Home » Stellantis Stands Tall With $18 Billion 2022 Profit And Big Bonuses For Workers

Stellantis Stands Tall With $18 Billion 2022 Profit And Big Bonuses For Workers

New Project

Good morning! It’s a great day to be a Stellantis worker or shareholder, a bad day to be a C8 Chevy Corvette, and (soon enough) a weird day when a Tesla equipped with “Full Self-Driving” version 11.3 hits the road. Let’s dive into today’s news—and let’s be careful out there.

Stellantis Makes The Big Money In 2022

03 New 500 France EV lease
Photo credit: Fiat

We know Ford binned it in 2022 while General Motors managed to print money instead. But what of Franco-Italo-Dutch-American automaker Stellantis, the combined Fiat Chrysler and Groupe PSA conglomerate whose name remains deeply embarrassing to say out loud?

Well, Stellantis kicked ass in ’22, according to annual results reported by Automotive News:

The automaker, which doesn’t disclose quarterly results under French financial reporting rules, on Wednesday said net profits for the year surged 26 percent to 16.8 billion euros ($17.9 billion) while global revenue rose 18 percent to 179.6 billion euros ($190.9 billion).

North American results paced the company’s 2022 gains. Stellantis said North American adjusted operating income rose 23 percent to 13.9 billion euros ($14.8 billion). Revenues surged 23 percent to 85.5 billion euros ($90.9 billion). Deliveries rose 2 percent to 1.86 million vehicles.

Since these gains were led by North American results, you’d expect strong Ram and Jeep sales to be the driving factor behind these numbers. But deliveries of those vehicles were off last year, thanks to supply chain issues. So what happened? According to various reports and statements by CEO Carlos Tavares, it was a few things: strong deliveries of EVs in Europe (which is funny, because Tavares has been a big EV skeptic compared to most), savings from the 2021 merger, and yes, new vehicle price increases. They say this year may not be so bad:

“Price increases were substantial in 2022 and they will be lower in 2023,” CFO Richard Palmer said on a call with reporters. “The challenge for 2023 is to offset inflation with pricing, but also with an improvement in the industrial efficiency.”

I tend to be skeptical when any company claims it’ll lower its prices, especially since so many supposedly inflation-driven increases have proven to be speculative. But there’s an upside for workers and investors here: Stellantis is offering a stock buyback of up to $1.6 billion, and a $14,760 bonus to each of its 40,500 employees affiliated with the United Auto Workers union.

That’s a hefty chunk of change! Think how many busted old Jeeps you could buy with $14,760. At least two, at current car prices.

Either way, this big merger between some of the world’s most notable auto brands seems to be paying off nicely.

What To Expect When You’re Expecting Mayhem Screen Shot 2023 02 22 At 9.10.02 Am

Elon Musk has a unique penchant for getting himself and his companies into hot water, but few Tesla-related things are as controversial as “Full Self-Driving” Autopilot—its supposedly advanced automated driving assistance system. I put that in quotes here on purpose because it’s nothing of the sort, and in case you need a reminder, there are no self-driving cars on sale today. And FSD has faced a number of federal and state crackdowns over its operational abilities as well as how it’s advertised.

But hey, Musk says Tesla’s potentially worthless until it can figure out the robot car thing, so the company powers forward. Version 11.3 of FSD is coming soon and already in the hands of employees and test drivers, reports Tech Crunch. The outlet has the scoop based on leaked documents posted to Reddit:

FSD beta V11.3 promises to allow a Tesla to drive itself completely from A to B through cities and highways by combining FSD and Autopilot capabilities into a single-stack system. Today, drivers can engage FSD on city and residential streets to handle things like responding to stop signs and traffic signals. Highway driving is available via Autopilot’s software, which can automate driving tasks like traffic-aware cruise control and automatic steering in clearly marked lanes.

The latest version enables FSD beta on highways, which “unifies the vision and planning stack on and off-highway and replaces the legacy highway stack, which is over four years old,” according to release notes shared on Reddit. The notes continue:

The legacy highway stack still relies on several single-camera and single-frame networks, and was setup to handle simple lane-specific maneuvers. FSD Beta’s multi-camera video networks and next-gen planner, that allows for more complex agent interactions with less reliance on lanes, make way for adding more intelligent behaviors, smoother control and better decision-making.

Overall, the new software “provides a more seamless overall experience, but it’s not entirely free from issues that have led to disengagements,” the story says. Great.

