Brian Brown
Back in 2018, I wrote an article predicting that General Motors (NYSE:GM) and other automakers like Volkswagen (OTCPK:VWAPY) (OTCPK:VWAGY) (OTCPK:VLKAF) Ford (F) and Toyota Motors (TM) would see a substantial reduction in their unit sales within a few years because of the growing penetration of ride-hailing services like Uber (UBER) and Lyft (LYFT) with their sharing options.
Obviously, the COVID-19 pandemic completely derailed that. But in fact there has been a substantial reduction in auto sales in 2022 – but for reasons obviously having little to do with Uber and Lyft. Despite this auto stocks have held up reasonably well under the Fed’s selling pressure, as the lower sales and high demand have produced a more profitable environment.
The general expectation is that auto stocks will do well, either because sales stay low and maintain pricing power or because volumes recover. While the pandemic has obviously scrambled my projections around ride-hailing, it has also produced other changes that may be less auspicious for autos.
In my view, the changes in work-from-home policies in corporate America may be causing longer-term deleterious changes in the US auto market, which the temporary pricing power from the pandemic is only masking.
Miles, Sales And Inventories
My argument in 2018 was based on the view that any substantial reduction in VMT (Vehicle Miles Travelled) must produce a concurrent reduction in autos produced and sold every year. This is because assuming a steady state of automotive penetration, new auto sales every year are basically just sales to replace cars that have worn down and reached the end of their useful lifespan.
The key words there are “assuming a steady state.” The US is one of the most highly-penetrated automotive societies in the world, with a personal vehicle fleet of over 260 million cars and trucks. Because this inventory is so large relative to annual sales – around 18 million run rate when there isn’t a global chip shortage – changes in the structure of automotive utilization can have even larger impacts than changes in the rate of utilization.
Thus, the impact of any change that reduces the need for automotive transport, like working from home and less commuting, must be analyzed on two levels: the impact on VMT – and through it the necessary replacement rate – and the impact on automotive consumer inventories, the cars that sit in homes’ garages and driveways.
The Role Of Commuting
Commuting to work constitutes a very substantial portion of the entire personal automotive mobility segment. Prior to COVID, 30% of all VMT was accounted for by commuting to and from work, while only about 6% of the civilian workforce was working remotely, which means that the average worker who is required to travel to the office was spending roughly $4,160 commuting.
Those commutes were also taking almost half an hour each way, so they’re also working an hour unpaid. But that’s less relevant to GM and its peers, I suppose.
The situation now is very different. Roughly 25% of the workforce is now working full time from home, and another 20% are working part time from home. There are growing indications that these changes will stick as workers demand more flexibility from employers – and considering how expensive commuting is, that’s perhaps not as surprising as some think. This means that the VMT from commuting has probably dropped by about 1/3.
Thus, approximately 10% of all VMT in the country may now be in jeopardy. And that’s assuming that the work-from-home effect is finished growing, which it may not be. Could the final number be 15%? 20%?
Translating Fewer Miles To Fewer Cars
Replacement Rate
This is bad enough if it simply produces a 20% reduction in the annual depreciation of vehicles, i.e. it reduces the number of vehicles that needs replacing by 20% per year. That alone could take an 18-million unit annual run rate down to 14.5 million.
Inventory Adjustment
But far more serious is the possibility, if not probability, that it will also reduce the amount of inventory that is needed to fulfill all automotive needs. If a 20% reduction in VMT also reduced the amount of cars that Americans needed to own by 20%, there would suddenly be a tsunami of no-longer-needed used cars flooding into the market.
Of course, commuting isn’t heading anywhere near zero, even if it remains much lower than it has been. So a full 30% reduction is nowhere on the horizon. But the ratio of inventory to annual usage is so high that even a fraction of that is enough to produce a severe effect, if the effect on owned inventory is anywhere near the VMT reduction.
This comes back to the fact that America’s high penetration of automotive ownership produces an unusually high ratio of inventory to sales. Retrograde changes in inventory greatly magnify the effects of a lower VMT because excess inventory has to be bled off through reductions in new car sales over and above what the lower replacement rate has already produced.
If, for example, a cut in half of commuting did produce a 15% reduction in vehicle inventory needs, it would reduce the number of vehicles needed by roughly 39 million. That would hold auto sales at current depressed levels – but without the rocket high prices that make such low sales levels profitable – for almost nine years if phased in over that same time period. If it was phased in over five years, it would cut sales down to 10 million units per year. If it were phased in over two, it could theoretically take sales all the way down to 0 for that time frame.
Disproportionate Impact?
There’s also the possibility that a commuting reduction’s impact on automotive consumer inventory will be outsized relative to its share of VMT. Commuting is relatively unique among our automotive subsets in that it is more likely than any other use to produce the need for an increase in automotive inventory. Being on time to work is one of the few uses of cars for which a person or family has little to no flexibility. This increases the need for a car on hand they can use without fear of it being tied up by another family member.
A household with two working parents is routinely cited as the quintessential example of a two-car household. If one of those parents starts working from home and sells their car, their VMT will drop by 15% (half of 30%) but their car inventory will drop by 50%! When a single-person household owns multiple cars, one of them is almost always devoted to commuting.
Substantial Risk
To be clear, I don’t seriously think there is any chance of US auto sales falling to 0. Even a fall to single digits would be a little difficult. The most likely impact of a severe drop in demand would be to bring more buyers out at the lower prices. But such sales would not increase miles traveled, so they would merely prolong the adjustment, not obviate it.
Even putting aside any outsized inventory impact, work-from-home seems likely to, when considering inventory and replacement rate in combination, substantially reduce VMT in the US economy and, by extension, new car sales. Auto manufacturers are doing well now because, with the chip shortage in place, supply shortages are masking the effect and producing temporary pricing power above what the long-term equilibrium of the market would support.
Investment Summary
Things definitely did not turn out as I expected with regard to Lyft and Uber and shared rides. But in the five years I have been avoiding GM, it has returned barely 2% counting dividends, while the broader market is at 44% before dividends, so I still feel I made the right call.
My larger point has always been that personal automotive travel in the US is exceptionally inefficient, and that a lot of newer, innovative companies are trying to come to market with ways to reduce the waste. So long as that remains the case, companies which lie at the center of the traditional, inefficient model are likely to remain under pressure.
I rank GM a Hold only, and will not be opening a position.
Be the first to comment