Cars as Differentiated Durable Consumer Goods

Mike Smitka
Retired Economist
GERPISA Steering Committee
Automotive News PACE Judge

The investor site SeekingAlpha just posted my most recent article, Tesla's Thin Model Pipeline. That article is an application of the economics of differentiated durable goods to the auto industry.

First, autos are highly differentiated consumer good. Car company strategies reflect that, since Alfred Sloan organized General Motors to offer A car for every purpose..." (the ad is downloaded from the digital archives at The Henry Ford). More generally, use cases differ tremendously, from utilitarian transport to vehicles meant to display wealth. Work vehicles are even more differentiated, with up-fitters adding bodies to frame/drivetrain/cab sets for a multiplicity of businesses. We have small commuter cars, family-oriented SUVs, performance cars, and full-sized pickups with towing packages for farmers and others – I go past the Virginia Horse Center every time I head to town, lots of fifth-wheeler duallies.

Empirically, that suggests lots of models and a fragmented market. I use model-level data for Europe and for China (the world's largest vehicle market) to show just that. In Europe, no single model (or model series, e.g. the BMW 3-series) holds even 2% of the market, while over 400 models were available. In China, only 3 vehicles (barely) cleared 2%, while over 540 models were available. And these are just passenger vehicles. I don't have comparable data for the US, so I did a "deep dive" into the market-leading F-150 to show that it in fact consists of a family of models, suggesting that the upper limit for market share is likewise 2%.

The second piece is that vehicles are durable, with the average car on the road over 12 years old, and light trucks even older (from memory, 14 years). Used car sales are roughly 3x those of new. Hence new cars compete with used, and the more years a new car model is on the market, the greater the cannibalization from used cars. Henry Ford discovered that the hard way in the mid-1920s, as sales of the Model T fell despite repeatedly lowering the price. Ultimately Ford shut down production.

The economics literature on durable goods pricing is thin, and I know of only one set of studies that directly address that issue, from Adam Copeland of the NY Fed, solo and with various co-authors. Transaction data are hard to come by, the datasets compiled by various consulting companies are expensive and need to be cleaned up to handle rebates to consumers and dealers. That most new car purchases are accompanied by trade-ins further muddies the data. Their findings include an annual price drop of nearly 9%, and a shift down the income scale as a model ages. US consumers are well aware of these trends, and those on a tight budget time their purchases accordingly. Such consumers are also more likely to compare new and used cars. My son just bought a Subaru Legacy, and he shopped both. Given the current (August 2022) distortions in the market, he found that if he could wait, he could purchase a new car at MSRP, sticker price. If he wanted one immediately, he had to purchase used and pay above sticker. He could and did wait.

...to grow, Tesla has to develop a portfolio of new products and regularly renew existing ones...

Combine both of the above and the result is that car companies offer a portfolio of products that they renew on a regular basis. On SeekingAlpha I detail that as well, looking at the release cadence of new and refreshed models by the luxury car companies with which Tesla competes. As to Tesla, they have only one concrete future product, the Cybertruck set to launch in the summer of 2023. This year they've also refreshed the interior of the Model S. However, they've not updated the sheet metal on any of their existing vehicles – the Model 3 is overdue for that – nor have they announced any future product, only vague promises to come out with a limited-volume Roadster. Furthermore, I detail the limited market for pickups in China and Europe. The Cybertruck may do well in the US, but it's not a global vehicle. The stock market values Tesla as a high-growth company. However, they have revealed no strategy to develop a portfolio of products or to regularly renew existing products. My analysis indicates they can't grow without doing both.

Tesla appears likely to launch the Semi in 2023, but that is not a passenger vehicle and again is aimed at the US market.

China's NEV Market: Rising Segments, Falling Segments will generate winners and losers

Mike Smitka
Prof Emeritus of Economics
Automotive News PACE Awards Judge

I've been tracking Chinese sales data for a couple years, and pulled together thoughts in a brief article, China NEV Segment Analysis, on the finance website Seeking Alpha. The midsized segments, both sedans and SUVs, are stagnant. That's where Tesla's Model 3 and Model Y compete. In contrast, the compact "A" segments are expanding. Players there include one one startup, XPeng, but has BYD, VW and Geely as major players. To my surprise, I also find SOEs with decent shares, particularly GAC's Aion [GAC is owned by the Guangzhou Municipal government]. Most of the SOEs have lived quite well off of profits earned by their joint venture partners, such as Toyota and Honda for GAC. Are these proper commercial ventures, drawing upon the experience of SOEs in designing vehicles and running factories? Historically the SOEs were poor at design and at marketing. Some of that is the home boy effect – you won't find many Aion vehicles on the streets of Shanghai or Beijing. I don't know the produzct, I'm not a "car guy," so maybe these are real ventures properly run with earning money as a goal. But they could also be the result of party hacks pushing management to follow the EV trend, using their joint venture profits to make up for a lack of business acumen (and a very crowded market).