Interestingly, V11.3 offers the ability to send a voice note to Tesla when a disengagement happens, so you can personally describe exactly what happened when your Model 3 decided to park itself inside an Arby’s again.

Anyway, expect this software to hit the road soon. What’s that? You, a non-Tesla driver, didn’t sign up for this public beta test? Tough shit, amigo. Disruption happens whether you, or the children inside your car, asked for it to happen or not.

Tesla: Friendship Ended With Germany, Now BIDEN Is My Best Friend

0x0 Modely 04
Photo credit: Courtesy of Tesla, Inc.

Say what you want about President Biden’s Inflation Reduction Act—the tax incentives on individual vehicles are deeply confusing and inconsistent at times—but it’s working. The goal of the IRA is to localize EV and battery production in the U.S. so our auto industry isn’t totally beholden to other countries, China in particular. One example is the big Ford battery and EV plant announced for Michigan last week.

Now, even Tesla indicates it may focus on battery cell production in America, not its new Berlin site, according to Reuters. It sounds like Tesla will certainly do battery production in Germany, but that could get scaled back and the U.S. could get more attention and output thanks to Biden’s tax incentives:

Tesla has begun assembling battery systems at its plant in Germany but will focus cell production in the U.S. in light of tax incentives under the Inflation Reduction Act (IRA), a spokesperson said on Wednesday.

The EV maker is also preparing to produce cell components such as electrodes, some of which will be sent to the U.S. from its site in Gruenheide in the state of Brandenburg, the spokesperson said.

Cars produced at the Brandenburg site, just outside Berlin, would in the “near future” contain batteries assembled locally, the spokesperson added.

“The focus of Tesla’s cell production is currently in the United States due to the framework created by the United States Inflation Reduction Act (IRA),” the spokesperson said in a statement.

EU leaders have expressed concern that local content requirements of much of the IRA’s $369 billion of subsidies will encourage companies to abandon Europe for the United States.

Understandably, the German government is looking into what’s happening here exactly, because you probably can’t count on Tesla or Musk to tell anyone anything. Maybe you’ll get a tweet if you’re lucky. We reached out to Tesla’s communications team for comment, but we haven’t heard back yet because they were all fired in 2020.

Either way, on the EV front, Dark Brandon is crushing it.

Corvette Production Paused Over Parts Shortage

2023 Chevrolet Corvette Z06
Photo credit: Chevrolet

The C8 Corvette hasn’t had the easiest time in production and it’s not even the car’s fault. No, it keeps getting hammered by chip and parts shortages, as it has since the pandemic started. (It also makes sense that GM might want to focus its more available resources on the volume-sellers rather than niche sports cars.) Now, once again, production in Bowling Green will be paused for a bit, Automotive News reports.

It should resume next week and this time it’s apparently not chip-related, but it’s another bummer for anyone who may have been on a Corvette waiting list.

Your Turn

I’ve driven many Teslas and I hardly ever use Autopilot unless I’m on a long road trip, and then I treat it like glorified cruise control. Why? Because I don’t trust the damn thing, and besides, I enjoy driving Tesla’s cars a lot. You know, myself. But it’s even worse if you’re just minding your own business in your own car and you have to deal with faulty Autopilot issues or some joker taking a nap in their Model Y.

So let’s talk regulation and rules for this sort of thing. What laws do you think should exist for these so-called “self-driving” systems, or even for the automated driving assistance systems? How can we allow this technology to develop, which it inevitably will, while keeping other motorists, pedestrians and cyclists safe?


Share on facebook
Share on whatsapp
Share on twitter
Share on linkedin
Share on reddit

44 Responses

  1. There are several ways to regulate self drive systems.The best would be to make sure government safety agencies are well funded and doing their f**king jobs.
    They need to be testing cars BEFORE they’re allowed on the roads.
    Any issues revealed in testing must be properly fixed and retested.
    Any egregious errors get an automatic 6 month ban.Bring it back after 6 months and we’ll test again.
    Stuff like not being able to see a human in the dark or not seeing non-moving objects on the road are insane.Allowing systems that bad on our roads is unacceptable

    1. Do you have ANY idea how much money they mint with Jeep?
      What is the top selling mid-size SUV, period? Jeep Grand Cherokee. For what year? Fucking all of them. No, not ‘some’ of them. Not ‘most’ of them. Not even ‘almost all of them.’
      Jeep has been the top selling mid-size SUV since before there was a ‘mid-size SUV’ class. The SJ Wagoneers were moving over 15,000 units a year until 1990, and those things were expensive as hell. And that was with absolutely no competition.

