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

[15Peter20]

$5K more at each $5 step down btwn $100 and $60? (DCA).

But first remind us why $60 is the bottom?

Cheers!

Just Numberwang at the moment.

 

[bkp_duke]

So in markets where the IRA doesn't apply, Tesla needs to close the price gap on Chinese LFP.

Raw materials are a minor part of the cost, most of the cost is in refining and manufacturing.

As per EVs, demand global for energy storage batteries is set to grow, the Chinese taking one slice of an expanding market doesn't stop Tesla taking a different slice.

As per cars, I agree Tesla needs to lower energy storage battery prices.

Per Tesla, their current advantage on Megapacks is that the setup is ~80% less when on site than other solutions.

 

[MC3OZ]

Check capex cost for a machine.
Check cycle time.
Check downtime between tool change.
Think through volume required to store completed castings so as to have the necessary castings on hand to keep Z line(s) moving from X casting machines.
Re-think your comment.
I've done the sums before on TMC if you go back and look.

Well how do you think Tesla are going to achieve a 50% cost reduction for Gen3?

Smaller casting machines cost less and probably have a faster cycle time.

Stamping moulds are expensive, as are body shop robots.

Floor space and the number of workers required, is a consideration.

Castings can be stored outside and stacked up.

We need to compare the cost of casting to the cost of what it replaces, and also consider the operational life of the machine and the moulds.

 

[15Peter20]

Maybe Tesla didn't issue a profit warning because they are making money hand over fist! Duh!

Gah! You made me view his ignored content!

 

[elasalle]

Well how do you think Tesla are going to achieve a 50% cost reduction for Gen3?

Smaller casting machines cost less and probably have a faster cycle time.

Stamping moulds are expensive, as are body shop robots.

Floor space and the number of workers required, is a consideration.

Castings can be stored outside and stacked up.

We need to compare the cost of casting to the cost of what it replaces, and also consider the operational life of the machine and the moulds.

start with 100% 4680 battery pack?

 

[generalenthu]

Well, we got ourselves a new volume record for the new year.

 

[Ogre]

Well how do you think Tesla are going to achieve a 50% cost reduction for Gen3?

Smaller casting machines cost less and probably have a faster cycle time.

Stamping moulds are expensive, as are body shop robots.

Floor space and the number of workers required, is a consideration.

Castings can be stored outside and stacked up.

We need to compare the cost of casting to the cost of what it replaces, and also consider the operational life of the machine and the moulds.

I keep going back to Musks comment a while back about stamping out a car like a die cast toy. Those have about 2 parts, top and bottom.

That would be a single bigger press rather than a bunch of small ones.

I’m not going to gamble on which way they go, but I suspect it involves reducing the number of welds/ joints as much as possible. Also reducing the number of points where manual labor is required.

 

[samppa]

For energy storage 4680 LFP is IMO exactly the right play.

4680 should result in the cheapest LFP, the fastest scaling, and it should be very easy to do, with a quick ROI on capex.

Given the IRA, the Biden admin is practically begging Tesla to do it.

4680 LFP in India, initially for energy storage. also something that seems to make sense.

Current lfp packs do not consist of cylindrical cells, they are prismatic. So I don't think there will be much benefit in going to a completely different form factor.

 

[generalenthu]

start with 100% 4680 battery pack?

4680 is not LFP (for now), which is 95% likely what gen 3 will use.

 

[MC3OZ]

start with 100% 4680 battery pack?

That is part of it.

But Tesla implied a rethink of how cars are built and lower capex per car.

If could be that I am wrong and the answer is higher volume factories.

 

[mongo]

I said something similar recently.




Mongo had the weird reply (around page 19315 Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable


and I had to reply

Tesla can say as a clarification after it happens.

I'm not asking for them to post before it happens. I'm asking for them to post later and have the answer on file. I want a link to that answer on the ordering page so there is no confusion once the answer is known.

I'm not sure why people can't see the need to put a clear answer on the tesla website on every page relevant to pricing and rebates.

