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

[ZachF]

1 twh manufacturing capacity in the US, vertically integrated.

As an American, this makes me happy.

 

[Nocturnal]

Yeah, Elon is waaaay too sloppy when talking about FSD. It's like he doesn't realize his words can EASILY be misunderstood. Popular press article will be written based on Elon's sloppy statements leading people to believe they don't need to ever touch the wheel (no one reads manuals) leading to AP crashes. Pretty predictable.

Eh not sure that matters. Ford and GM advertise their crap as entirely hands free and those systems randomly shut off when they feel like it.

 

[Skryll]

Eh not sure that matters. Ford and GM advertise their crap as entirely hands free and those systems randomly shut off when they feel like it.

Do they disengage like FSD when it missed my hands on the wheel, and more than once did that while doing a curve with me majorly cursing and salvaging the trajectory?

 

[Ogre]

Looking back at some numbers, Drew said they were making 1,000 packs per week (4680 cells). This suggests the current run rate is 4.25 GWh. And Musk said growing exponentially. Wasn’t clear what the split in terms of location, but feels like 4680 ramp is finally hitting it’s stride.

 

[ByeByeJohnny]

Looking back at some numbers, Drew said they were making 1,000 packs per week (4680 cells). This suggests the current run rate is 4.25 GWh. And Musk said growing exponentially. Wasn’t clear what the split in terms of location, but feels like 4680 ramp is finally hitting it’s stride.

Do we know if the ramp is because of better efficiency though? Maybe they just put in three times as many lines?

 

[Xepa777]

Eh not sure that matters. Ford and GM advertise their crap as entirely hands free and those systems randomly shut off when they feel like it.

...so we're holding ourselves to the standard of Ford and GM now? Hope we can do better than that...

To me, "FSD" is not financially meaningful until it can either (a) robotaxi or (b) enable people to move around in cars while distracted/asleep. You see a lot of anecdotal evidence even here/reddit on FSD's pitfalls, and objective data from FSD aggregation sites on interventions. It's just not there. Thankfully Tesla's valuation doesn't depend on it, but there is real potential that it becomes a financial liability (and class action lawsuit risk) if it doesn't hit that level by the end of this decade.

Do I think that will happen? No. I think Tesla undergoes some slipperly slope where they gradually introduce new hardware (like to cameras or HW4) and makes people pay for those upgrades from previous cars for "enhanced FSD." That's why their terms and conditions around FSD changed when the SR+ was released (around the time I bought my Model 3) for greater legal liability protection.

I also think worst case scenario robotaxi can be introduced in geo-fenced neighborhood suberban areas and still make an absolute financial killing.

It'll be interesting to see what happens with the robotaxi/cheaper car platform, order of operations, and what those end products actually become. That'll reveal a lot.

 

[Xepa777]

Yep, I believe @Gigapress did a breakdown earlier this week or last week just on this. That the Semi beginning 2023 will be a gigantic money print machine for Tesla.

Probably the same for Megapacks too: cha-ching!

...or it could not, because of ramp-up cost, not using 4680s, and using Giga Nevada, which might be sub optimized with a lot of manual labor involved. But Semis need to come out to satisfy upset corporate customers and to gain enough confidence in a certain deployment threshhold to invest in a highly automated process.

But naw, everything's automated and Tesla Bots will make that *sugar*, meaning gross margins of >60%. More likely to hear highly upvoted 2000 word posts around that lol.

 

[Nocturnal]

...so we're holding ourselves to the standard of Ford and GM now? Hope we can do better than that...

No. My point is that I find the claims that adult humans who pay 60-150k on a vehicle and 15k on software aren’t stupid enough to hear a snippet of an earnings call and decide to go to sleep in their cars after ignoring all of the warnings and bypassing safety measures.

You won’t find many with a lower opinion of people than I, but I find those critiques to be ridiculous.

Functionally, I see the worst case is FSD becoming a world class L3 system and it’s not far off. Maybe robotaxis never happen. If so we will still be rewarded handsomely.

 

[Nocturnal]

Do they disengage like FSD when it missed my hands on the wheel, and more than once did that while doing a curve with me majorly cursing and salvaging the trajectory?

From user experiences I’ve seen it will just suddenly turn off which is not what you want if your hands are not on the wheel. People attack Tesla for driver monitoring but there is no system today that I trust not having my hands ready to take over.

 

[Cosmacelf]

Elon started out saying Tesla could be valued at Saudi Aramco + Apple valuation or greater. Or twice Saudi Aramco. I took that as comment about the price. I can retract my bingo if needed :-(

Wups, you are right! That was absolutely a stock price comment. Yea, I got Bingo too then!

 

[Cosmacelf]

I was a bit distracted when listening to some parts - did Elon say 20,000 Semis for next year? If so, I think that's huge.

 

[Chunky Jr.]

serious operating leverage...

https://twitter.com/x/status/1582869952415272960

 

[woodisgood]

I was a bit distracted when listening to some parts - did Elon say 20,000 Semis for next year? If so, I think that's huge.

He said ramps take one year, 50K for 2024.

 

[Chunky Jr.]

