Tesla’s Investors Day has generated lots of expectations and speculation. Most are not interesting enough to post about, Farzad Mesbahi is the exception. Of Musk’s many uttering maybe the most relevant is that Tesla’s most important product is “The machine that makes the machine.” Giga factories that obsolete the old ways of making cars. I’ll let Farzad tell the rest of the story.
Tesla’s $25k Car Will Break the Auto Market (and the Internet)
The Captain
Does it?
Tesla has high margins. Much higher than most automakers. But comparatively high margins can come about from either being more efficient than your competitors in making your product or being able to charge more than your competitors for your product. Or some combination of the two factors, obviously.
Is there any analysis that Tesla is significantly more efficient at producing its cars than other automakers? That the Gigafactories do indeed represent some real advancement in how cars get made? It wouldn’t be surprising if Tesla’s factories were more efficient than the average automaker. We would expect that any collection of relatively brand new factories would be more efficient (other things being equal) than the average car factory, merely by being newer. After all, we would expect that the 10% newest auto factories from conventional automakers would be more efficient than auto factories overall.
But apart from that, is there anything that shows that Tesla makes its cars more cheaply than competitors rather than, say, being able to leverage its brand, software packages, and first/early mover position into the field into higher prices for vehicles with relatively conventional production costs?
'Cause if we’re unpacking general statements about the future of Tesla (as you do upthread), there’s Musk’s comment about self-driving that has to be accounted for as well:
" “But the overwhelming focus is on solving full self-driving,” Musk said in an with the YouTube channel “Tesla Owners Silicon Valley,” published Tuesday.
“That’s essential. It’s really the difference between Tesla being worth a lot of money or worth basically zero,” he said."
That’s really inconsistent with the notion that Tesla has some material edge due to manufacturing prowess or the Gigafactories. If Tesla’s really got such a huge advantage in manufacturing efficiency, the value of the company wouldn’t be entirely contingent on software advances. It is, however, consistent with the idea that Tesla is kind of in the same league as other automakers in terms of manufacturing efficiency, and instead is able to charge high premiums based on being early to the market (which advantage will erode over time) and software.
That’s the danger in relying on the gibberish printed by Business Insider. It makes your conclusions just as stupid as their disingenuous assertions. You do know the company is run by Henry Blodget, right? Wall St. criminal, permanent ban from the securities industry (Henry Blodget - Wikipedia). You should just assume that everything he prints is a continuation of his fraudulent practices by other means.
Musk was (when taken in context) saying that everybody would have autonomy at some point, so that Tesla had to have it too or be worthless as a car company. It was essentially meaningless hyperbole.
From a couple of years back, regarding production time per car:
In comparison, Diess says that VW is at over 30 hours at Zwickau, and it is currently aiming to bring this down to 20 hours, which would still be twice as much time as Tesla.
This is a huge advantage for Tesla, and has just gotten more extreme since then.
-IGU-
@Albaby -
Please tell me that you’re just trying to play devil’s advocate with this argument!
Even I, as a way left field Tesla follower can believe that Tesla can be way ahead of their legacy competition in terms of productivity.
I think of my career consulting experience as a “change agent” to design and implement new systems and processes across various legacy industries. I would say that the most challenging aspects of completing those initiatives was meeting resistance to change - time and time again.
Then there were those clients where we needed to design a system and a process from scratch, and we were able to wring most inefficiencies out of the design (6-Sigma).
I’d say without using too much brainpower that Tesla has a significant advantage in being able to have no legacy BS dragging the operation and the design (and the entire process to get there) down into the weeds.
My 2 cents!
'38Packard
Does Tesla have an advantage in production costs? Let’s run some numbers.
| Tesla | Ford | GM | Toyota | BMW | |
| Sales (000,000) | 71,421 | 149,000 | 143,914 | 241,380 | 112,662 |
| CGS (000,000) | 51,108 | 134,397 | 126,892 | 186,544 | 92,792 |
| Units (000) | 1,314 | 4,231 | 3,579 | 8,230 | 2,522 |
| Price/unit | 54,354 | 35,216 | 40,211 | 29,329 | 44,672 |
| cost/ unit | 38,895 | 31,765 | 35,455 | 22,666 | 36,793 |
(Not the prettiest table I’ve ever done, but it will have to do.)