I should be able to customize an order and see the rebate apply or drop off as I cross a threshold.

Weird? I resemble that remark.
At the time of my post, Treasury had only said they would release non-authoritative guidance by the end of the year and the offical, law triggering, 'proposed guidance' in March (which would be in violation to the Dec 31 deadline, but whatever...).

At this point, a link to the IRS page is the most that seems safe.
After the official guidance in March, yes, definitely, a link to the IRS page and add a 'potentially eligible' indicator on the configurator.

Spoiler: more words

My point was that Tesla can't say what the unreleased guidance will be.
Given the lack of certainty, I prefer them not saying anything so that no one can claim they were mislead.
Even now, will 5 seat Ys ultimately be included? If someone buys a 7 instead due to a notice on tesla.com (based on the current unoffical release) is Tesla liable?
Even sticking with currently eligible 7 seat Ys, what will the credit be once mineral and component requirements replace pack size? When in March will that change be triggered? Would Tesla be liable for a loss of credit if they promote it, but deliver late?

 

[thesmokingman]

Current lfp packs do not consist of cylindrical cells, they are prismatic. So I don't think there will be much benefit in going to a completely different form factor.

This. LFP won't benefit from cylinders. Musk has stated this before, etc etc.

 

[Todd Burch]

I bought 45 more at $108ish.

 

[PeterJA]

I’d like to try to give a little bit of sunshine on a gloomy day:

I was one of those who upgraded to FSD beta on January 1 when it was pushed out to me, despite never having requested it. For a little background I have over 225,000 miles driving Teslas with advanced auto pilot. I used it a lot.

I only have a couple of days and about 200 miles with it so far, but my initial impression is I am impressed, almost to the point of being stunned! Do not make the mistake of judging Tesla‘s progress on this by what auto pilot or even advanced auto pilot is doing currently, it gives a false impression. Watching the car politely pause for pedestrians to cross in front of it, cautiously but smoothly drive around parked UPS trucks sticking out in the street, and just generally driving so well that in most cases an unaware passenger could be fooled (I don’t do this) Into thinking you are driving the car.

Where it falls short is in sometimes being extremely cautious, especially when pulling out where there is limited visibility or not a clearly defined lane. In my opinion this is precisely what it should be doing right now, and intervention usually involves just a light application of throttle similar to grazing the flanks of your horse with spurs just to let them know to keep moving forward. This is something the sped up YouTube videos don’t really accurately portray because it can give the impression of the car moving out with more alacrity than it sometimes does. Once again, I view this overall as beneficial as it this point like a new teenage driver it should err on the side of caution not confidence.

I put in a small buy order with the remaining funds in my Roth IRA at $109 which of course hit fairly quickly this morning. So I guess I’m not just talking the talk.

I partly view this issue from the perspective of someone who has done a lot of rideshare driving. Without going into gory detail, it’s easy to verify with a cursory investigation, the current major companies are still unprofitable and yet are now raising rates on passengers while dramatically increasing their take and further cutting compensation for drivers. I believe the key source of their own unprofitability can also be seen in the drama currently being played out at Twitter post Elon‘s take over where companies located in the greater silicon valley area have developed an extremely wasteful attitude towards running a frugal, efficient business. In short, I would describe them as highly vulnerable even to a new regional player who develops an efficient model that is profitable while still creating an environment for the drivers who render the service the ability to be profitable as well. What a self driving network would to do to them would be self evident.

We are clearly in a new era where a fairly large amount of patience will probably be required. However, I think I understand a little better now where those who have been around since 2012 say this is nothing in comparison. I have an even higher degree of confidence that this company has and is not yet done changing the world.

HODL

And you have FSD Beta 10.x. Version 11 is coming soon.

...Tesla's competition will work to exhaustion, very likely quitting in the process of trying to reproduce what FSD can already do....

