AUTO LOAN PORTFOLIO RISK MANAGEMENT: PROBLEMS TO AVOID

Experian’s State of the Automotive Finance Market – A look at loans and leases in Q2 2018 provides an excellent, easy-to-digest snapshot of the current market. It acknowledges slowing market growth, but paints an optimistic picture of borrower willingness to assume debt as a result of increasing rates and vehicle prices. Key findings from the report include:

• Record high auto loan balance of $1.1 trillion;
• Rising rates across all tiers—3.47% for super prime, 14.93% for deep subprime;
• Credit scores showing year over year improvement—average of 715 for new vehicles and 655 for used vehicles;
• Longer term loans (61 to 72 months) make up ~42% of loans;
• Deep subprime and subprime balances combined fall below 19% share of the loan market;
• New deep subprime loans average 72-month terms;
• More than 90% of deep subprime loans are for used vehicles; and
• Improvements in 30- and 60-day delinquency rates.

None of this comes as a surprise to experienced auto lenders. A growing economy and declining unemployment rate are prime factors in the willingness of borrowers to assume greater debt. Buyer preference for trucks and SUVs with a high price tag is another factor contributing to the record auto loan balance.

With respect to risk, lenders who focus on super prime and prime markets may be on easy street for now. Lenders who focus in non-prime and below segments need to be cautious, particularly as loan terms extend beyond 60 months. Regardless of your market segment, auto loan portfolio risk management needs to be a continual focus in your lending practice. There are always opportunities to improve portfolio performance by reducing risk, assuming you can identify it. The best tool for auto loan portfolio risk management is analytics.

Lenders who have adopted analytics report improvements across all areas of their lending practice. However, the ways that you implement and use analytics can be quite varied. Based on our experience in working with lenders, we offer guidance to help you avoid two major problems and gain the greatest benefit from your investment in analytics.  

Auto Loan Portfolio Risk Management: Problems to Avoid

Problem: Homegrown Analytic Tools and Methods

Only in recent years have industries that are data-intensive realized the value of using analytics to obtain ever-greater insight and value from their data. With the volumes of data acquired and generated throughout the lending cycle, lenders have come to acknowledge the value of analytics. However, for lenders that still use legacy lending software analytics is absent. When that software was developed years ago analytics was not even on the radar scope. As a result, lenders resort to home-grown analytic tools created with Excel or Access and built by the in-house guru, who is often the only person who really understands the functionality.

Lenders that we’ve worked with have mentioned several problems with this homegrown approach to analytics, including:

• Dependency on the individual(s) who created the one-of-a-kind analytics tool to maintain the software (what if they decide to go elsewhere?);
• Data definitions and formulas—LTV, PTI, DTI—used in the analysis may not accurately reflect the formulas used in the underwriting process, causing confusion when interpreting reports;
• The process required to extract, translate, and load (ETL) portfolio data and run the analysis and distribute reports to individual users is not automated; and
• High IT overhead for maintaining homegrown analytic tools.

Solution: A Modern LOS with Fully-Integrated Analytics

Modern lending solutions that provide fully-integrated analytics capabilities avoid these problems by providing:

• An intuitive interface that allows business users—underwriting, funding, servicing, field reps, and executives to easily create customized analytic reports that provided summary and detailed portfolio analysis of interest to them;
• Immediate and secure access via the internet. There’s no need to wait for someone to run the reports and distribute them;
• Accurate data definitions and formulas that eliminate any confusion regarding the interpretation of reports; and
• Near-real-time results that show the current state of the business for an accurate and timely information, allowing you to quickly address any issues.

Analytics is an essential tool for auto loan portfolio risk management. If your lending software does not include fully-integrated analytic capabilities, you’re not getting a timely and accurate picture of your portfolio. You’re forgoing the opportunity to identify potential and actual sources of risk throughout your lending processes and portfolio and quickly act to mitigate them.

Problem: Narrow Analytics Focus and/or Underuse

A modern lending system with fully-integrated analytic capabilities lets you explore and investigate the wealth of process and portfolio data you’ve acquired and created. However, the value of the analysis is a function of the analytic questions you ask. Lenders new to analytics often underestimate the power and reach of analytics in identifying risk.

A good analytics strategy takes a lifecycle approach, focusing on each stage of the lending cycle, regularly monitoring trends, and looking for anomalies. Apply that analytic insight to make adjustments to policies and processes to reduce (and ideally eliminate) risk and improve performance.

Solution: Analyzing Every Stage of the Lending Lifecycle

Taking a lifecycle approach, a few examples of the types of analysis that can uncover potential risk in you lending processes and portfolio might include:

Applications  

• Do any sources—dealers, geographies show increasing evidence of fraud? If so, what fraud avoidance strategies have you put in place?
• In what areas are applications increasing/decreasing? Do the trends correspond to your business plan or are these unexpected?
• What is the distribution of application sources—mobile, online, dealer? Are there any trends that show changing buyer trends that you can exploit?

Underwriting

• Who are your most/least efficient underwriters in terms of process time?
• Do any underwriters demonstrate an unusual number of overrides?  What are the reasons for the overrides? How have past loan overrides performed?
• What percentage of loans were auto-approval? With the improving economy could you loosen credit policies somewhat for auto-approvals without increasing risk?
• What percentage of applications were auto-declined? Would any of these applications be considered if both bureau scores and alternative credit data were used to determine creditworthiness?

Servicing

• Do any dealers show unusually high delinquency rates for specific loan classes or vehicle models? Is this the result of a new finance manager at the dealership?
• Has the borrower’s credit score changed since loan origination? Does that change present greater/lesser risk regarding ability to pay?
• What is the historical probability of subprime loans with 72-month terms and 60 DPD will default? How should credit policies be modified to reduce this risk?
• Is there a strong correlation between defaults and loan type, underwriter, dealer, or other loan characteristics that can be remedied by addressing these risks early in the lending cycle?

These examples only scratch the surface of the types of analytic questions a lender may ask.  Powerful analytic capabilities fully integrated into your lending solution give lenders endless means to analyze lending data and dig deep to uncover the root causes of potential or actual risk. The biggest problem lenders may face with analytics is the failure to use its power and reach to the fullest potential.