If Tesla can sell every car they can produce for the foreseeable future, as Elon claimed in the call, then there should not be any need to lower prices (unless Elon is taking lowered prices into account and not telling us).

 

[bkp_duke]

...or it could not, because of ramp-up cost, not using 4680s, and using Giga Nevada, which might be sub optimized with a lot of manual labor involved. But Semis need to come out to satisfy upset corporate customers and to gain enough confidence in a certain deployment threshhold to invest in a highly automated process.

But naw, everything's automated and Tesla Bots will make that *sugar*, meaning gross margins of >60%. More likely to hear highly upvoted 2000 word posts around that lol.

Dang, someone's grumpy.

Let's assume the 2170s going into the Semi are a full MW (original specs, although probably a tad more efficient). Let's also assume the cost of those, at the pack level is $120/kWh (Tesla is probably less, but let's factor in some inflation).

Pack cost = $120,000/kWh minus $45,000/kWh = $75,000 cost for the pack after IRA production tax incentives.

Tesla is selling them for $180,000 (or they were when the prices were up for the Semi - let's assume some commercial buyers have locked that in).

$180,000 - 75,000 = $105,000 to break even. Add in R&D, and cost of the cab itself, and that still leaves a lot of room for a healthy profit margin.


Additionally, beginning Jan 1, 2023, the buyers are eligible for a $40,000 IRA tax break on purchases of the Semi. Given that Tesla has removed pricing from their order page, let's make a reasonable assumption they will plan to raise the price, and capture half of that ($20,000 price increase).


What am I missing here? R&D costs are amortized over multiple years, the facility these are built in is not close to a GF in size, it's relatively cheap to build.

How would this not be a money-printing machine?


(and look at that, far less than 2000 words).


EDIT - I could fully see the Semi being a loss for the first year without the IRA in place. But with it, that's a lot of extra money coming Tesla's way, for a product they had already planned out years ago (and probably planned for it to be reasonably profitable).

 

[Nocturnal]

So both the plaid and semi moved from plans to use the 4680 cells to 18650 and 2170. Is that a short term issue or is 4680 tech destined for mid tier applications? Financially that doesn’t matter if they can make them cheap and at volume.

 

[MC3OZ]

So both the plaid and semi moved from plans to use the 4680 cells to 18650 and 2170. Is that a short term issue or is 4680 tech destined for mid tier applications? Financially that doesn’t matter if they can make them cheap and at volume.

We don't know if 4680 cells are currently cheaper, eventually they will be.

4680 cells are best used in cars with a structural pack, currently Model Y.

Tesla never announced a public plan to use 4680 cells in the Semi or Plaid Model S.

There was strong speculation that the Plaid+ used 4680 cells, but Tesla never publicly announced that.

If we take a 5 year time horizon we might find all Tesla vehicles use 4680 cells in around 5 years time, but the Semi and Model S/X might be the last vehicles to migrate.

 

[ZenRockGarden]

So both the plaid and semi moved from plans to use the 4680 cells to 18650 and 2170. Is that a short term issue or is 4680 tech destined for mid tier applications? Financially that doesn’t matter if they can make them cheap and at volume.

It looks to me like the 4680 failed to ramp as intended, Tesla is having to take the effort in-house which means it will get going, but not in the timeframe and scale since the Semi's gotta go to customers right now.

So you reluctantly build some near term product out of existing tech - just like Austin is gonna make 2170 Model Y's for a bit.

I'd expect the long term goal is to still make 4680 work, and go into the highest volume products for the efficiency and cost gains. They just arrive later than desired.

 

[MC3OZ]

Semi 500 Miles at full capacity with full cargo load on level ground, no cargo restrictions compared to a regular class 8 truck.

 

[wipster]

Dang, someone's grumpy.

Let's assume the 2170s going into the Semi are a full MW (original specs, although probably a tad more efficient). Let's also assume the cost of those, at the pack level is $120/kWh (Tesla is probably less, but let's factor in some inflation).

Pack cost = $120,000/kWh minus $45,000/kWh = $75,000 cost for the pack after IRA production tax incentives.

Tesla is selling them for $180,000 (or they were when the prices were up for the Semi - let's assume some commercial buyers have locked that in).

$180,000 - 75,000 = $105,000 to break even. Add in R&D, and cost of the cab itself, and that still leaves a lot of room for a healthy profit margin.


Additionally, beginning Jan 1, 2023, the buyers are eligible for a $40,000 IRA tax break on purchases of the Semi. Given that Tesla has removed pricing from their order page, let's make a reasonable assumption they will plan to raise the price, and capture half of that ($20,000 price increase).


What am I missing here? R&D costs are amortized over multiple years, the facility these are built in is not close to a GF in size, it's relatively cheap to build.

How would this not be a money-printing machine?


(and look at that, far less than 2000 words).


EDIT - I could fully see the Semi being a loss for the first year without the IRA in place. But with it, that's a lot of extra money coming Tesla's way, for a product they had already planned out years ago (and probably planned for it to be reasonably profitable).

And if I remember correctly, they were laughing when they were discussing it. Clear sign it's gonna be bigggggg.