Several competitors in self-driving have already quit.

https://www.cnn.com/2022/10/26/business/ford-argo-ai-vw-shut-down/index.html

Even After $100 Billion, Self-Driving Cars Are Going Nowhere

They were supposed to be the future. But prominent detractors—including Anthony Levandowski, who pioneered the industry—are getting louder as the losses get bigger.

www.bloomberg.com

 

[Gigapress]

I keep seeing this meme that Tesla Energy is going to save the day, the year, the decade. Not just from you but all over the place. Especially here on TMC. However much I might wish that to be the case (and as a shareholder I would like it to be so) the actual facts do not support the hopium on offer.

So sorry if this seems like a snippy response to your post, really it is a response to all energy hopium posters.

1. Tesla solar is pants. It is not even a rounding error on global solar sales. It shows no signs of changing nor do I expect it to in this decade, if ever. Approx 315 GW of global solar was installed in 2022, and Tesla will have done barely 0.3 GW of that. Can you say 0.1% very slowly ?

2. Tesla wind is worse. It is to be precise zero. So of the 110-120 GW of wind installed globally in 2022 there is nil attributable to Tesla. Yes, that is 0%. Nada, zilch, rien.

3. The evidence I see suggests that Tesla Energy is slipping in the storage space. In domestic storage it is not even on the leader board for most of the world, excepting USA. In utility storage it has no penetration in China, is market leader in USA, and RoW is very murky but with plenty of non-Tesla wins.

4. For the next several years the vast bulk of the cells going into the storage market will be LFP. That is good, the characteristics of LFP make it a good match for storage. Remind me again how much LFP Tesla makes. Yes, zero. Tesla does not have a competitive advantage in LFP manufacture. The people who do make LFP such as BYD, CATL sell to everyone in the storage market, including themselves for packaging into client-ready cabinets. So the people who are most vertically integrated in this area are BYD et al, not Tesla.

5. Putting LFP into a 4680 form-factor does not seem to be an attractive or high-priority thing for the next few years, if ever. Prismatic LFP is plenty good enough for storage. Load-bearing prismatic LFP seems to be attractive in auto as well, but that's not that relevant in storage (except double-wrapping becomes unnecessary). So by the time that 4680 form-factor reaches LFP (if ever) then there is no reason to think that Tesla will have any particular magic sauce to add.

6. The biggest other cost in storage is the bi-directional power electronics, i.e. the inverter-charger. The vast majority of these in the world are made in China. The Tesla inverters are good, but they are not a game-changer. Again, the vertical integration is better in China than in Tesla, or anybody else in USA or Europe/etc for that matter.

7. The actual products coming out of China increasingly look generic, whether they are at domestic scale or utility scale. This is because there is increasing convergence on a dominant design for each segment's core-product. (The intermediate scale, aka Powerpack, for the commercial segment, looks to be an evolutionary dead end). Now I don't think we are quite at a fully generic dominant design quite yet, but we are getting close. Especially in the domestic market where I increasingly need to check the logo before I can figure out who is actually making/selling product coming out of China. This means that huge numbers of companies who you've never heard of are suddenly making LFP domestic plug'n'play product that is remarkably good quality, fully specced, absolutely functional, low priced, and is really a good deal with not much 'pray' required. That is also increasingly the case in the utility market as well. Take a look at this link and you'll see one (of many) wins of yet another fairly unknown utility-scale supplier of containerised storage by the half-GW, with not a Tesla logo in sight.

China connects 220 MW/440 MWh battery to grid

Robestec has connected a 220 MW/440 MW battery storage system to the grid in Ningxia, China. It is reportedly China’s largest standalone energy storage station, and uses lithium iron phosphate (LFP) batteries from Hithium Energy Storage.

www.pv-magazine.com

8. There is no logistics or cost advantage to Tesla in making this stuff in the USA, whatever the belief system one ascribes to. Roughly speaking one container-load of stuff disappears into one door of a factory and one container of product comes out the other door of the factory. Pretty much that is how it is both by weight and by volume. Irrespective of whether it is a domestic-scale product or a utility-scale product. The two big lumps that go in are the LFP and power electronics; labour and dumb steel get added; software gets loaded; testing happens; and a finished product comes out ready for mass-shipment. The tightest logistics integration comes from doing all this in China. Any claims that somehow doing all this in the USA lead to a better manufacturing/logistics supply chain are plain baloney. That is precisely why storage competitors of Tesla are out there around the world offering more product now than Tesla, even when Tesla's ramp is constrained by chip and cell supply shortages. Believe me, I have some on order from a competitor - the Tesla product is not available, not certified, does not integrate well with the rest of my system(s), and is in any case twice the price for no corresponding gain in functionality.

