Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

[replicant]

If I understand correctly, structural battery packs require 4680 cells, which must be nickel-based (yesterday's call) and are reserved for LR 3/Y, S/X, Semi, CT (battery day).

I thought they planned to use LFP for robotaxis, because they don't need to be LR and thanks to LFP's lower cost, higher life cycle and material abundance. But that means Tesla won't enjoy the benefits of structural packs for the vast majority of their cars

Could Tesla be working on producing their own LFP cells now, and looking for better integration (e.g structural packs for prismatic/blade LFP)?

 

[Singuy]

https://twitter.com/x/status/1486589517280821249

 

[Artful Dodger]

Macro is weighing down on TSLA.

That's why we feel beat up.

Macros not weighing on TSLA at the moment though: (07:30 ET)

Nothing we can do about it unfortunately.

Kinda feels like this was always shortze's plan (since Jan 3rd, anyway).

"Ignore the man behind the curtain".

 

[nativewolf]

And only Elon could authoritatively state that, thus he was on the call.

Which is silly to me. There's demand with the 3 models: CT Y 3 to hit those numbers. Probably hit that with Y alone across the world. I don't see FSD moving people that were not going to buy an EV. I don't see other OEMs having capacity to build enough EVs. Therefore...I don't like the answer.

 

[H Mak]

That was the first Tesla call of any type over many years, when I didn't come off the call buzzing with excitement and itching to buy more shares. Record car deliveries, record revenues, record margins, record earnings and 50%+ annual growth affirmed - I know, I know, how is that not amazing! I suppose because there was so much emphasis on things that might happen in the future. Personally, I'm discounting the "things that might happen in the future" and concentrating on cars in my valuation model. That is still pointing to a mult-trillion dollar company within a very few years, along with a share price of several thousand dollars. My investment is safe. Maybe I need a bigger brain to understand robots and so on. I certainly need another coffee.

 

[Singuy]

https://www.reddit.com/r/teslainvestorsclub/comments/sdt606

https://www.reddit.com/r/teslainvestorsclub/comments/sdt7et

 

[mongo]

If I understand correctly, structural battery packs require 4680 cells, which must be nickel-based (yesterday's call) and are reserved for LR 3/Y, S/X, Semi, CT (battery day).

I thought they planned to use LFP for robotaxis, thanks to their lower cost and higher life cycle. But that means they won't enjoy the benefits of structural packs for the vast majority of their cars. Maybe Tesla is working on producing their own LFP cells and/or looking for a better integration (e.g structural packs for prismatic LFP)?

What do you think?

A pack with modules cost more than a pack without.
A pack without modules is also termed structural.
The car frame can be built such that it does not require a structural pack.
Total car cost depends on frame, pack, and cell chemistry costs. So a dual use frame may be lower overall cost across the pack mix versus making an LFP pack structural.

 

[Prunesquallor]

See this is what I do not get.
How?
Imagine someon approached you, and said I think you and nine neighbours should just use this one car instead. It will be available for the most part with minimal to no wait. Would you do that? I can’t see a fast shift in this at all.

An then there is the peak time problem..ie going to and from work, where you might have to wait.

I feel many here envision being the owners of this taxi service rather than the users.


If this shift happens significantly in the next 5 years I will be shocked. I am thinking a moderate sized shift away from personal vehicles over the next twenty years.

There are nearly 4 million Uber drivers worldwide.

 

[Sancho]

To me the most important information on the call was that batteries are no longer a constraint to auto production. That they will have enough batteries for all their auto production plans. This is huge. Unconstrained by battery limits, they can ramp up four factories far, far more than 50% growth.

I did not get a sense from the call that chip supply is a likely limiting issue this year. What I heard was something like: “Yes, it’s still a pain-in-the-rear thing we need to manage, but we are managing around it, and during the year we expect the chip shortage to ease.”

 

[jerry33]

As others noted above, most people don't want to share a car with a bunch of random strangers. These $25k FSD enabled cars will likely be shared by small groups of "trusted" people (think friends, family, neighbors).

I expect this to have the same impact as car pooling. e.g. a minuscule amount. There are just too many barriers such as work locations vs home locations, work start times (if everyone starts at the same time, they will have to start much earlier to get everyone to work on time, and work end times. No one can have a job where they sometimes work late. The idea of private car ownership is that you can go anytime in privacy and there is no limit on flexibility. Robotaxi might work if the taxi always arrives within five minutes of the request, and have no no-shows the way Uber does.

