A leading automotive forecasting firm now thinks that U.S. auto sales will fall at least 23% in 2020 -- and that's the best-case scenario. 

ALG is a subsidiary of online car-shopping service TrueCar (TRUE -0.71%) that provides forecasts of auto sales and used-car prices to help banks set the prices of new-car leases. The company's analysts keep a close eye on auto sales trends, and they think the effects of the COVID-19 pandemic are shaping up to be worse than they expected just a few weeks ago.

Three scenarios for U.S. auto sales in 2020 -- none of them good

As it did in its last update in March, ALG sees three possibilities for U.S. auto sales in 2020. But its new forecast is weighted more toward the negative outcomes than in March, and even the optimistic case is a lot less optimistic.  

A row of cars on an auto dealer's lot.

Image source: Getty Images.

The optimistic case: 10% chance

If government restrictions on movement are lifted nationwide in May, and if the U.S. government's stimulus and relief packages have their intended effects, then U.S. auto sales could total 13.1 million this year. In March, ALG's optimistic case called for sales of 15.3 million vehicles.

That would represent a decline of about 23% from 2019, when Americans bought just over 17 million light vehicles. "Light vehicles" is the industry term for the kinds of vehicles bought by individuals: cars, pickup trucks, and SUVs, but not big trucks. 

That would be bad, but not awful. But right now, ALG thinks there's just a 10% chance things will work out that way. 

The mixed case: 45% chance

In this scenario, movement restrictions get lifted in May, but the economy doesn't recover quickly, and local outbreaks of the virus cause some disruptions later in the year. Total U.S. light-vehicle sales in this case would be about 12.6 million, down 26% from 2019. In March, the mixed-case forecast was for 13.2 million vehicles sold.

ALG gives this scenario a 45% chance of happening.

The cautious case: 45% chance

Some experts think the virus could have a second peak in the fall. If so, and if that leads to new movement restrictions, then ALG sees a slower recovery through the end of 2020, with higher unemployment levels lasting longer. 

The upshot: 11.3 million light-vehicles sold, down 34% from 2019. But it's up slightly from March's cautious-case forecast of 11.2 million vehicles sold.

ALG now thinks there's a 45% chance that this could happen.

Why so pessimistic? 

Eric Lyman, ALG's chief analyst, said that with movement restrictions now extended into May and unemployment skyrocketing, the industry's faint hope of a full recovery to 2019 sales levels isn't happening. 

Lyman said he thinks 12.6 million sales -- the mixed case -- is the most likely scenario. But that scenario depends on two things, he said: First, the stimulus packages successfully get most people through the spring; and second, people get back to work quickly in the summer, once movement restrictions are lifted.

How about after 2020? Morgan Hansen, ALG's chief data scientist, said that the underlying "replacement rate," or the number of vehicles people will seek to replace in an average year, is around 15.5 million vehicles. He believes that if sales in 2020 end up below that number, as now seems likely, many of those will be made up over the next couple of years. 

What does this mean for auto investors?

For auto investors holding shares of companies like Ford Motor (F -0.84%) or General Motors (GM -0.10%), the implication is clear: Whether or not the larger U.S. economy is in recession, auto sales are functionally in a recession right now. 

That's not good for auto stocks or automakers' profits. But the silver lining is that historically, auto stocks, and automakers' profits, have tended to bounce early in economic recoveries, as buyers return to dealer showrooms. 

Assuming the U.S. and Europe recover from the coronavirus downturn without a prolonged recession, as in ALG's optimistic and mixed cases, there's a chance that bounce will happen this time around, too. But as ALG makes clear, there are a lot of variables still in play. Proceed with caution.