I also argue that the car market faces many headwinds. China has worked through the demographic dividend generated by falling birth rates, but now the working age population is shrinking, and probably the overall population as well. It's easy among the monthly NEV sales hype to overlook that the overall Chinese car market peaked in 2017. Then there's the end of the real estate bubble, evidenced by the failure of Evergrande, and the prospect of continuing lockdowns, an ironic side effect of China's initial success in using a combination of testing, contact tracing and quarantines to suppress the pandemic. There's no concrete left to pour after the huge infrastructure expansion that kept China from suffering the worst effects of the US real estate meltdown. That ammo has been depleted, and while Beijing talks about boosting car sales to offset the slowdown, consumers can't sell the condo that represents the biggest part of their life savings, and on top of that are worried that their bank may be the next one to shut its doors.

In any case, it will be interesting to see who does well in the NEV market, which just (barely) set a new sales record in July of 560,000 units. I think it's the bigger players who will fare well. VW's R&D spend is $3.6 billion a quarter; that's more than the combined revenue of XPeng and Nio, two of the current "pure play" EV favorites. Cars are in the end a consumer product, and having a broad, regularly refreshed portfolio of models on offer is of the essence. So my belief is that in the end the global OEMs will do well, and one or more of the private Chinese car companies: BYD, Geely and Great Wall. BYD is already in 6 of the 9 segments I tracked, as is Geely. VW, despite its late start, is already in 7. They cover most of the bases already. In contrast, Tesla can't seem to get new product out the door, and is stuck with but two models, and those in the stagnant mid-sized segment. They aren't heading towards failure, but they do risk being left in the dust as an also-ran.

Car Price Inflation: I Expect a Rapid Reversal

Mike Smitka
GERPISA 2022 Planning Committee
Automotive News PACE Judge
Prof Emeritus of Economics

Auto industry researchers, journalists and others, please consider attending the June 14-16 GERPISA conference in Ann Arbor, Michigan. A one-day price is available for locals who don't want to attend the full conference. A Tuesday dinner/reception and a Thursday Gala at the Ford Piquette Avenue Museum in Detroit. Breakfasts and lunches included. Morning events will be available to virtual participants; afternoon events are aimed at those attending in person. Go to GERPISA.org for details.

New car prices – and dealership profits are at record highs. Base models are unvailable, and high used car prices add to the mix. From another angle, the two line items are the biggest elements driving our 7% inflation rate. I believe, however, that "normalization" – greater availability, softer prices – will be sudden, muted only by the low level of cars flowing into the used vehicle market by car rental companies and other fleet operators. Why? – the market for high-end vehicles just isn't that large, so the shift from sticker-plus to discounting will occur over just a few months. Once it does, car makers will face a stronger incentive to produce vehicles with less expensive trim levels and in less expensive price segments. They won't have to discount the latter immediately, because such vehicles simply haven't been available. That won't last long, perhaps another 6 months.

To understand my logic, it's necessary to step back and look at the dynamics of the market since the onset of the pandemic. New and used markets are tightly linked, and the sources of "new" used cars (that is, sales by first owners) are part of the story. Yes, there's a chip shortage, and a war. However, that's but one factor, and you should be wary of such simple explanations. Analytics would be much simpler if we lived in a monocausal world, but we don't.

I'll put an arrow chart at the bottom, but I think a chronological exposition highlights the multiple feedback effects and channels of causation. So grab onto your seats, here we go!

