Ongoing digitalization, unpredictable inflation, increasing cost of funds, changing consumer expectations and an emerging wave of defaults have combined to create turbulence in the auto finance industry.
The average interest rate on a new car loan rose to 8.95% in March, up from 5.66% a year earlier, according to Cox Automotive. For used cars, the rate hit 11.3% last month, up from 7.7% in 2022, according to Edmunds.
Delinquencies are on the rise, especially for younger borrowers, according to a recent study by the New York Federal Reserve.
Lenders are responding by undertaking a variety of strategic shifts. Some have raised prices indefinitely while others are pausing to regroup and reprice with an eye to avoiding undue risk exposure. Still, others are pricing aggressively to gain market share. These shifts are reverberating across the entire auto finance landscape.
Is the concern enough to cause strategic or technological shifts, and if so, where do auto lenders go from here?
Auto Finance News spoke with Be’eri Mart, head of banking and auto finance at Earnix, a global AI provider focused on optimizing and personalizing consumer lender experiences for a variety of financial institutions, on how lenders are managing their portfolios. What follows is an edited version of the conversation.
Auto Finance News: Words like “unprecedented” have almost lost all meaning over the last few years, but that still seems to best describe where auto lenders find themselves today. When you speak with them, what strategies do you see them employing, and are those strategies working in this environment?
Be’eri Mart: There seem to be as many strategies at work as there are lenders. Competition is fierce among the various players.
Some, such as banks and finance companies, are simply pricing their loans high to ensure they don’t lose money, even if interest rates continue to climb. The downside, of course, is that they may be pricing themselves out of the market and losing business to more aggressive competitors.
At the other end of the spectrum, captives are hoping to drive demand with very attractive rates. Credit unions, likewise, are pricing their loans aggressively hoping the new auto loans may lead to other products down the road. Both are betting that the strategy won’t load their portfolios with poor-performing loans or lead to rising delinquencies down the road.
Others, including some non-traditional lenders are backing out of certain market segments and focusing on higher-income borrowers, or exiting the subprime market, for example, or just writing fewer loans overall.
I would say there is no “right” answer, but one thing we see is that many of these decisions are being made with a “broad brush,” lumping consumers into target segments, but those segments may be too large and undifferentiated to make analysis truly effective. We see a need to inject more precision into the equation.
AFN: As a data analytics company involved with many auto lenders and banks across the globe, what is the most frequently asked question you hear from your customers today?
BM: The overarching theme is this: How should lenders look at setting loan and lease pricing and managing portfolio risk in a market that is constantly changing?
We have been hearing for many years questions around the correlation between loan pricing and portfolio management and how to optimize the two. Pricing is a key lever to manage the trade-offs between profitability and market share, and lenders are looking for technology solutions to help manage those trade-offs and optimize their portfolios.
More recently, the discussion has expanded to include how lenders can also use pricing to control the risk profile of their portfolios, given the recent increases in cost of funds and the trend of rising delinquencies.
A key challenge in using pricing for risk management is that many lenders often use a “one size fits all” approach, only taking into account one or two broad parameters, because their existing technology is incapable of solving for segment-level pricing. Solutions are needed that allow granular pricing management to influence demand across narrower, more targeted risk segments.
AFN: How can lenders maintain volume without pricing so low as to threaten profitability and increase delinquencies?
BM: We urge our customers to take some very concrete steps right now.
First, quickly and proactively assess the current market and how it could affect the lender’s total portfolio. This must include shifting Treasury rates, competitors’ pricing, delinquency rates, car prices and macroeconomic trends.
Second, develop a detailed, intelligent plan and response. In terms of pricing, this should include a highly agile pricing function (driven by the latest artificial intelligence- and machine learning-powered pricing solutions) that combines analytics, go-to-market strategy planning and competitive intelligence.
Third, turn to technology. The best pricing software solutions today offer powerful automation capabilities – critical to eliminating excessive internal handoffs and accelerating time to market – as well as ML and AI capabilities that use self-learning cycles to improve pricing strategies and results over time. Pricing is the key lever and common denominator in the discussion, and AI and ML help make pricing granular and smarter by helping lenders understand their customers better.