Sales and cost of sales are in millions. I’ve converted Yen and Euros to dollars with an eyeballed average exchange rate. (130 for Yen, 1.18 for euros)
Data is from the most recently reported fiscal years. Specifically, Ford, Tesla, and GM are calendar year 2022. Toyota is their FYE 31 March 2022. BMW is calendar 2021 (their 2022 numbers aren’t out yet). Since Tesla is more of a luxury car, I really wanted to get a luxury maker in the comparison. MB reporting was a mess, so I used BMW. These companies all have non-automotive groups. Fortunately, they all report automotive results separately from other divisions. So these figures are for the respective automotive divisions. Finance arms are excluded.
So what did I notice? Well, on a strict dollar cost per car sold, Tesla is actually the highest. Kind of puts a damper on that low-cost producer idea. But that’s not the whole story. We really need to adjust for the price level of the cars.
As I mentioned, Tesla is more of a luxury car. So the best comparison in this brief list is probably to BMW. Tesla spends a couple thousand more per car to produce them. But wait - BMW’s numbers are a year old. And inflation during 2022 was anything but negligible. If we increase BMW’s figures by 5% to account for a year of inflation, their costs are basically the same as Tesla - just shy of 39k per car.
My take - Tesla doesn’t really have a cost advantage over BMW, in spite of the newer factory technologies. And in spite of the huge difference in the number of parts in their respective drive trains.
Let’s also look at sales price. Tesla is clearly at the top of this sample of car makers. And if we give BMW the same 5% inflation adjustment to their sales, that puts their average car price at about 47k per car. Well below Tesla’s 54k.
Here, it looks like Tesla has some advantage in the market - able to command a higher price per vehicle than BMW.
So if I were to opine on the source of Tesla’s large advantage in operating margins, I’d attribute it mainly to their pricing power rather than lower costs. That also makes sense with what we see in the market place. To buy a Tesla, you place your order and wait for Tesla to build the car. To buy a BMW, you can walk onto a dealer’s lot and buy something right away. Tesla can sell all the cars they can build. BMW has an inventory of cars available for sale at their dealer network.
Which also points out another notable difference. BMW sells cars at wholesale to their dealer network. Those dealers add some markup to the price of the cars. Likely not a lot, as we’ve discussed many times that dealers make most of their money on service, not on new car sales. But there is some markup to cover the cost to the dealer of selling cars - commissions, flooring, showrooms and storage lots, for a few examples. Tesla’s sales are at retail. So they are capturing that markup which would otherwise go to a dealer. But those selling costs should not be in the cost of the cars, they would be reported in S,G,&A costs. That would be an interesting further analysis for another time. Still, though, I think Tesla has some significant pricing power in the current market. Perhaps not quite as much as this simple analysis suggests, but probably still material.
At any rate, there’s some numbers-based analysis for you all to chew on.
–Peter
As a fellow ex (management) consultant I agree with this statement 100%. One approach to overcoming this negative bias was to let people think that these were really their ideas. Coming is as the expert from afar was the worst thing one could do. One way to create team work was to share, I would tell my customers, “You know your business, I know information technology, let’s work together.” At Colgate Palmolive the Comptroller asked me for a sales report sorted in the traditional fashion, by region, sales rep, etc. I said that I would prepare a report sorted by revenue. He didn’t like the idea so I said to him “Sr. Ellis, if you don’t like my report I’ll make another like you are asking for.” He never did. My report set in motion a series of changes that cut our delivery time in half, from 48 hours to 24 to match our competitor, Procter and Gamble despite the fact that their factory was in Caracas while ours was in Valencia, over 100 miles distant and over 2 hour drive each way. BTW, we also closed the Caracas warehouse as it was no longer needed. I didn’t tell Mr. Ellis the should smarten up and catch up with the times. 
The Captain’s
contribution was just a sales report, they figured out the rest of it 
Hey Captain -
Thanks for your response.