1. The initial total shutdown froze both production and sales. Nothing happened – except in the used car market. Without business travel, Enterprise, Hertz and their peers were lift with fleets of cars that lost value by the day. We're not talking small numbers – Enterprise had about 2,000 vehicles just at the Detroit airport, and nearly 200 at their rural Rockbridge County, Virginia store, to serve a population of a mere 35,000. It didn't help that a couple of these companies were in poor financial help. What resulted was a fire sale, as rental companies "de-fleeted" as fast as they could, even if it meant booking a loss on the vehicle. Used car prices crashed.
2. However, that process ended once fleets had been slashed. In a normal year Enterprise buys 1 million new vehicles, the world's single largest purchaser. The flip side is that they also normally sell 1 million used vehicles directly to consumers and dealers, or dispose of them through used car auctions. Now none were flowing into the market. Consumers weren't driving much, remote work lessened commuting and attendant accidents, and with fewer new cars being sold, trade ins were also lower in number. Consumers who normally bought used cars held onto their cars longer, too, so that source also fell.
3. When demand returned, prices had nowhere to go but up. With used car prices rising at their fasted rate in history, demand shifted to new cars. Well, new car prices rose, initially because discounts shrank, and then vanished. Enter fleet purchasers. Normally they would opt for lower trim levels and include a higher mix of sedans than the overall market. But car companies weren't willing to discount those, or even to produce them. Instead they had to hold onto cars longer, and try to offset higher purchase prices on the SUVs they did buy with higher rental prices. So while there were more tradeins, the huge flow of one-year-old, moderate mileage rental vehicles of pre-pandemic times didn't return. Meanwhile used car purchasers found they couldn't afford to get rid of their old ride for one that was a bit new and more reliable.
4. You have surely noted that I've not mentioned supply chain problems. That's because I believe they are secondary to the interactions of new and used vehicle markets, and the flow of used vehicles. Lost production hurts, because when car companies are constrained in what they can produce, they move upmarket in trim levels and vehicle segments. For those old enough to remember, that happened in Spring 1981 when Japan agreed to Reagan's "Voluntary" Export Restraint that limited Honda, Toyota and their competitors to a combined 1.68 million units. I was about to buy my first new car, and watched the price on a subcompact rise 25% almost overnight. I didn't have to buy a more expensive trim package. Instead, I was forced to buy add-ons, a paint protection package that began peeling within a couple months, from a vendor no longer around. Grrr. But such is common in today's market. Clearly chip shortages add to the price swings, but even if we didn't have them, I am sure we would have seen a sharp spike.
5. So where is the market heading? Well, the dynamics I've sketched propped up the prices of luxury SUVs and fully-loaded vehicles. It won't take much to reverse that, because we all have run through stimulus money while wages aren't rising enough to push ordinary workers into the 15 million or so individuals with an income high enough to purchase a new vehicle. I've already heard from a couple reliable (but off-the-record) sources that big pickups and the largest class of luxury SUVs aren't selling. The rest won't be far behind.
6. Now the permanent shift will come once car rental agencies are able to re-fleet. Hertz coming out of bankruptcy is in my judgement showing desperation in striking deals to contract for large numbers Tesla EVs and now Polestar EVs. (I suspect the actual contracts are rife with escape clauses so that the numbers prove more PR than a purchase commitment.) Again, I am reading hints of re-fleeting, versus none a month ago. Within 9 months those cars will be cycled out, because from then on it will be easier for rental companies will able to buy more cars. Prices for used vehicles will drop rapidly, and across the board.
7. Note one implication: we will go from sharply rising new and used car prices to sharply falling ones in the space of 3-4 months. Gasoline prices are another major component for US inflation. Petroleum prices are set in global markets, and Russia is 11% of global production. The US is 20%, but is a modest net importer. So Russia accounts for a larger share of global exports. How much short-term flexibility do Saudi Arabia and the other Gulf states have to increase output? How leaky with Russian sanctions be? I don't know, but I don't expect further price increases. If so, then the two biggest sources of inflation will turn negative, and energy prices become neutral. I've already been surprised by how long inflation has lasted, but I remain convinced it's transitory.

I have opted to focus this post on the components of new and used vehicle markets, while shying away from data. OK, typically 2 or more used cars sell for every new car, and I can with some effort put specific numbers on that. It would make for a long and turgid post, or longer and less readable one. It's also very, very hard to get current data on fleet purchases and other components of a very dynamic market. I've also simplified, with no mention of interest rates or leasing. I don't believe my argument depends on such details.

From my presentation at the June 2021 GERPISA virtual conference.

Tesla China: Upside is Vanishing

Mike Smitka
Prof Emeritus of Economics
GERPISA.org 2022 conference team

The following appears on the SeekingAlpha finance blog. I wrote this last week, editorial approval took a long time due to COVID.

Were I to rewrite it, I'd probably start:

"...in 2021 Tesla exported an average of 15,000 Model 3's a month to Europe. At some point, though, Tesla Berlin will start production. Where will the Model 3's go? – they're in the "B" midsized sedan segment in China, which is the slowest growing EV segment in China. Furthermore, within the segment Tesla lost market share quarter by quarter in 2021. So the one place they won't go is to Chinese consumers. They want SUVs and fresher vehicles.

Tesla China: As Good As It Gets?

Jan 19, 2022 03:09 PM | Tesla, Inc.(TSLA) | By: Mike Smitka

• Tesla's CY2021 sales of 321,145 units are stellar, bolstered by December sales of 70,602 units. The good news is that the Model Y is holding onto its share in the "midsized SUV" segment, and that segment expanded 3.5x in 2021. CY2022 should be good.
• The bad news is that the Model 3 lost share in the "B" sedan segment throughout 2021, and that segment grew the slowest in 2021.
• With no new models until 2023 or later, CY2022 is as good as it gets. Read the full article now »