      The first year of the Grand Cherokee was ’92, right behind the death of the SJ, and they sold 129k units. That year. Alone. In 2021 they sold 264,444. With an average price of let’s be generous and say $55k, that’s $14,544,420,000 in revenue. From one model. And Jeeps have always been high margin vehicles.
      The only thing that gets close to the Grand Cherokee is the Highlander. Which would take the wind out of people’s sails, if the #4 best selling mid-size SUV wasn’t the Wrangler at 204,610 units. (Wrangler margins are insane thanks to still being a pre-1980 platform.)

      Chrysler always said they bought AMC solely for Jeep. That’s why. Even in the SJ days, it was a money printing operation. They did not wait for the XJ to be a smash hit. They tried to buy Jeep without the rest of AMC. Where Kaiser couldn’t make a penny, AMC never lost money on Jeep. Ever. Jeep was so strong that it not only covered every other model’s losses, but turned a profit.
      A 1978 $65M loss on ‘conventional’ swung to a $36.7M profit. Just from Jeep alone. They forecast 40,000 combined for the SJ and XJ in ’84 – they instead did double estimates, for both.

      Grand Cherokee and Wrangler by themselves are over $23B in revenue. Combine that with profiteering and already insane margins, and you could sell every sedan and minivan at a loss and still make a profit.

      1. This always makes me think: Why dont they cut Jeep loose?We’ve heard for years that consolidation is the only way forward but Jeep is proving it can survive just fine alone.
        I’m sure shareholders would approve.Get rid of the other lossy brands and watch the share price skyrocket

        Of course this requires the management actually admit it has lots of bad cars…

    2. It’s been the curse everything-except-Jeep that made up AMC, for what, 60+ years now? From Studebaker to Stellantis (I know, I know, that meger didn’t go through, but the lure of alliteration was too much) , making money making cars is hard

  2. “I tend to be skeptical when any company claims it’ll lower its prices”
    He didn’t say he’d be lowering prices. He said the increase would be lower, which just means he’ll be raising prices by a smaller amount in 2023 than they did in 2022.
    The second part of that quote is the key. “Offset inflation with pricing, but also with an improvement in the industrial efficiency” is CEO-speak for raise prices and cut costs.

  3. 1. Mandate the names of the systems include what level of “autonomy” they are; ex “Level 2 Super Cruise” with the explanation chart included on all Monroney Stickers.
    2. Make the companies legally and financially responsible for Any and All incidents that happen while systems are engaged or just prior to systems disengaging.
    3. Require 10,000 hours of direct supervised testing (including closed loop, and real world) done by the manufactures with NHTSA reps (not customers) before systems can be engaged by customers.
    4. Require 5,000 hours of direct supervised testing done by the manufactures with NHTSA reps (not customers) before major system updates can be sent out to customers.
    5. Require 1,000 hours of direct supervised testing done by the manufactures with NHTSA reps (not customers) before minor system updates can be sent out to customers.
    6. Require NHTSA, CPSC, and FCC approval before making claims about future software promises and deliver refunds when promises are not met.

  4. I think the Stellantis “lower in 2023” remark referred to the magnitude of the price increase. It was a large increase in 2022, but it’ll be a smaller increase in 2023. I, too, do not believe prices are going to go down for anything new anytime soon.

  5. I think GM (maybe others?) has the right strategy in allowing its Super Cruise only on a certain marked set of roads, mostly highways. That’s both where self-driving is easiest for the machine, and most useful for the driver, compared to busy city streets.

    Let it develop there and learn, gradually introducing it to more and more of the map as the capabilities improve.

    I also think there is going to need to be at least one giant precedent-setting case to mete out responsibility between driver and manufacturer when this kind of thing goes wrong. Right now I think it’s in some kind of limbo and no one really knows.

    1. Someone is going to have to tell me which words in the above comment set off the moderation filter, because usually I can guess, but this time I am clueless.

    2. THIS!
      yes, if manufacturer bears no liability, then it’s an easy decision for them to let half arsed products into the wild.
      currently they can point to the owners manual and joeblow gets stuck with the bill.
      make the manufacturers have some skin in the game.
      how about a regulatory presumption that manufacturer is e.g. 50% at fault in accidents involving level2 systems?
      that would permit market forces (insurance companies and lawsuits) to drive behavior.