Modern technology allows lenders to use predictive models to understand, at the customer level, differences in price sensitivity, profitability and risk profiles.
Combining optimization algorithms with these predictive models allows lenders to determine optimal price points at a very granular level, even down to individual consumers, a level of granularity far superior to looking at broad swaths of consumers, such as those with certain credit scores, income levels or loan histories.
Lenders also gain a new level of agility and the ability to react quickly to market changes and market feedback, rapidly assessing the effectiveness of their pricing strategies and adjusting as necessary in a “test and learn” or ”self-learning” cycle.
Finally, monitor and adjust in real time. In addition to developing better pricing strategies and offers and getting them to market faster, auto lenders should also continuously monitor pricing performance in real time. Ideas to consider here include real-time price deployment, dynamic A/B testing capabilities and ongoing monitoring in order to make the best decisions possible.
AFN: What is the role of data in assessing pricing options and attempting to predict the market response? What type of data should lenders be analyzing, and what kind of analytical tools should they use?
BM: Data plays a huge role in designing the solution and in its everyday operation.
To gain intelligence and paint an accurate picture of the current state of the market, lenders need to be able to factor in external data in real time, as changing market conditions (interest rates, competitors’ pricing, demand for new loans, macroeconomic data, etc.) must be considered.
Technology that provides a forecasting framework is absolutely necessary, one that allows changes in the cost of funds or competitors’ behaviors to be instantly reflected in demand and profitability forecasts.
Internal data on current customers and loan performance also factors in. With delinquencies on the rise, some customers may be paying late or teetering on the brink of default, and you may want to reach out with an offer to renegotiate.
All this data is instrumental in powering prescriptive analytics to not only optimize pricing strategies, but also to assess and adjust those strategies based on real-time performance data. This gives lenders more confidence that they can balance new customer acquisition efforts, financial performance and regulatory compliance to optimize their portfolios.
With modern analytics and the right data to feed the models, lenders can be much more agile in response to changing conditions, and rapid outcomes analysis allows them to assess the effectiveness of their strategies and adjust quickly if necessary.
AFN: A lot of lenders seem to feel that they need to overhaul their legacy systems before they can take advantage of newer technologies, which can be costly and time-consuming. How do you look at implementing new technology for loan pricing?
BM: You’re right. Many lenders have gone down the “all or nothing” path, only to be overwhelmed with the task and with the resource requirements, and as a result they never achieve the promised results.
At Earnix, we take a more agile and modular approach. Our strategy is one of building what we call composable solutions.
Some existing systems, such as loan origination systems (LOS), are perfectly fine as they are and need only be connected to a new pricing solution. And much of the data lenders need for smarter decision-making is already in-house but inaccessible due to disconnected, siloed systems and internal data management issues.
Through modular construction and the use of technologies such as application programming interfaces (APIs), we avoid “throwing the baby out with the bath water.”
That allows lenders to retain what works and to focus on quick wins, such as operationalizing their existing data, segmenting markets with more precision, deploying scenario modeling and automating their pricing strategies. This strategy allows for rapid solution deployment and reduced time to market, while always keeping the long-term strategy in mind.
AFN: What recommendations do you have for auto lenders right now for improving their pricing analytics? Are you optimistic about how lenders can adapt to current conditions?
BM: No. 1: Operationalize the data you already have at your disposal. As I mentioned before, customer behavioral models and loan profitability algorithms have often been developed over the years but need to be drawn together in order to perform portfolio analysis and to model various “what if” scenarios.
No. 2: Implement an integrated, purpose-built pricing analytics framework that will allow lenders to understand within minutes how a change in the market will impact their businesses and the best way to respond to those changes.
No. 3: This combination of data and modern analytics allows for the real-time deployment of pricing, giving lenders a competitive advantage through immediate responses, automated test-and-learn cycles and rapid deployment of new pricing models.
While the current environment may present “bumps in the road” for auto loan originators, it can also offer new opportunities. We’re optimistic that lenders can move away from cumbersome, time-consuming pricing approaches and embrace new technologies for faster, more-effective pricing, gaining a competitive advantage and delivering long-term financial performance.