Tesla Model 3 teardown points to $28k in potential material and production cost | Hacker News Reader

Tesla Model 3 teardown points to $28k in potential material and production cost

electrek.co

64 pointshippich45 comments

”They claim that their cost analysis resulted in materials and logistics costs of $18,000 and labor costs of $10,000 for a total cost potential cost of $28,000.”

If a factory worker costs $10,000 a month, that would be a person-month of labor per car, or 4 persons to produce a single car in a week, or 20,000 to produce the 5,000 each week that Tesla aims for.

Because of that, I doubt that $10,000 labor costs number is correct.

Reading the referenced https://www.wiwo.de/technologie/mobilitaet/elektroauto-zerle..., it talks of production rather than labor costs, so I think my suspicion is right.

That labor could have been done by robots (this is just an estimate) or contractors farther in the supply chain.

I have visited the Toyota factory in Ontario, Canada. What stood out to me was how little people was involved, literally everything was automated and the only labour people were doing was testing the cars by driving it (or even just checking the colour) when it came out or standby mechanics to maintain the machines.

I see Tesla as incorporating more labour than regular cars but not extensively as modern car plants have pretty much eliminated the need for humans.

That is rather interesting, since Tesla is famous for wanting to turn the whole assembly line into a giant robot, but Toyota is still in a cycle of being very open to using more people. There's plenty about this in the automotive and business press, the following isn't a great example.

https://qz.com/196200/toyota-is-becoming-more-efficient-by-r...

Labor does not mean only people but the cost of assembly. Robots aren’t free they need to be maintained by highly skilled and in the current market also highly paid workers and theya aren’t free to run either. They have expensive consumeables and they need to undergo constant maintenance for even basic things as maintaining their zero.

Yea I included the mechanics who do maintenance in my paragraphs when I was talking about labourers, you might have to reread that part, when I was there, they largely hang out in one roofed area within and arent utilized as heavily as you expect.

I think from a layman perspective we expect mechanics to be constantly busy and fixing and maintaining but they only fix if there are issues and with a company like Toyota the realization is that they're so experienced and familiar that the maintenance costs reduce with time simply because its cheaper and much more efficient that way and because its much easier to fix the same issues.

As well, a mechanic that is fully utilized is actually a bad thing as that means you plant is constantly being shut down to fix things and not producing cars, as thats how you fix things in a TPS system.

Also car plants nowadays are wonderfully automated, the materials were being shipped around using driverless carts with sensors and music to alert people or things in their way. Its also suprisingly devoid of human noise aside from the machines clanking.

The biggest cost of any car is R&D.

So this article doesn't tell us much.

It tells you what Tesla's margins are going to be, as R&D cost per unit decreases/is amortized over more units as production scales up. Check out their free cash flow [1]. You'll see that it's always in the red before production scales up (Model S first, then the X, and now the 3).

TL;DR If production continues to scale up, the Model 3 will be wildly (relatively speaking for the auto industry) profitable (and the battery cost is possibly under $100/kw, which would be big if true).

Sidenote: Elon's enormous pay package [2] that was approved requires one final condition be satisfied; four consecutive quarters with 30-percent gross margins. If I had to bet real money, I'd say 2018Q3 is when those margins begin to be realized, his compensation delivered 2019Q4/2020Q1, and his plan to go to Mars accelerate (possibly as a direct SpaceX customer).

Sidenote #2: The Model 3's battery pack has the lowest amount of Cobalt in the industry (2.8%, known state of the art is ~8%). Panasonic (Tesla's battery production partner) is currently working to remove the need for any Cobalt whatsoever [3]. Good for margins, good for production throughput, mixed bag for conflict areas Cobalt is sourced from. Can't win 'em all.

[1] https://i.imgur.com/HjyKxTm.png

[2] http://money.cnn.com/2018/03/21/investing/elon-musk-tesla-pa...

[3] https://ca.reuters.com/article/businessNews/idCAKCN1IV14Y-OC...

