Outlook from the 2024 AFSA Vehicle Finance Conference

I recently attended the American Financial Services Association (AFSA) Vehicle Finance Conference in Las Vegas as a representative of Digital Matrix Systems.? This was my second consecutive year attending, and as I anticipated I found my time there valuable.? With the challenges currently facing the auto finance space including increasing delinquencies and fraud, I was eager for the discussions and presentations.? I also made some great connections during the conference.

Topics included the state of the market, legislation, AI, fraud, and the EV market.? I enjoyed a keynote presentation from Gian Paul Gonzalez, a ninth-grade teacher from New Jersey who had an inspiring message that started in a talk with the NFL’s NY Giants about being “ALL IN” including how words and actions can have so much more power than we realize.? He shared that an ‘all in’ approach is easy when you are winning.? Being truly “ALL IN” is about true commitment, especially when things aren’t going well.?

A great takeaway from his message was this: fame is about what we do for ourselves; greatness is about what we do for others.

Here are a few takeaways from the other sessions I attended:

State of the Market with Rob Kurnick, Penske Auto Group and Kyle Birch, GM Financial

• Sales forecast for 2024 of 15.5 – 16 million new cars
• There is pent up demand for both new and used cars
• Credit is available and financing has been strong, mainly in the prime sector
• Longer finance terms are causing downstream issues with respect to negative equity for loans that have 72- and 84-month terms
• Car buying is still a relationship business, and dealerships play an important role in terms of customer engagement
• Electric vehicles (EVs) are posing an interesting challenge in the market in terms of general buyer anxiety and more specifically, range anxiety.? The maturity (or lack thereof) in terms of infrastructure needed and reliability is a key factor.? ICE vehicles are generally still better for the dealer across the board — margin, service, and sales cycle.? A more measured approach may be to encourage a move to hybrid vehicles first to ease the eventual transition to EV and in the process eliminate the anxiety that EVs present to US buyers.
• Fraud is often perpetrated by employees and weaknesses in the regard versus system issues.
• For the next 12 months, a significant industry challenge is vehicle affordability.? Interest rate reductions and a return to ‘normal’ inventory levels will help correct this.

Takeaway:? in spite of some challenges, the market should see an uptick as things start to balance out

AI Risks and Opportunities – Scott Zoldi, CAO Fico and Sanjiv Yajnik, Capital One

• AI (generative) has resulted in false accusations --- can/should it be trusted?? As one example noted by the presenter, a Stanford study found false answers to medical questions.
• Issues with lack of explainability and accountability have resulted in some organizations banning use.
• Hackers are using ChatGPT and other Generative AI tools to create phishing emails and commit other fraud.
• A survey of 100 North American Financial Services companies found that 43% struggle with AI governance, 71% indicate that AI ethics and responsible use of AI is not part of their current operations, 50% lack AI model development standards, and 27% haven’t started refining responsible AI capabilities.
• Robust AI initiatives take appropriate time and resources for data prep, analysis, model development and testing.
• There are issues with explainability of AI for models and in many cases it is insufficient for regulatory purposes; leading to adoption of interpretable machine learning (ML).? ML isn’t viewed as either good or bad, but is designed to minimize prediction error and can find proxies for protected classes.
• An audit trail is needed to enforce model governance when using these methodologies.

Takeaway:? AI and generative AI should be explored, with an emphasis on learning and a steady (neither too fast or too slow) approach.

The Electric Vehicle (EV) Market – Nishit Madlani, S&P Global Ratings; Elizabeth Krear, JD Power; Jonathan Smoke, Cox Automotive

• A significant supply and demand imbalance is creating price pressure for EVs.? On the supply side there is a global race for battery production.? On the demand side, consumers are hesitant to adopt due to affordability and range concerns.
• US infrastructure in its current state is causing range anxiety.? In 2023 there were 178K public chargers available, with projections indicating there will be 1.2 million chargers by 2030.
• The EV industry remains predominantly premium vehicles such as Tesla, resulting in an affordability factor.? Residual values are improving.
• Projections indicate that retail share will be 46% by 2030.
• There could be an increase in leasing of EVs for those who don’t qualify for tax credits.

Takeaway:? Cost and convenience are significant barriers to adoption of EVs, generating a need for innovative financing approaches as well as overcoming consumer anxiety, but growth in this space is coming.

Vehicle Finance Fraud – Panel discussion with Erika Deitrich, ACI Worldwide;? Frannk McKenna, Point Predictive; Geoff McClellan, Ford Motor Credit; Richard Tsai, Trans Union and Capt Brian Cole, Las Vegas PD

• Synthetic fraud is on the rise and no longer just a subprime issue.
• The time period from 2018 – 2021 saw a decline in activity partially due to the pandemic, however since 2021 there has been a drastic spike.
• Cybercrime is considered one of the top global risks over the next decade.
• An increase in training and education is needed.
• Anticipate an increase in activity as AI makes fraud easier to commit and harder to detect.
• Fraud as a Service is emerging as subscription models make fraud schemes easy to deploy.
• AI models built to detect fraud can help identify different types of fraud including account takeover, synthetic ID, and credential stuffing.
• Organizations should create a multi-layer approach to fraud detection, and tools to measure the impact of fraud.
• Balancing a seamless/frictionless approach with the appropriate level of interrogation is critical; KPIs related to fraud should be created.

Takeaway:? Cybercrime, particularly schemes powered by AI will continue to increase and it is imperative that organizations implement tools, measurement, and investigative techniques to minimize the impact.

?I’d love to hear your thoughts and any additional takeaways if you attended AFSA Vehicle Finance as well. Where do you think the EV market is headed??? Has your organization employed AI models or leveraged generative AI??? What use cases have you found most effective so far??? With respect to fraud, what schemes and challenges are causing you the most concern?

Thanks for taking the time to read my notes.? This was a great conference and I look forward to seeing everyone next year in New Orleans!