9. And last time I checked Chinese workforce were a darn sight cheaper than a US workforce, highly motivated, and very productive. In fact there is a Tesla plant or two in Shanghai that prove this (auto, and supercharger) and Shanghai is probably the most expensive place to make stuff in China. The only advantage the USA (and by extension, Tesla-USA) has is a cost one driven by a taxpayer subsidy called the US-IRA and various tariff and non-tariff barriers (whatever happened to free trade ?). That does not seem like a sustainable competitive advantage to me.

10. I've yet to see any evidence of a Tesla software advantage in this sector vs peer competitors. Yes there are advantages vs non-peer competitors, but not vs peer competitors. Want a utility-scale VPP or a trading/operating/metering platform, buy Kraken. Want to integrate with your car and your solar, buy Zappi. Or many other products. The (domestic-level) problem in the software area is not a lack of offerings, or the lack of standards, it is a lack (until now) of a dominant design that allowed this problem-set to be cleaved, and the desire by vendors for lock-in in the scaling wars. My personal guess is that the emergence of dominant designs will soon create the conditions that allow this to be solved. Yes Tesla is a nice little Apple-style walled garden, but don't mistake it for the only game in town. And whilst sometimes walled gardens are very productive Apple-style enterprises, more often they are a fast road to ruin - Yahoo, AOL, DoCoMo, are just some that tried and failed.

11. Look at the actual numbers.

a) If you think that Tesla Energy is going to grow to become 50% of Tesla Incs revenue then either the rest of Inc is going to have an epic fail, or I have to ask how many decades are you looking ahead - best I can figure out is 40% by 2030 with a lot of favourable winds to reach Tesla's own target:

View attachment 891832


b) Ditto profits, well only 37%,
View attachment 891833

c) Ditto cell usage, max 50/50 by 2030:
View attachment 891834


12) So overall if Tesla Energy is going to have a crack at world domination in storage it will need to scale much faster than it has been doing; to recognise that the competition is in fact here; and to spread its plants globally like now, and especially into China and Europe. If it is not careful it will enter the death spiral of Tesla solar and be globally irrelevant.

13) No more hopium please.

Tesla has a 40 GWh per year stationary storage plant coming online right now. That’s about 30-50% more than the entire grid storage industry deployed globally in 2021, depending on whose estimates you look at. This is an order of magnitude jump all at once that will vault Tesla straight to first place in volume with roughly 40% market share if the growth trend for the rest of the industry continues in the next couple years. Tesla has said they chose to allocate scarce chip supply to cars instead of storage. I think they basically decided Megapack was good enough to scale and just went straight for a bigger factory that would be ready to ramp sometime around when the chip constraint was expected to relax. Here we are now and it is ramping. Give it another year and see if we're still concerned about Tesla's market share in storage.

Economies of scale and manufacturing efficiency will matter for battery cost. All else being equal, merely being the biggest producer will lead to lower average unit costs. Since Lathrop is supposedly about to make Tesla the biggest producer by a wide margin, this bodes well for Tesla’s cost competitiveness. We don't have a lot of public data on the Megapack, but we can draw a strong inference from the fact that Tesla says Megapack has a huge backlog extending two years to Q4 ‘24 and growing demand and they decided to build the Lathrop factory to meet it. The length of the backlog and plans for scaling imply that Tesla has roughly 50 GWh worth of orders enqueued. That’s about double the world’s 2021 storage deployments. If Megapack was not a competitive product compared to other offerings on the market then Tesla would not be aggressively scaling like this. I don’t know exactly why so many customers are lined up to buy Megapacks, but apparently Tesla is doing something right.