Usual disclaimer: TSLA investor

The Ford F150 makes a gross profit margin of $10,000 to $13,000/vehicle. It sells 800,000 F150s each year, for a profit of $8 billion or more on just the F150s. By the way, F150s sell for about the same as the Model 3.

Porsche has gross margins of $18,000 to $23,000 on its vehicles, which are comparably priced to Tesla's other models.

Really, if you drill down into the actual numbers, you'll see that Tesla isn't significantly more profitable than other car makers on a per-vehicle basis, and they're certainly not wildly more profitable.

With the Model 3, Tesla needs to scale up just be in the black. It would need to sell somewhere on the order of several hundred thousand Model 3s each year to approach the per-vehicle gross margins of the Camry, and more than a million Model 3s each year to achieve the gross margins of the F150. Unless they drastically improve quality control, that's simply not happening.

“In North America, Ford had a record operating margin of 12.9 percent. This compares with rival General Motors' first-quarter operating profit margin of 8.7 percent, and Fiat Chrysler Automobiles's 7.2 percent in North America.”

Note that none of these margins approach 30%.

https://www.cnbc.com/2016/04/28/ford-motor-sets-records-for-...

Are you seriously comparing the operating margins of profitable companies to a company that has yet to make a profit based on its operating activities? (Selling credits doesn't count, that's a one-time stunt.)

The numbers for Ford, GM, and Fiat Chrysler include all of their global activities, including their sales of lower-margin cars, compliance cars (i.e., no-margin or negative-margin vehicles, massive R&D, international operations, etc.). You are literally trying to compare the gross margin on a single Tesla model to the overall operating margins of entire companies...that's not even apples to oranges. That's comparing apples to zucchinis.

I'm making observations with all publicly available information I can get my hands on.

I'm not saying Tesla's margins are 30%. I am saying that if they're able to attain their production goals, it's likely they will hit their target margin, which are above legacy automakers.

You do understand that Tesla's gross margin that you are arguing around don't include R&D, and I'm not sure if covers SG&A too.

While the numbers you cited include every expense.

“So how does Tesla actually do on a per-vehicle basis? To that, we turn to the quarterly reports. Before Model 3 production became significant — aka, just S and X sales — Tesla was earning a 25% non-GAAP margin / 27,9% GAAP margin in their automotive division. These are very healthy margins. As Model 3 production ramped up — and famously encountered difficulty — Tesla’s gross margins fell, bottoming out at 13,8% non-GAAP and 18,3% GAAP, before rising back to 18,8% non-GAAP and 19,7% margin.”

https://m.dailykos.com/stories/2018/5/29/1767826/-The-War-on...

And you can't compare these numbers to other car makers, because they include all the lifecycle costs for each car model, including the design, production line spin-up, etc.

If you're an auto startup, computing numbers the Tesla way makes sense: Tesla is growing rapidly and will outgrow its current high fixed costs. If you're an established automaker that's not growing, adding in everything makes sense.

>Check out their free cash flow [1]. You'll see that it's always in the red before production scales up (Model S first, then the X, and now the 3).

That's a charitable interpretation. From your link it's nearly always significantly in the red, outside of a couple anomalous quarters 5 years ago, and the one quarter they sold a bunch of ZEV credits.

It certainly doesn't look like "it's positive, except when they're scaling up!" to me. I guess one sees what they want to see.

This is because they're always spending money to grow. They do this because they want to grow and are growing. This graph shows you they spent money to launch the Model S, then went profitable. After this they invested to launch the X, then went profitable. Then they invested to launch the 3, and now this year they aim to be profitable again. And you know what they're going to do after that? Spend money to launch the Y and Semi, then go profitable.

Can they use the profits from S/X to deliver the 3? Yeah, but not on the timelines that their customers and the market demands. People want their Model 3s today.