Interesting that I also worked for a division of Colgate-Palmolive from 1985 - 1987. I was in and out of their HQ in NYC a number of times as we implemented a new mainframe order processing system for the division - Princess House, Inc.
I also worked for P&G from 2005 - 2008 (through the Gillette acquisition) and spent lots of time in Cincinnati and Warsaw, PL leading a global team implementing an enterprise data warehouse and analytics platform.
It was great to work for some amazing Fortune 500 companies (including Gillette, JNJ, National Grid, TJX, Biogen, 3M) throughout my career.
Cheers!
'38Packard
→ happily retired now!
IGU, I know you would never miss an opportunity to savage the business press you disdain - but it should be obvious in context I wasn’t relying on Business Insider for any assertions. I was just providing a citation for the quote. The BI article was just the first one that came up in my google search - there’s scores of other citations, if you prefer another one.
As for this:
…that doesn’t make any sense as an explanation of his quote. Everyone will have autonomy at some point - that doesn’t mean that everyone will “solve” autonomy at some point. Most automakers won’t “solve” autonomy. It won’t make them worthless. They’ll just end up licensing autonomy tech from the firms that actually develop it. Not as lucrative as being the firm that develops the tech, but not something that makes your company worthless.
If Tesla really does have an especially unique and valuable innovation in making factories - one that could support the “machine that makes the machine” quote - then there’s no way not “solving” autonomy could render the company valueless.
Of course, the “machine that makes the machine” quote could also be “meaningless hyperbole,” as you put it. And I suspect that’s the reality.
They can be ahead of their legacy competition based on this dynamic - but that doesn’t mean they are.
Yes, legacy BS can hold back from implementing new ideas. So Tesla certainly has the freedom to experiment and implement new manufacturing ideas. That doesn’t mean, though, that they’ve actually found any new manufacturing principles that are so revolutionary or innovative as to support the “the machine that makes the machine” type rhetoric that Captain quoted above. Or even the type of analysis in the video he quoted. Indeed, a few of the ideas that Tesla was able to implement because they weren’t held back by legacy “BS” ended up having to be walked back (hi, Fluffbot!). Not every departure from past practice is going to be a winner - and departing from past practice doesn’t mean your new ideas are materially better than the old ones.
Again, Tesla’s got fantastic margins - but margins don’t tell you whether Tesla’s doing better at making cars more efficiently or selling them at higher prices, relative to competitors. Whether they actually have some manufacturing special sauce, as opposed to pricing advantage or better software packages or a more valuable brand.
Tesla certainly produces them fast, as IGU noted above - and we know they use somewhat different processes than other companies. But as ptheland’s back-of-the-envelop analysis above points out, building them faster doesn’t necessarily mean they’re being built more efficiently in terms of lower production costs per unit. Much less by such a large amount to support the idea that the factories themselves are Tesla’s special product - the “machines that build the machines” idea.
Building cars one has two sources of expense - the cost of what goes into the car and the cost of the factory to produce it. If one is producing cars twice as fast, isn’t the factory cost portion much less?
This is the proverbial apples/orange comparison. I think we can all agree that ICE production costs are generally lower than BEVs of equivalent level. That’s why few BEVs are currently profitable. The comparison should be between the BEV costs of Tesla compared to the BEVs of legacy companies.
One example is the Ford Mach-E, considered a direct competitor of the Model Y and priced roughly the same. The Mach-E is not profitable while the Model Y has industry leading margins.
No, we can’t! I might be true for legacy car makers using old style production methods. That’s where Tesla’s advantage lies, they found out at Fremont how poor legacy manufacturing plants were and they changed course. Tesla build purpose designed factories to implement the Agile Manufacturing methods and listened to Sandy Munro when he told Tesla why their first models were unsatisfactory. The giga-press idea came from Sandy Munro, not from Elon Musk, and Elon is a great listener and learner.
The Captain
Not necessarily. It depends on whether there are higher costs to obtain that 2x production speed.
If it costs you the same and you go 2x faster, then your factory cost portion is likely to be much less. But if it costs you (hypothetically) 2.1x more to go 2x faster, then your factory costs might even go up! Sometimes faster costs more.
