An assessment of supply is critical to understanding future residual price trends, though difficult to calculate. Our aim here is to help readers better understand the drivers of used-vehicle supply in the U.S.

In 2015, about 17.5 million new cars and trucks were sold in the U.S. Let’s assume that all of these vehicles were 2015 model year and that such vehicles are not sold in adjacent years. In the future, this cohort of 2015 vehicles will be traded in and repoed, bought from used car dealers, exchanged privately, totaled by insurance companies, driven into the ground, repaired, stripped for parts and ultimately scrapped. Some 2015 model year vehicles will be exported to other countries, while a much smaller number of foreign vehicles will be imported. 

What we know for a fact is that 2015 vehicles will never again be produced. The stock of such autos can and will decline in future, but it can never increase.

There are actually only three ways to increase the supply of 3-year-old cars in the U.S. First: produce an extra new car today and then wait for three years. Second: import an appropriately aged used vehicle from overseas. Third: rescue a vehicle that, for whatever reason, had been removed from the current cohort of 3-year-old vehicles.

The first of these is the main driver of used vehicle supply. We find, quite reliably after controlling for demand-side ructions, that declines in used car prices lag increases in new vehicle sales by between 18 months and two years. 

In terms of used vehicle exports, data are somewhat spotty, but reveal that in 2013 about 800,000 used vehicles were transferred abroad. Used cars and trucks typically flow from richer countries to poorer countries so apart from the odd vintage car enthusiast bringing home a beloved 1964 Porsche 911, we can surmise that vehicle exports exceed imports by a healthy margin. One can imagine that exports are impacted by domestic prices, rising when U.S. prices are suppressed and falling when local prices are relatively high. 

The third supply driver — the rate at which older vehicles are retired — is perhaps the most interesting of the three. We know from the Cuban experience that when new vehicle production or imports are severely curtailed, the value attached to older vehicles rises apparently without limit. In the U.S., the supply dynamics are less stark but people still make supply decisions when determining, for example, whether to repair the blown head gasket on their 1998 F-150 or buy a different car. Cars are often retired when they still have many years of safe motoring ahead of them. People may even retire perfectly good cars just because they like the features offered on newer models. If the propensity to retire vehicles ever rises, this represents a supply contraction that will potentially impact used vehicle prices.

Just as important is a consideration of the things that do not increase the supply of used cars. If, for example, my neighbor offers to buy my three-year-old hatchback and I agree to sell it, there is no change in either the demand or supply position of the market. Yesterday I was consuming the vehicle and my neighbor was catching the bus to work; today our roles are reversed.  

The next question we must address is whether these dynamics change if a dealer acts as a middleman to the transaction between myself and my neighbor. While the dealership is providing an indispensable retail service by matching my needs with those of my neighbor, the only way a used car dealer can impact supply is by opting to send a vehicle in their possession off to the salvage yard.

There is no doubt that dealer inventories can rise if there are too many sellers and not enough buyers. This issue is currently a major concern because of the flood of off-lease vehicles that have been present in the industry over the past few years. The critical thing to keep in mind is that the total number of 2015 vehicles in circulation in the U.S. is unaffected by the nature of their original financing arrangements. For off-lease volume to be a genuine problem we would need the demand for clocking miles in three-year-old cars to decline relative to the normal level. There is no evidence that this phenomenon has recently occurred.

So what explains the declines we have witnessed in used car prices over the past few years? For one thing, prices entered this period at an aggregate level that was well above trend. This occurred because new vehicle sales fell by half during the Great Recession — causing supply to contract — and only slowly recovered in the years immediately afterward. The paucity of 2008 model year vehicles is a market feature that is still in evidence today. 

New vehicle production and sales then increased at a steady clip in recent years, causing prices to fall back toward trend levels. Americans have also shown an increased propensity to keep older cars in operation for longer and this has also helped to boost available supply.

Trading used cars among ourselves is great for generating retail service fees for dealers but it has no bearing on vehicle supply. When analyzing possible implications for residual prices, market participants would be wise to ask whether a particular event adds to or subtracts from the size of the vehicle fleet. 

It is only these events that can influence vehicle supply, and therefore vehicle prices. 

Tony Hughes is a managing director at Moody’s Analytics, where he leads the development of used-car price forecasts.