Intro to Natural Demand
The concept of Natural Demand for automobiles in the US market was first established in 2012. At that time the global economy was still in recovery from the Great Recession. GM and Chrysler had suffered through bankruptcy, layoffs rippled through all layers of the industry and collectively automakers had scaled back vehicle production in the US by roughly 30%. It was painful, but necessary medicine that reinvigorated the US auto industry, allowing the sector to lead the macroeconomic recovery of North America.
Once the 2009 production cuts were in place, took years to ramp up to pre-recession levels. Many of the assembly plants were shuttered and reopening a plant is no small task. The resulting shortage of vehicles, combined with a robust rebound in consumer demand for automobiles led to (at the time) all time highs in new car average transaction prices and used vehicle values. (BTW.. Does this all sound familiar?) Discussions of pent up demand were rampant. Indeed, between 2008 and 2012 there more vehicles sent to the junkyard than new vehicles sold contributing to a big imbalance of demand vs. supply. Automakers were scrambling to maximize production where they could. It was only a matter of time until vehicle production caught up with consumer demand. But when would that occur? What fate awaited the auto industry once pent up demand was satisfied?
In an effort to understand these questions, I developed the “Natural Demand” calculation. The formula was borne of basic economics: understand the underlying metrics of consumer demand and cross reference it against vehicle supply. Like an ideal dinner recipe, the ingredients are succinct, readily available, easy to follow and go back decades (with delicious results). They are:
SUPPLY INGREDIENTS:
Vehicles in Operation (VIO) - the number of vehicles currently registered in the United States (aka the car parc). This data set represents all vehicles from the most recent vehicle processed at the DMV to Jay Leno’s 1931 Duesenberg.
New Vehicle Production - units being introduced into the car parc
Scrappage Rate - units being removed from the car parc
DEMAND INGREDIENTS:
Driver’s Age Population (16 years and up)
Driver’s License Penetration Rate - how many people within the Driver’s Age Population cohort possess a driver’s license
Vehicles per Licensed Driver - the average number of registered vehicles for each licensed driver
I love analogies and I think of Natural Demand as the landscape of a dam with a lake behind it and a river downstream from the runoff. The lake itself represents vehicles in operation. Rivers of newly produced vehicles continue to fill the lake. The dam itself is made up of the driver’s age population. As the population grows, the dam gets bigger and able to hold more water in the lake. And lastly, the behavior of that population determine how much water is released through the dam, resulting in scrappage.
Let’s unpack the demand ingredients first. Not only is it easy to obtain data on the current population of people 16 years and older, it’s easy to forecast as well. Today’s 10 year olds will be 16 in, you guessed it… six years! The other two demand factors are easy to establish and remarkably stable over the last 30 years. This makes them easy to forecast as well. In fact, the average number of vehicles per licensed driver has been a steady average of 1.21 since 2001 with a standard deviation of just 0.013. The high during the last twenty-two years was 1.237 in 2006 with the low being 1.190 in 2010 (just after the government sponsored cash for clunkers program removed many vehicles from the car parc). It’s safe to assume this factor will continue at 1.2 for the foreseeable future.
Driver’s License Penetration Rate is a bit more volatile than vehicles per driver, but not by much. Since 1990, 87.2% of eligible people have had a driver’s license. The standard deviation of this factor is also low at 0.010. With that said, the high and the low points of this factor have occurred in two of the last three years. The high of 89.37% was in 2022 and the low of 85.95% in 2020. Occam’s razor suggests that shuttered DMV offices during COVID lockdowns contributed to that 2020 figure and the high point in 2022 was the rebound once the pandemic was largely in the rearview mirror and folks sought freedom of personal mobility. It also appears the theory that Millenials and Gen Z aren’t interested in driving did not come to fruition, meaning a penetration rate of 87-89% is reasonable to expect going forward.
For supply side factors, scrappage rate is nearly as consistent as the above demand factors. Average scrappage rate since 1990 is 5.17% of all vehicles with a standard deviation of 0.61 percentage points. As with license penetration rate, the 30 year low occurred recently in 2021 when only 4.2% of vehicles were sent to the junkyard while the high was way back in 1992 at 6.71%. It’s reasonable to assume that scrappage will hover around the 5.0% range given the increasing lack of affordability among new and used vehicles.
New vehicle production is the wildcard. History has proven many times that when given the chance, automakers will accelerate production. The capital intense development cycle and insatiable quest for market share seems to always result in the oversupply of vehicles into the market. Most industry analysts agree that a return to the pre-COVID total sales rate of 17M units is not too far off and the automakers will likely push that number towards the 18M range by the end of the decade.
Lastly, Vehicles In Operation (VIO) is just the outcome of all these factors. It has climbed steadily over time with population growth and has only deviated slightly from the long term trend line during prolonged periods of high incentives (the aughts) and supply shortages following the pullback in production due to recessions and marketplace disruptions (early 2010s and 2020s).
PUTTING IT ALL TOGETHER
You still with me? Of course you are! We must accept that all market conditions have flexed and adjusted appropriately to arrive at the number of units the car parc at the present moment. In other words, the vehicles were produced and sold and a price necessary to move them off a dealer’s lot working in conjunction with the level of inherent Natural Demand. The price vehicles sold for reflect the supply and demand dynamics we’ve been discussing. So we use the latest 2022 data as a starting point and apply our highly consistent and reliable factors to understand what the Natural Demand for all registered vehicles in the United States will be throughout the rest of the decade.
The last step is subtracting for the scrapped vehicles from the car parc and adding vehicles that arrive in the form of new production and sales. This data is represented below in the red bars and represents the forecast based on the factors provided in the table at the bottom of this page. The variance between these values informs us of the imbalance of supply relative to demand. A supply shortage will cause prices to increase and srappage to decrease (thereby increasing US fleet vehicle age). Oversupply will cause vehicles produced to be forced into the market via price discounting.
The below chart represents the imbalance of Natural Demand vs. the Car Parc forecast for each year. Discounting should accelerate in 2026 with additional pressure in 2027 when the oversupply of vehicles in the market exceeds 3 million units.
Below are the individual factors contributing to the Baseline scenario. In further posts we will explore how variability in these factors influences the expected Natural Demand and Car Parc forecast.
NOTE: This post was updated to provide details of each individual factor in the baseline forecast and fix an error in the initially published calculation.