 

[petit_bateau]

Why do think think they cannot use a cast front with a non-structural 2170 pack? Frame rails appear to work with both types of packs.
Tesla shows off new structural battery pack with 4680 cells at Giga Berlin - Charged EVs
View attachment 760838

I am not totally certain, but this picture seems to be of a pressed front. Maybe I'm mis-seeing it ? If it is pressed, then is it a local Berlin pressing, in which case that would be a dead-end. Or is it an imported presssing to go along with imported 2170s. We know Berlin can cast these, so why are they not using the castings. Or are they using the castings and I'm not recognising it sufficiently well ?

(By the way I'm very happy to see the design cleverness that has cleavage points to allow for these sorts of production flexibility.)

 

[mongo]

There was a lot of coyness and word smithing around the battery question. They clearly still have 4680 issues but I perhaps did not realize that the chip issue was such a limiting factor that it dwarfed battery production. It's clear they are obscuring something. Berlin was designed to use 4680 and anything less is not ideal. Panasonic is not even going to be close to ready to produce 4680 this year. So who else is? CATL? LG? Samsung? BYD? So does this mean the chip limitations are so bad that Austin can run off of Kato? If that bad why produce any 3s? Why not just make Ys? Dual cast Ys will have fewer long term maint issues (one would think). Anyway I found the battery/chip discussion points to be somewhat confusing and even contradictory.

Say what now?
Tesla (TSLA) Q4 2021 Earnings Call Transcript | The Motley Fool

Drew Baglino -- Senior Vice President, Powertrain and Energy Engineering


Sure. Yeah, yeah, sure. So throughout 2021, we focused on growing cell supply alongside our in-house 4680 effort to provide us flexibility and insurance as we attempt to grow as fast as possible. As we sit today, sales from suppliers is actually -- it sort of exceeds our other factory-limiting constraints that you mentioned, Elon, in 2022, or to say differently, 4680 cells are not a constraint to our 2022 volume plans based on the information we have.

But we are making meaningful progress of the ramp curve in Kato. We're building 4680 structural packs every day, which are being assembled into vehicles in Texas. I was driving one yesterday and the day before. And we believe our first 4680 vehicles will be delivered this quarter.

Our focus on the cell, the pack, and the vehicles here is driving yield quality and cost to ensure we're ready for larger volumes this year as we ramp and next year. And the 4680 and pack tool installations here at Giga Austin are progressing well with some areas
producing first parts. And the internet has also noticed that.
Elon Musk -- Chief Executive Officer and Product Architect


Yeah.
Drew Baglino -- Senior Vice President, Powertrain and Energy Engineering

Yeah, I was touring the factory -- the cell factory here. I'm super pumped. It's like a really exciting accomplishment for us to bring everything into one Austin factory here in Texas.
Elon Musk -- Chief Executive Officer and Product Architect


Absolutely. And just to repeat Drew's point, we are still -- we still expect to be part or primarily chip-limited this year. So that's the thing -- that's actually the driver.
Drew Baglino -- Senior Vice President, Powertrain and Energy Engineering

Yeah.
Elon Musk -- Chief Executive Officer and Product Architect


And that chip limitation should alleviate next year. And then probably, we transition into a cell limitation battery, you know, total gigawatt-hours of cell limitation, which is when the 4680 will become very important.

Berlin starting with 2170 has been known since at least the open house last year. Their cell manufacturing building doesn't even have walls yet.

 

[CHGolferJim]

My projections did not factor in the payroll tax on Elon's exercised options. I don't believe anyone else had that factored into their forecasts.
It is a one time item that analysts should back out. Excluding the payroll tax, EPS would have been $2.84.

The payroll tax thing hasn’t appeared in the P&L’s before? If not, what’s different about these tranches? Is it an end-of-year entry for all tranches recognized in 2021?

 

[NicoV]

https://www.reddit.com/r/teslainvestorsclub/comments/sdt606

https://www.reddit.com/r/teslainvestorsclub/comments/sdt7et

Credit Suisse seems to think that lower interest expense is a one time bonus this quarter. I’m not an accountant but I think that if you pay back loans, you don’t have a related interest expense anymore in any of the upcoming quarters.

 

[petit_bateau]

There was a lot of coyness and word smithing around the battery question. They clearly still have 4680 issues but I perhaps did not realize that the chip issue was such a limiting factor that it dwarfed battery production. It's clear they are obscuring something. Berlin was designed to use 4680 and anything less is not ideal. Panasonic is not even going to be close to ready to produce 4680 this year. So who else is? CATL? LG? Samsung? BYD? So does this mean the chip limitations are so bad that Austin can run off of Kato? If that bad why produce any 3s? Why not just make Ys? Dual cast Ys will have fewer long term maint issues (one would think). Anyway I found the battery/chip discussion points to be somewhat confusing and even contradictory.