  6. There should be some type of certification when you buy a car with these kind of systems, the limitations, etc. I don’t use it much on my car since it makes my drive more stressful when the car does something that I wasn’t expecting (always trying to take the exit on the highway when I am driving on the right lane for example). The steering wheel gets hard as a rock when you try to manipulate the car when the system is on, and sometimes the car start telling me to hold the steering wheel when I am actually holding it. I only use adaptive cruise control since I can focus on the steering wheel and the car can handle speeding up, braking, keeping a safe distance with the vehicle in front of me in case something happens.

  7. Car safety systems should augment the driver and not go much farther, not fully automated on public roads with 40 year old no-tech cars and Semis running about. It’s not just accounting for all the regular road things like staying in the lane, but then accounting for random guy in his 1992 F350 just remembered this is his exit and has to cut over 3 lanes seems insurmountable.

    If people want a full self driving automated vehicle, that’s where Trolleys/Subways/Trains fill in.

    Also hopefully Stellantis(ask your doctor if Stellantis is right for you) should put some of that money back into updating their platforms a little, Dodge and Chrysler’s lineups are practically gone.

    1. Agree. I’m not sure why we’re developing anything much beyond adaptive cruise control. Trying to get computers to understand and overcome the randomness of real life on the roadway seems like a lost cause and I’m not sure what problem it is trying to solve anyway.

    2. If they finally updated the Charger/300 platform that’s from the Cambrian period and put the Hurricane 6 in it I’d buy one in a second. The wife and I are working on expanding the franchise and I have some concerns that the Kona N is going to prove to be too small in the next few years. If that’s the case a big, fast sedan would be a no brainer.

  8. So I understand that folks appear to be concerned with “self driving” cars on the roads, but the logic of “I don’t want this on the road with me” is very lacking. I don’t want drunk drivers. I don’t want sleepy drivers on the road with me. I think drivers looking at their cell phones are way more dangerous than “autopilot”. I don’t want them. BUT. We have seen people fall asleep with autopilot engaged and *nobody got hurt*. So if I’m going to freak about something on the road with me, I’ll freak out about cell users because they’re more likely to hurt someone. I know it gets the clicks, but really?

    1. Yeah, really.
      It isn’t about self driving cars on the road, its about untested, beta versions of self driving cars on the road.

      We legitimately don’t know what these cars will do. However we are at a point where we can control how and when we let them on the road. Your argument is basically the same as saying “who cares if phones are in cars” in 1995.

      1. Not really, because you appear to think that it’s less safe than the status quo. Have more situations been averted than created by this technology? It appears that answer may be yes. Where’s the data? Because for every “parked in an Arby’s” we appear to have a “was stopped in traffic while asleep/drunk, no issues.” That should be the discussion.

        1. No.

          The situations averted by self-driving are clearly human error, with humans to blame.

          The situations caused by self-driving are imposed on humans who made no error beyond trusting something called “self-driving”, and that is the moral difference that is unacceptable.

          Until we solve for the moral difference, “self-driving” does not belong on roads not specifically designed for it.

    2. One of the issues is that the presence of this “beta” version of self driving is not as capable as advertised, and is actively training more people to be less attentive. The utility for drunks and the like is not debated. It’s the large collection of people on the fringe of being responsible drivers. Without this unreliable safety net, many of these middling folks would be “good enough” drivers. But with it, they feel ok taking a nap and then running into a parked fire truck. Or getting easily distracted in an adverse condition where AP is not effective, because they spend most of their time behind the wheel doing anything but driving.

      1. Good point. However, it appears the issue is people, not the software. Someone made a really interesting suggestion in a similar thread a while ago: add autodrive status lights to all cars running such software. The only way we’re going to get this stuff to work is get it out on the road (because Machine Learning needs massive data sets) — so why don’t we add blinky lights to indicate a system is active? Jason can obsess on where we add them, other drivers are warned, and we get a useful marker for video take from other systems. Thoughts?

        1. I would be strongly in favor of blinky lights on self-driving vehicles – and would love to see Jason’s mock-up of how they would be mounted appropriately.

          However, we also need to make sure these cars can easily be identified in bright daylight, when lights may not be so easily visible. Perhaps some sort of distinctive roof furniture is required?

    3. Being drunk or asleep at the wheel is illegal. The fact that people still do it shouldn’t then lead one to think that beta testing unproven software on the roads is somehow okay. That’s like saying that since we have one problem, it’s fine to have two problems.