One sure mark of a hater is that there doesn't exist a configuration of reality that will satisfy their objections. Is Tesla spending borrowed money to accelerate their growth and deliver products sooner? They suck, they should be profitable. Are they using solely their profits to expand more slowly? They suck, because they're having trouble delivering their products on time. Great, awesome contribution.

>This graph shows you they spent money to launch the Model S, then went profitable.

The graph doesn't "show" anything. You are interpreting it in a specific (favourable) way. I don't agree. Also, free cash flow positive is not "profit".

>One sure mark of a hater is that there doesn't exist a configuration of reality that will satisfy their objections.

Oh, there it is. Any criticism means you're a hater.

I'm looking at the numbers. That's reality. You, on the other hand, are operating on future assumptions that may or may not come to pass. "Becoming profitable" isn't a given, just because Elon said so. As of right now, they are losing money at an accelerating rate. I guess we'll see.

And by the way, I'm not hater. I can see Tesla being the next Mazda, who make some of the most popular cars in my area. Great cars, great company. Do you know what its market cap is? $8BB. At Tesla's current market cap they need to build millions and millions of cars at substantial profit. You can believe that they'll achieve that and an investment today still wouldn't be worth it. Does that surprise you?

What you're missing is that Tesla does not want to be a Mazda, nor a Ford, nor even a Toyota. They want to be an industrial behemoth that powers every part of the carbon-free economy. They want to be a GM, GE, ConEd, Exxon and Uber rolled into one. This is not some bizzare internet conspiracy, this has been stated in plain English multiple times and has been voted on by their board.

Now, you might think that's too risky, too ambitious, or they'll never be able to execute etc. Whatever. But you can't judge them based on your own made-up ideas and standards.

He already pretty clearly said what he thought. I'm looking at the numbers. That's reality. You, on the other hand, are operating on future assumptions that may or may not come to pass

Meanwhile your counterpoint is a very vivid fantasy of what Tesla wants to be beyond a car company, which is something it’s not doing particularly well. You’re engaging in just the kind of fantasy “Future assumptions” in question, and really just buying into empty PR and marketing. The evidence suggests that Tesla is a car company, and struggling to be just that. It’s an act of intense mental gymnastics to say that no, they’re not just a car company, they’re a whole economic sector that started as a car company... before they’ve even pulled off step 1. Maybe they will be, but the odds are not in their favor, and it’s going to be a moot point if they can’t even turn a profit on their cars.

Arguing from an imagined future should be the job of marcom, not you. It’s not a failure of imagination or a lack of faith for rational people to look at what is, not a vision of what may or may not ever be.

>They want to be a GM, GE, ConEd, Exxon and Uber rolled into one.

They have a long way to go to be compared to any single one of those companies, let alone all of them combined.

Betting on any company to be the biggest in the history of history is a fool's errand.

>But you can't judge them based on your own made-up ideas and standards.

I can't judge them by my standards? News to me.

What if they’re GE and not Mazda? The world needs an incredible amount of stationary storage to use renewables alone for generation.

Can they use the profits from S/X to deliver the 3? Yeah, but not on the timelines that their customers and the market demands. People want their Model 3s today.

Is it about what people demand, or what they promised when taking deposits?

Per unit cost is an extremely important metric to an auto company.

Typically the biggest cost per new car is the factory and tooling, not the R&D.

That's why they say potential cost.

And it's only selling for $35k? I thought the rule-of-thumb was that you charge 3x the "bag of parts" for a product you make, sell, and service. Apple certainly got this memo... Elon is leaving a lot of money on the table.

The as-configured vehicle they examined was sold for at least $49,000 (but probably more likely $54,000 with Autopilot), as they're only selling large-battery and high-trim interior now. The headline lumps material and labor together, but materials was reported as $18k, which is a surprisingly exact 1/3 of $54k.

The base model would have to have $12k of materials for the same ratio. I don't think that's likely (since much of it's invariant and especially since the $54k figure includes the Autopilot upgrade, which is only software) but it will have 1/3 less battery, which is a decent part of the cost of the car. So the base model will probably not be as profitable, but I think that's true for most cars.

