Peak Auto? Our Latest Proprietary Density Survey Suggests Otherwise

Overview:

• "Peak Auto" is based on forecasting that fails to capture vehicle density and installed base/scrap shifts.
• Our latest analysis suggests a more resilient U.S. auto demand outlook than consensus incorporating our proprietary vehicle density survey and scrap age analytics.
• We are raising 2025 to 2027 U.S. seasonally adjusted annual rate (SAAR) estimates and expect low-single-digit industry revenue growth.
• Macro risks are elevated, but our work reinforces the above-consensus estimates.
• Our analysis bodes well for both automaker and supplier stocks.

The TD Cowen Insight

"Peak Auto" calls have long been based on forecasting methods that fail to capture two critical trends. We've developed proprietary tools to incorporate these trends – our vehicle density survey and scrap age analytics. Both point to a more resilient U.S. auto demand outcome than consensus expects, reinforcing our generally above-consensus estimates for the group.

Our Thesis

Traditional auto forecasting methods have misled investors on numerous occasions over the past 15 years, including through recent years' Peak Auto calls that continue to weigh on auto trading multiples. The issue with traditional methods is that they tend to miss two complex yet critical trends – the future direction of vehicle density (vehicles per household) and installed-base/scrap shifts. Based on our work, we disagree with the Peak Auto narrative. Our June survey points to a 3.3%-plus expected rise in U.S. vehicle density over the next two years, which is slightly better than the 3.0%-plus observed in February.

Similarly, the "More/Fewer" ratio remains resilient at 1.64x vs. 1.62x in February. Our latest scrap age work also suggests gradual upward pressure in the coming years. Combined, these factors should yield a stronger U.S. auto demand outcome than consensus expects. Though macro risks are elevated, our work leaves us more confident about our generally above-consensus estimates for auto stocks.

What is Proprietary?

In early June we ran our proprietary vehicle density survey, which we believe gives us an edge in forecasting the future direction of vehicles/household. Without a survey like this, we think it becomes very difficult to properly forecast U.S. auto demand.

We are also introducing a unique vehicle scrap model that we feel gives us a further modeling edge.

Financial and Industry Model Implications

Programs from warehouse clubs have generally been more successful than those from other subsegments of retail. Further, the more a program is indexed to higher income customers, the more attractive it is to both retailers and issuers. Other desirable traits include fast growth and top-of-wallet status. We believe this will continue to be the case for retailer card programs, and retailers with such characteristics should have more negotiating power.

We also note that credit card programs drive on average 40-60% of department store earnings before interest and tax (EBIT) as credit card revenue could represent 4-6% of net sales and flow through an estimated 95% EBIT rate to operating income.

What to Watch

There are two unique industry trends to consider when modeling how U.S. auto demand will show through in the numbers. The first is new vehicle inventory, which once again sits at low enough levels to somewhat limit immediate SAAR growth. The second is lingering vehicle affordability constraints. Given these factors, if our call on underlying auto demand proves accurate, it will be interesting to see which path automakers take to leverage that demand: Do automakers raise production and lower price to unlock volume? Or, do they maintain lean inventory and ultimately capture price/mix tailwinds at the expense of SAAR. Our second half of 2025 bias is to the latter scenario while our out-year bias is to the former. Tariffs will likely also shape automaker decisions in this regard.