    4. Your logic is a bit lacking… No one wants drunk drivers on the road, that’s why it is illegal. No body wants sleep drivers on the road, that’s why falling asleep while you’re driving is illegal. And guess what… using your cell phone while driving is illegal too (in most states?)!

      So by your logic, “self driving” cars should be just as illegal as driving drunk! Which in the case of untested “beta” applications, I completely agree.

  9. I have no idea how you regulate something based on machine learning. By their very nature they have to deal with an enormous number of possible inputs, too many to actually test. If you have a fixed set of inputs you probably aren’t bothering with ML.

    What I’ve seen of such systems thus far tells me that they will look amazing at first glance, right up until they shit the bed and fail in some spectacularly dumb way. How do you validate a system based on petabytes of historical data (that might be constantly changing, depending on the system) and a whole bunch of fuzzy logic to pick the most probable response out of that data? I haven’t the faintest idea, and AFAICT neither do the people working on this stuff.

    I remain of the opinion that these systems are useful as a first step in accomplishing a task, but they still need a human validating all of their answers to catch the obvious mistakes. As Torch has repeatedly noted, that’s a terrible fit for driving because humans are bad at constant vigilance activities.

  10. Regarding Stellantis’ profits… So out of on $17.9 billion in profit, $14.8 billion came from North America.

    I have to again wonder out loud why FCA needed to merge with anyone. And I also have to wonder that given the vastly greater profitability of was is clearly the old FCA part of Stellantis, why wasn’t it FCA buying PSA?

    Well I can tell you… the problem was top management at FCA and a big shareholder named ‘Exor’.

    Chrysler and FCA were fine on their own and had the money and resources to do anything they needed to do.

    The root issue to this mess was Bob Eaton… the guy who took over as CEO of Chysler after Lee Iacocca and sold out to Daimler to get his golden parachute.

    Iacocca should have picked Bob Lutz instead.

    Regarding Tesla and Germany… I wouldn’t say it’s ‘friendship ended’. Tesla is just doing what all for-profit companies do and put their investments where they get the best bang for the buck. Now having said that, Tesla isn’t gonna leave Germany. They’re gonna continue to assemble vehicles for EU consumption for EU tariff reasons alone.

  11. I maintain that the tech bros obsession with autonomous driving can mostly be explained by San Francisco Bay Area traffic and the fact that even the billionaire class isn’t immune to it.

    Of course the obvious solution would be “more and better public transportation” (at this point the scary price tags aren’t even so much when compared against freeway maintenance and potential expansion projects). But that would involve folks like Elon sharing space with The Poors and not getting to skim sweet rents from the top so clearly no.

  12. I’m going to take the “Your Turn” question slightly differently.

    I believe that to make AVs happen, there is of course going to need to be rules about how manufacturers implement, improve, and update these systems. However, they can’t do it on their own! There must be improvements to infrastructure, not only to fix what’s crumbling already, also to help with systems like those used by AVs.

    There are going to need to be standards agreed upon for V2I and V2V communications so that any AV driving out there has an idea of what is going on and what to expect. Manufacturers are going to need to take care of their part of V2V, and the government is going to have to step up to upgrade and install V2I compatible infrastructure.

    Of course, this is going to require money be invested by all parties, and agreement on how to standardize the tech as well as write the laws for it, so I’m not exactly hopeful of the above being implemented in a reasonable amount of time

      1. That’s a hell of a question, actually! I don’t think any of us on this site want AVs to take over completely, so we’re never going to advocate any of that.

        However, AVs are coming, and this is the sort of stuff that needs to happen for it to be safe.

  13. You know, for my entire life there have been people traveling around in cars who, except for occasional verbal commands, didn’t have to control speed or braking and did not actively steer the car or maintain position in traffic lanes. We called them PASSENGERS.

  14. If people want to penalize the oil companies for making too much profit, then shouldn’t the car companies who produce the users of the oil be also penalized for too much profit?

  15. if the few Stellantis employees I know represent the majority (and I’m almost certain they do), metro Detroit cocaine purveyors have to be hyped on the huge bonuses

  16. Now all Stellantis has to do is bring some of their cool French cars here.

    Dodge: Fiat
    Chrysler: Peugeot/Citroen

    Dodge Panda
    Dodge Berlingo
    Dodge Ducato
    Ram 700
    Ram 1200

Leave a Reply