Tesla is not going to get into wind energy because their position is that solar is the primary winning solution. This was discussed in the first paragraph of the original Master Plan: "...the overarching purpose of Tesla Motors (and the reason I am funding the company) is to help expedite the move from a mine-and-burn hydrocarbon economy towards a solar electric economy, which I believe to be the primary, but not exclusive, sustainable solution." Wind doesn't make practical sense in as many locations as solar and it doesn't have as steep of a cost decline curve, and it's definitely not suitable for being situated within 1-1000 meters distance from the load in most cases like solar is able to do easily. Solar will dominate the renewable energy transition in the long run.

Tesla Solar is not in a death spiral. Total MW deployed is now slightly exceeding the previous peak in 2018 before resources were reallocated to Model 3 and Y. Tesla Solar has problems for sure and it is currently on the order of 0.1% market share globally as you've noted, but mostly it’s still in a transition phase from traditional mounted panels to solar roof tiles. Tesla Solar is, in my view, a long-term bet on the direction the solar industry is headed. I still believe localizing solar close to the load and eliminating most of the cost of transmission is going to win in the long term and so will dual-function solar roof tiles. Tesla has the best solar roof solution on the market. It remains to be seen whether they'll solve the cost and complexity challenges, but I wouldn't count them out just yet. Tesla has been focusing on bigger priorities for the last five years.

I think there's a strong case to be made for housing and light commercial construction, and Tesla's home energy products, to modernize by moving into a proper industrial factory setting. This could drastically cut the soft costs that currently constitute the majority of the total cost structure for residential solar power, batteries, heat pumps, and fancy integrated HVAC systems. The fact that Elon bought one of the first Boxabl prototypes and still has it in his backyard at Starbase is highly suspicious in my opinion. We already know Tesla is partnering with other homebuilders like Brookfield for new construction, and we know Tesla focuses on manufacturing and first principles thinking. I wrote an essay about this here: Tesla Home Energy Products Should Be Installed in a Factory.

Attributing Giga Shanghai’s superior cost efficiency primarily due to the wages and work ethic of the workforce misses other important factors. Fremont is a retrofitted facility originally built in the 1960s that’s now making advanced EVs. The factory layout is heavily compromised and if I remember correctly some of the equipment is still old stuff Tesla inherited from NUMMI. Shanghai was a greenfield project custom-designed by Tesla for Tesla. The flow is clean, the vertical dimension is utilized better, and trucks drop off loads directly beside the line. Shanghai is just a fundamentally more efficient factory and if you magically replaced Fremont with a replica of Shanghai and kept the same Californian workforce, then margins and productivity would increase substantially. It is therefore hard to determine how relevant the difference is to a hypothetical comparison of a Megapack factory in the USA vs one in China. Megapacks also probably can be built with a higher proportion of automation per dollar of revenue than cars because cars have a bunch of manual operations in General Assembly that aren’t necessary for a big battery box. This might make labor a less important consideration for Megapacks compared to cars, diminishing whatever advantage China has in that department.

 

[MC3OZ]

I keep going back to Musks comment a while back about stamping out a car like a die cast toy. Those have about 2 parts, top and bottom.

That would be a single bigger press rather than a bunch of small ones.

I’m not going to gamble on which way they go, but I suspect it involves reducing the number of welds/ joints as much as possible. Also reducing the number of points where manual labor is required.

For a smaller car with a structural battery pack they need smaller front and rear castings.

If they also cast the sides, those are larger castings, but now it is essentially a matter of joining 4 parts, then attaching the battery pack.

 

[petit_bateau]

Well how do you think Tesla are going to achieve a 50% cost reduction for Gen3?

Smaller casting machines cost less and probably have a faster cycle time.

Stamping moulds are expensive, as are body shop robots.

Floor space and the number of workers required, is a consideration.

