and extracts missing data from borrower and dealership documents, synchronizes data across systems of record, and reduces exception queue review times by minutes per loan. It can auto-fund loans with no stipulations and auto-fund loans once stipulations are corrected, cutting overall processing time by 90 percent. As touch points by staff decrease, efficiency and data integrity increase.

From the minute a client's loan information is submitted into a machine learning platform, the data is being effectively classified, assembled and flagged for inconsistencies. All condition and stipulation review is automated, and the data from retail installment contracts and dealer jackets is being auto-compared and auto-calculated. Suddenly, a tedious job spread over days or weeks by several people can be completed to regulatory standards in mere minutes.

By turning the document review process into an exception-based process, staff members are able to eliminate quality control issues associated with outsourced processing services and provide more quality service to customers.

While most banks accept the risk of manual auto loan processing, some leaders are looking toward the future of data processing by implementing trusted machine learning algorithms for their compliance systems. In the competitive auto lending industry, banks willing to invest in automation have an edge as FDIC and CFPB requirements demand data integrity. The time to automate is now.