Thats for smaller ticket items. Boeing isn't 3x'ing the cost of a plane, and neither are car manufacturers. Toyota makes an average of $2800/car, and GM makes only $654/car [0].

Porche makes $17k/car... but that's on an average price of $100k.. so only 17%[1].

0. https://www.detroitnews.com/story/business/autos/2015/02/22/...

1. https://www.autoblog.com/2017/03/21/porsche-17250-profit-per...

But the parent said "bag of parts" which implies only the cost of materials, not the full per-car cost of production, and I can't tell if that also includes things like R&D, marketing, insurance, etc.

The selling point of the Model 3 is a $35k electric luxury car. It's not leaving money on the table if no one is willing to pay $100k for the car. Even if they sold it at 2x the cost, the selling point isn't "small electric luxury car" it's "inexpensive electric luxury car".

And trying to meet that target before technology allows them to is what is driving Tesla's current financial problems.

The cost of production has absolutely ZERO bearing on the sale price. Price is based on what a consumer is willing to pay for the product, that’s it. The cost of production only comes into play after that is determined, to ensure the company doesn’t sell at a loss (in which case they would probably cancel their plans), or whatever profit margin they are willing to accept.

The models currently being sold (high-capacity battery, premium upgrades) are closer to the the 50-55k range.

It depends on the product. Most automakers only make single digit percent profit.

The 3X rule assumes a middle-man distributor. When selling direct, as Tesla is, 2X is more common.

It's not clear from the article if this cost figure includes software engineering cost which would be significant in Tesla's case.

No, gross margin doesn't include R&D. That money's spent and gone, so is pretty much irrelevant towards future costs & profitability. Relevant from an academic perspective, but not from a business perspective.

Or at least that's the traditional view; purchasers of Model 3's now expect OTA updates so there is a little bit of R&D cost in the future but it's very minor compared to what's already been spent.

That also doesn't include self-driving R&D, but that's not included in the $35K base model so is also irrelevant.

>purchasers of Model 3's now expect OTA updates so there is a little bit of R&D cost in the future but it's very minor compared to what's already been spent.

Toyota spent almost $10 billion on R&D last year. Ford $8 billion. VW spent $15 billion.

Tesla spent $1.5 billion. What is it that makes people think that Tesla's R&D spend is so significant, and that it will diminish drastically in the future? This is an R&D intensive industry.

Toyota will spend roughly $0 in R&D on the 2017 Camry in 2018.

Having worked in some Fortune 100s we never counted software development as R&D unless it was true blue-sky work unassociated with a project.

All the routine work was included in Run The Business budgeting just like tangible material purchases.

On a number of occasions, as a (software) program manager at several very large firms, I was tasked with taking every single piece of work that happened in the previous calendar year and seeing whether we could classify it as 'R&D' in order to qualify for one of the many tax breaks on offer in the various places that we did business.

The criteria for qualification usually boiled down to 'was there an element of unknown quantity in the project?' - which left enormous room for interpretation, as basically anything that wasn't maintenance (RTB) could be justified.

We would regularly, across the firm, get $50m tax benefits a year out of this effort.

I'm not an accountant, but I'd be interested in just how rigorous the notion of 'R&D' is in a company's annual returns. It smells of PR to me.

I think its pretty clear that this is an estimate of the material cost data point.

The cost of the car for manufacturer might be: 60% goes into into the factory and tooling, 20% into the parts and 20% is the rest.

If parts alone cost $28k and the car is sold for $35k, there is no profit in the Model 3 even if the car is successful.

What I really want to know is how much Tesla pays for product recall insurance per car.

The one they tore down has a price of at least $49,000. They haven’t started producing the $35,000 version yet.

They mention parts cost $18k in the article, not $28k.

Reply on news.ycombinator.com