Castings can be stored outside and stacked up.

We need to compare the cost of casting to the cost of what it replaces, and also consider the operational life of the machine and the moulds.

I don't have any particular insight into Tesla's learning from Cybertruck, which are of course by no means completed.

That may greatly influence the direction the Gen 3 mechanicals take.

As to electrical and control platform, that is obvious: Tesla drivetrain, LFP, equipped for FSD + human-in-the-loop.

 

[PeterJA]

In my opinion ....

- Global domestic clients are voting with their wallets for generic Chinese LFP storage. Except for the sort of client for whom the sticker is important, they buy Tesla. Oh, and some Americans. It is already game over for scale-dominance in this segment and Tesla lost the game.

- Global utility clients are cannier. They will never allow anyone to become scale-dominant. The ones who are susceptible to FUD ("no-one got sacked for buying IBM") or who cannot bring themselves to buy Chinese, buy Tesla. The rest buy Chinese LFP with perhaps a side-order of non-Chinese shrinkwrap splattered with Western badges as camouflage. Tesla is - I think - quite rapidly moving from a wining position to a losing position.

I skim read a lot of other folk out there on the web. Most of them get hung up on trying to figure out a bit of the Tesla storage offer. Mostly they miss the storage competitors. And they don't understand industrial dynamics. And they have no idea re the wider energy game. Plus they don't understand well how this meshes with automotive. So they've zeroed in on one leaf on one tree in one wood inside one forest on one continent in a big planet. Plus other stuff. That chap(ess) you linked to is at leaf-level.

After 30-years in the energy game; from oil well to inverter; wind to solar; hydro to battery; 12V to 1.2 million volts; classroom to factory to field .... I try to put things in context.

But that chap(ess) is still saying things that are worth reading, not ignoring.

(And if Tesla think I'm wrong, and wish the market to understand that, then Tesla should publish regular, reliable facts about both their own business and the global market; and answer properly posed questions from knowledgeable people.)

Are Tesla Energy competitors doing virtual power plants like Tesla is? Do they offer an equivalent to Tesla's Autobidder software?

Tesla virtual power plant is rocketing up, reaches 50 MW

Tesla's virtual power plant in California is growing at a fast pace with more Powerwall owners joining. It can now output up to 50 MW.

electrek.co

Tesla quietly built a virtual power plant in Japan

It can feed power back into the grid.

www.theverge.com

Tesla to make 'virtual power plant' with some TX Powerwall users

Tesla Powerwall owners are all part of the company's big plan.

www.mysanantonio.com

 

[thesmokingman]

I keep going back to Musks comment a while back about stamping out a car like a die cast toy. Those have about 2 parts, top and bottom.

That would be a single bigger press rather than a bunch of small ones.

I’m not going to gamble on which way they go, but I suspect it involves reducing the number of welds/ joints as much as possible. Also reducing the number of points where manual labor is required.

OPTIMUS...

I can't imagine them being able to lower costs any further considering they already have the lowest cost possible. But robots!

 

[Skipdd]

Per Tesla, their current advantage on Megapacks is that the setup is ~80% less when on site than other solutions.

Still weighing and measuring the pros and cons for believing that Megapacks will have a meaningful impact on 2023 and beyond earnings. I get that order books are full based on the ordering page - that’s clearly a positive. But when I read @petit_bateau ’s posts (and he has the experience and knowledge to be considered a SME), it’s clear the data point to Tesla having a niche play here, which is concerning. Not sure the advantage that Tesla is touting carries the weight they ascribe to it. Maybe they need to apply some “first principles engineering.”

I’m definitely looking for positives, including beyond automotive. I do feel like I need to question things more. You know the saying, “pain is a good teacher..”. FWIW, while our investment is significant and we are down ~60%, it is a relatively small part of our overall portfolio. As it should be given our age and proximity to retiring. We have delayed that another 6-18 months to the extent we control such. But if what is a paper loss became a real loss, it would be very unpleasant. I have DCA’d a bit to lower our cost basis, but am holding for now.