FSD- agree with you. Mud He should never have opened his mouth on the optimus stuff. Silly. They have no credibility on FSD timelines (and I think Tesla is ahead of everyone and I like the approach). FSD will require new/better hardware as many had stated (though it didn't stop others from arguing). This looks to me to be a huge gaping question, I saw little risk before in Tesla now the uncertainty on this one issue just increases risk. If FSD doesn't work out...and that is not a 0 probability Teslas future earning now seem to be in question.


Berlin looks to be sucking hind tit so to speak...btw that expression never really makes sense until you see pigs or dogs nursing a litter. At this point, and granted they did not forsee the global supply chain issue, why did they bother rushing to open it? Just slow slow slow roll it. They have been slow rolling but why obfuscate it.

Indeed.

I think all that CATL LFP from Shanghai has come onstream in capacity ahead of Tesla expectations. That in turn is giving a lot of 2022 production flexibility. So the sensible thing for a prod mgr is to question how best to deploy that. Perhaps - and this is my suspicion - they are switching more Fremont to LFP, freeing up 2170 to pump-prime Berlin, and that in turn allows the birthing pains of 4680 to occur once in the transition from Kato to Austin, not simultaneously at Berlin (and anyway the permits are slow at Berlin, so Tesla can also be political point-scorer for free : the message will have been understood in Germany). But with no permissions in hand for Berlin or Austin last night there are still moving pieces, so no reason to 'fess up to everything on the call.

Furthermore I think they may have 4680s in some cars in Austin, but not client-ready, or even certification-ready (because we know they are not through Austin-certification yet). And - simply because it was not said - that suggests it was Kato 4680s not Austin 4680s. So, again, I think the Austin 4680 line is not running yet. Question is, what is the yield on the Kato 4680 line ?

CATL delivers first LFP cells to Tesla Giga Shanghai | electrive.com

www.electrive.com

Report: CATL's New Battery Plant In Shanghai Already Supplies Tesla

CATL shows some extreme expansion pace, as its latest lithium-ion battery plant in Lingang, Shanghai, China reportedly started production.

insideevs.com

 

[TSLA Pilot]

I would hae liked to see Tesla put more emphasis on the 25k car . It sounds like its in the scrap heap. Cybertruck is cool, but IF supply chain was no problem, a 25k car would advance the mission more than any other model they make now., in my opinion. I am not in the FSD is a panacea to carbon ouput camp. I agree with some here that state, the majority of people, especailly with family/work tools/ etc will require a decade or more to transition to a shared vehicle approach, if ever. Would I use it? Maybe here or there, but not a lot. At this point i still want my OWN vehicle, and I feel most others do as well. If not ride-sharing /joint ownership would be way more popular.

Now it is Optimus. Few years back was the solar roof. Once again I feel Elon's timelines are optimistic. They learned over the years to temper guidance on vehicle production, but continue to over promise big idea timelines. I have no doubt in Teslas future success but feel we need to stop burning fossil fuels asap, so feel disappointed in the focus. Sure FSD will make tons of money but I invest for the mission and ultimately thats where my priorities lie.

Where does optimus fit into the master plan? How does a robotic workforce help with respect to global warming/climate change ?

Think bigger.

Optimus won't:

1. Shoot other employees in the parking lot (a la recent Fremont murder).
2. Sue, and win HUGE awards, for discrimination lawsuits (while trying to get family members hired at the "oppressive environment" Tesla; can't make this stuff up).
3. Complain about a 24/7 work schedule.
4. Get tired.
5. Take vacations.
6. Require medical insurance.

Optimus moves the needle at Tesla by being a force multiplier, allowing for greater production with less humans, especially the kind of humans that sometimes get hired, but shouldn't have been . . . .

 

[mongo]

I am not totally certain, but this picture seems to be of a pressed front. Maybe I'm mis-seeing it ? If it is pressed, then is it a local Berlin pressing, in which case that would be a dead-end. Or is it an imported presssing to go along with imported 2170s. We know Berlin can cast these, so why are they not using the castings. Or are they using the castings and I'm not recognising it sufficiently well ?

(By the way I'm very happy to see the design cleverness that has cleavage points to allow for these sorts of production flexibility.)

View attachment 760843

Wow, you're right, great point! Definitely looks like all stampings in front.
Here is cast from Texas.

 

[Sancho]

Here’s another reason why TSLA is relatively undervalued now: not only are profits projected to rise substantially over the next two years, but the business risks that require investors to discount future profits is relatively much lower than other high growth investment alternatives.

1. Demand is pretty much a lock — even if the economy dives into a recession.

2. There’s no looming competition that would necessitate price discounts.

3. There seems to be no constraint on raising prices, so inflationary cost of parts isn’t a real concern.

4. A tightening job market doesn’t appear to be a concern. Tesla hires who they need, and the robots don’t seek unionization.