As we’ve discussed before, that was a temporary circumstance. If you click through to the underlying article, Ford thought all their e-products would be unprofitable at first, but the Mach-E was profitable from the day it launched:
Back in March, Ford CFO John Lawler admitted that the automaker’s all-electric vehicles wouldn’t be profitable until the second generation arrived, but the [Ford Mustang Mach-E proved to be an exception to that rule early on with the EV crossover was turning a profit from the day it launched, as Ford Authority reported last September.
https://fordauthority.com/2022/06/ford-mustang-mach-e-is-no-longer-profitable-for-the-blue-oval/
Rising supply costs changed that, and the Mach-E went unprofitable for a bit. But Ford responded the way that Tesla did, and raised prices and just kept selling 'em:
…and then when materials/supply chain issues were worked out and input prices reverted, they lowered prices again (also like Tesla). But they’re still higher than where they were:
So with Ford getting a higher price on the vehicle, and input materials costs having fallen from back in June 2022, it is entirely possible - perhaps even probable - that the Mach E is again profitable for Ford.
No small part of that factory cost is labor. Go 2X faster and one is likely to cut the cost of labor significantly. Moreover, the films we have seen of Tesla factories in operation compared to other car factories appear to show far fewer people. More saving. And, yes, the cost of the machine to do a giga casting might be high, but the labor saving and the small parts saving will be substantial and on-going.
I mean, that’s certainly a possibility. But it can also be the case that the added cost of the equipment needed to go 2X faster and use fewer workers could balance out the labor savings. Or even exceed the labor savings. You can’t tell things like that based on films of the factories.
I suspect we’ll find out soon enough, as the EV and BEV production volumes of legacy automakers increase. As they grab market share, there will be more data to figure out whether Tesla’s very large margins were due more to Tesla being able to make BEV’s more efficiently, or whether they were due more to Tesla being able to charge a premium for their vehicles.
Another piece of the puzzle is that I often hear claims that BEVs require a whole new assembly line. That you can’t rework an existing line. If that is the case, every car maker will be investing in new lines as they bring BEVs to the market. Those new lines will likely have all of the latest cost saving innovations in them.
That would imply that Tesla will not have a significant cost advantage. Being among the first to produce BEVs in volume, they might even be at a disadvantage as their lines get older and may lack the latest innovations. Alternately, they have to spend money to improve those older lines, again increasing their overall costs.
It will be interesting to see how this plays out.
—Peter
Don’t forget SARE: sales, administration, research, and engineering. Not to mention marketing and advertising.
Accountants can do wonders with these numbers. Do you want them larger or smaller?
All these numbers are fuzzy at best. They are only approximate. When used for decision making, investigate carefully.
Everything you say is true but you are forgetting the elephant in the car. It is estimated that for most BEVs the battery accounts for about 30% of the cost. Battery costs are declining, but at the moment they are why ICEs are generally cheaper to produce than BEVs.
Battery production may be Tesla’s biggest cost advantage over competitors. That and the fact they seem to be the only BEV manufacturer to design their car like a computer. That was illustrated by Tesla’s remarkable ability in 2022 to reprogram their cars to use different computer chips in response to supply problems. No other car company was able to do that.
Where (exactly) did the idea that Tesla produces cars faster come from? Maybe I missed it but I thought the former VW CEO was talking about the assembly line labor time, per car. VW was at 30 hrs, hoping to get to 20. While Tesla was at 10 hrs per car. Maybe this is the same thing as being faster…but maybe it is just more efficient (i.e. the cars come off the assembly line at about the same speed but requires fewers employees.) I don’t know the answer to this.
There are lots of reasons why Tesla could have lower assembly time not even considering the gigapress castings. For example the Model 3 design puts the glass roof on very late in the assembly line allowing much easier/faster assembly of other parts via more robots. Anyone could do this as well.
Tesla also has the octavalve HVAC system that integrates the heating/cooling of the cabin and the battery. More efficient for MPGe but also many fewer systems and small parts to assemble. Anyone could do this (assuming no patents), but they haven’t.
Mike