Machine learning for auto lenders: How to boost productivity
There’s an increasing amount of buzz and excitement around machine learning (ML) and artificial intelligence (AI). These terms are often used interchangeably, but they’re not the same. AI is a broad term describing technology that behaves and solves problems like a human. ML is a subset of AI in which machines are given access to data sets. The machine learns, adapts and responds based on historical and new data, without additional programing required.
Most people have already experienced machine learning
ML is already here, and lenders probably use it every day. Years ago, Amazon developed a recommendation engine to suggest products to customers based on their purchase history and other factors. Some estimates indicate that one-third of Amazon’s revenue has been attributed to its recommendation engine. Streaming services like Netflix make recommendations using ML and algorithms. Google uses machine learning in its search engine, Google Autocomplete, and advertising through responsive search ads. Spotify uses ML to build recommended play lists.
Machine learning for auto lenders
There are several practical applications for ML in auto lending. For example, it can be used to increase efficiency and accuracy of credit decisioning capabilities and verify borrower-provided documents.
Improving underwriting, risk and predictions with ML
Credit Scoring models built by the lender’s risk teams have traditionally relied on custom attributes. These attributes are built from data obtained from the application, bureaus and other data sources. These models are resource-intensive to maintain since they must be both compliant and standardized across the major bureaus. More recently, lenders have opted to purchase attributes from providers. As a result, they get access to a more extensive data set which are sometimes hundreds of attributes — while offloading the compliance and standardization burden to the vendor. Equifax‘s Advanced Decisioning Attributes and Experian‘s Attribute Toolbox are popular products in this space.
A newer trend in modelling involves the use of ML. Several providers, Zest.ai for example, specialize in using ML to analyze a lender’s portfolio and then build an ML-based scoring model for credit decisioning. The use of ML and AI in underwriting typically raises compliance concerns. This is one of the main reasons to consider working with a ML vendor: They have the ability to explain the ML decisioning model to satisfy regulator concerns.
Saving time on loan verifications with machine learning
Verifying borrower-provided documents, such as a driver’s license or paystub, is both time consuming and vulnerable to human error. ML can be used to digitize this process. Typically, a lender’s funding team or an outsourced function manually examines the documents to perform data entry and verify authenticity. Trained ML tools can examine documents for authenticity and complete the verification process quickly with a high degree of accuracy. Providers like Informed.IQ offer ML-based document verification services that are already being used by lenders like Ally Financial.
Two roads to bring learning to life
The most straightforward path forward for lenders looking to start using ML is to partner with a vendor that specializes in the technology, which allows lenders to benefit from both the technology and the vendor’s knowhow. Some lenders already have the expertise in-house or the resources to make it happen. Those lenders may opt for just the technology by working with a provider like H20. For a more integrated option, lenders can choose leading loan origination software providers like IDS (formerly White Clarke Group), that currently work with third-party integrations to make the software easier and faster for lenders to deploy and use.
No matter which option lenders choose, those who embrace technology traditionally outpace their competition in growth and customer satisfaction.
Shim Mannan is the executive vice president of product and business development in the Americas for IDS (formerly White Clarke Group).
Auto Finance Summit, the premier industry event, returns October 27-29 in Las Vegas. The Summit continues to bring together the best and brightest in the industry year after year for unparalleled networking and professional education. To learn more about the 2021 event and register, visit www.AutoFinanceSummit.com.