5. Tesla has already proven it’s the best in the industry in managing chip shortages.

6. Another Covid surge could possibly close one of the factories for a few weeks, but otherwise the pandemic is not a risk factor.

Any other company with near term profit growth potential like Tesla would need to discount that potential for the multiple business risks that don’t seem to be a factor for Tesla.

Bullish AF

 

[nativewolf]

GPUs are an established, growing ecosystem.
Dojo is all new.
It's like GM competing with Tesla in EVs, there is both a technical and performance hurdle to overcome while the target is also moving.
Dojo can catch up, but it may do so on the basis of cost of additional capacity basis as opposed to straight replacement of existing clusters. It also sounded like some things do not map well to the existing GPU architecture

Tesla (TSLA) Q4 2021 Earnings Call Transcript | The Motley Fool


Why do think think they cannot use a cast front with a non-structural 2170 pack? Frame rails appear to work with both types of packs.
Tesla shows off new structural battery pack with 4680 cells at Giga Berlin - Charged EVs
View attachment 760838

So IANAME (not a Mech Eng) but if that structural battery pack is providing any structure....it begs the question as to a car made without one vs one made with the structural battery pack. Would you have to get a different certification on safety? Where is the structure coming from to replace the engineered aspects of the structural battery pack. Sandy Munro and team discussed this in some detail in one show and I took it to mean that it was providing value to the entire body. I would not think that's something you'd just easily dispense with but what do I know. IANAME. Dual casting and 4680 structural battery packs (any structural battery pack) seem to offer tremendous advantages in assembly, fit, and long term reduction in service. EDIT_Ahh..I see you replied above. It's an interesting thing. On the one hand Kudos to Tesla's team for allowing the optionality!

Say what now?
Tesla (TSLA) Q4 2021 Earnings Call Transcript | The Motley Fool

Berlin starting with 2170 has been known since at least the open house last year. Their cell manufacturing building doesn't even have walls yet.

The wording had been if required they would use it, they designed a fall back. It was dependent on 4680 production coming online. In the transcript they say...no 4680 constraints. But Berlin was designed to use 4680 structural and can fall back to something else. If anything less there is a constraint somewhere. Either they have or don't have enough 4680s. The glass half full would be how great is it that Teslas engineers foresaw the need to allow flexibility and that it is designed into the product..

 

[jeewee3000]

They have no credibility on FSD timelines (and I think Tesla is ahead of everyone and I like the approach). FSD will require new/better hardware as many had stated (though it didn't stop others from arguing). This looks to me to be a huge gaping question, I saw little risk before in Tesla now the uncertainty on this one issue just increases risk. If FSD doesn't work out...and that is not a 0 probability Teslas future earning now seem to be in question.

Credibility is in the eye of the beholder. From your comments it is clear you are not convinced FSD will be achieved in 2022, nor with the current hardware.

I disagree since:
1) Elon was talking about "FSD software outperforms the average human", i.e. amount of miles driven by FSD without any incident/intervention is higher than the amount of miles driven by the average human between incidents. Since the average human is a combination of lousy, accident prone drivers and immaculate drivers that never have an accident, the average is not hard to beat, according to Elon. He said so literally: the "better than the average human-benchmark" is a low bar, one that he feels very confident they will beat this year, with current hardware.

Elon has all the data. We don't. He even acknowledged FSD timelines were off many times in the past, but still he wanted to go on the call and make these claims. After all he's learned regarding underpromising and overperforming, I'm highly convinced human-bar-FSD will indeed happen this year, with current hardware.

2) for the march of nines they will indeed use DOJO and some more years. FSD is not a static milestone, but will be finetuned in the years/decades to come. This is indeed something that will take time, and most likely hardware upgrades. But that is not what Elon was talking about on the call.

3) even if FSD takes more time, Tesla's future earnings are not in question. They are selling a $50k product at high (and growing) margins. Tesla is not doomed without FSD. Of course it's valuation will be (way) higher with FSD, but still a great investment.

Other notes regarding FSD:
- notice how the voices shouting "LIDAR is necessary for FSD" are shouting less and less hard these days? Every quarter FSD beta progress improves, without silly expensive LIDAR. Waymo and other competitors are wetting their trousers about this, I'm sure of it.
- the fact that Elon is already switching over to Optimus in his AI step-by-step program in his mind, shows that solving FSD is almost a done deal. I mean the path to FSD is so clear for Elon it isn't a question mark anymore. He is already thinking of what's next for real-world AI applications.

FSD might be off-topic most of the time, but today I took the liberty of adressing it since it was one of the biggest reveals on the call.