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

[Discoducky]

James Stephenson always aims to be conservative with his published numbers, too, so this should be viewed as a fairly pessimistic estimate.

Looking at his whole Twitter thread, he's projecting for 2023:

• $36 billion of GAAP net income
  • This would be an all-time world record for highest car company profits by a large margin
  • The best 12-month period Toyota (including Lexus and rest of company) has ever had was $29B earnings (source), and that was an anomalous rebound in 2021 from bad results in 2020. Normally Toyota was making about $20B per year pre-COVID.
• 2.2 million vehicles sold
• 31-32% gross margin excluding FSD, leasing, etc.
  • Up only slightly from 30.6% actual in Q1 '22
• Average cost per car (excl. leasing) of $43k by Q4 '23
  • Up 13% from $38k actual in Q1 '22
• $20k average gross profit per vehicle by Q4 '23
  • Up $3.7k from $16.3k actual in Q1 '22

Especially, those delivery and gross margin estimates are very conservative.

Mr. Stephenson forecasts Shanghai production of just 982k cars in 2023. As of now, Shanghai is at 864k annualized production rate based on 72k produced in June. Thus, James is projecting merely 14% growth for Shanghai between now and then, which is ludicrously low considering that they're making major upgrades soon and they have a whole Phase Two area coming online later this year that should basically double production capacity once ramped up. Before the government lockdowns, Giga Shanghai historically was growing output much faster than what James is predicting, as shown below. James is predicting average monthly output in 2023 of 81k, whereas linear extrapolation of the trend resuming now after a first-half of 2022 disruption would lead to a 2023 estimate more like 120k per month or over 1.4M for the year.

If Shanghai were to sell half a million cars more than James' model shows, using his estimate of $20k profit per car means adding an extra $10B to income, for a new GAAP total of $46B.

View attachment 824718

The Berlin and Austin projections are 340k cars and 395k respectively. That's less conservative than the Shanghai number, but still very cautious. At the Giga Berlin Grunheide County Fair event in October 2021, Elon stated at 21:25 in the video:



So basically at the time, Elon said that Giga Berlin is targeting a 250k to 500k annualized production rate by the end of 2022. I don't recall guidance having been changed since then. If this is so, then they will produce far more than 340k in 2023, and likewise Austin would probably beat James' published estimates. If they produce 1 million combined, this adds 265k cars to the total. If we again assume $20k per car, this adds another $5.3B to net income. However, these factories should have the highest margins due to using 4680s and front castings, only making Model Y, etc, so really $23k is more reasonable if we still are sticking with James' overall gross margin estimates. So then it's an incremental $6B in earnings. Add that to the Shanghai adjustment from before and now we're at $52B total.

We need to account for tax on the incremental earnings added in this analysis, which would pull approximately $2B off the total. $50B total net income.

Higher margins would multiply this result even higher. James' model is estimating just $3.7k increase in profit per car even with all the price hikes; more Model Y, S & X in the mix; front casting; less shipping and tariff expenses; newer, more efficient factory designs; structural pack and 4680s; dilution of Fremont contributions vs Shanghai/Berlin/Austin; and less struggling with chip problems. If we model with the new total delivery estimate just shy of 3M cars and add an extra $3k to gross profit per car, that's another $9B for profit, or maybe $8B after tax. Now the adjusted GAAP income estimate is $58 billion, which would be $49 GAAP earnings per share fully diluted 1.18 billion outstanding shares projected by the end of 2023.

$49 GAAP EPS. Roughly $70 annualized for Q4 '23. All of this excluding FSD, leasing, etc.

Your posts continue to astonish at the level of detailed analysis.

Thank you!

 

[Gigapress]

These margins will not last. The Chinese are coming, from the Electric Viking:

Will China be prevented from competing in some of our markets? If so, then maybe our margins will survive somewhat longer?

Over the long run, autos will act like commodities. Whoever has the lowest cost of production will win. That's why Elon says the car business is not where Tesla's value lies.

We need to win real-world AI.

That car is 16% less in length than Model Y and 18% narrower, with a less luxurious interior. It has a 400 km range using the very optimistic Chinese rating method, vs 545 km for the SR Model Y per the Tesla China order page. It's slower than the Y and probably doesn't handle as well.

Cheap EVs have been for sale in China for several years but Tesla is still crushing the market there with no end in sight. For example, the Xpeng G3 is half the price of the Y.

The big thing is we have no idea what margins Geely is getting on this car. That $14k price tag could be at a loss for all we know.

 

[Knightshade]

James Stephenson always aims to be conservative with his published numbers, too

That is... a relatively recent change.... from James His June 2020 #s for example (so this was not pre-covid) were let us just say overly optimistic on production ramps... He had 2021 deliveries projected at ~1.1 million for example...including almost 50,000 Semis he expected to start shipping Q1 2021... and Cybertruck shipping Q3 2021. He had Austin and Berlin both shipping cars Q1 2021 with Berlin shipping almost 100k vehicles in 2021 and Austin doing 145k in 2021.

He even had the new compact model that still doesn't exist shipping by Q4 2021 out of Shanghai.


I appreciate he's dialed things back a bit since then, and I do find his stuff very interesting and useful to read these days.




EDIT-
Ah, found a note I'd written when his 2021 forecast came out... he definitely got more conservative compared to 2020... his June 2021 slides knocked deliveries back almost 200k (meaning he adjusted, halfway through the year, to land on nearly the right number versus the way high one he'd predicted 12 months earlier).... he STILL had both CT and Semi delivering in 2021 though (albeit not till Q4)...he dropped the "imaginary compact" model entirely.... but he was still way optimistic on Berlin (Q3 2021 deliveries with almost 15k in Q4 2021)...and Austin producing in Q4 2021 though "only" 2k.

He also had very optimistic #s for Model X, 6 months after it was clear there were refresh issues.

His June 2021 numbers DID have projections for 22 and 23 though-

https://twitter.com/x/status/1404124757688193028

Q1 2022 he got pretty solidly, Q2 obviously less so He has Q4 this near 500k which I think is still a bit optimistic.



EDIT2-


Oh, and checking his just-posted June 2022 predictions, he dialed back again from his '21 ones.

Instead of 500k Q4 he's nearer 400k, with Q3 around 375ish. He's got 500k showing Q1 23 now.


So yeah looks like he's generally overly optimistic about future next-year stuff (though not quite as badly as he used to be) and then dials it back halfway through that year when he realizes they're not gonna remotely hit his original estimates.

https://twitter.com/x/status/1543650561114923012

 

[JustMe]

Whenever I see this guy on a YouTube video or a tweet of his shared, whether it’s P&D or quarterly earnings, it’s always overly optimistic.

Does anyone have a track record of this guys accuracy?

Perhaps you shouldn’t assert that James is always overly optimistic if you haven’t bothered to check his track record.

 

[bkp_duke]

Not sure how Gary is getting that number. Especially listing an exact number like that.

Would love for it to be true, but if it was, I have no clue as to why Tesla wouldn't have included something like "June's deliveries topped 157,000 for June, at a annual run rate of 1.89 million vehicles now" in the June P/D report besides just saying "June was highest production month for Tesla ever"

As for tomorrow, I'm seeing an extreme amount of FUD and wouldn't be surprised if there's a major bear raid. In fact there's a nasty FUD filled article on the Berlin 2 week shutdown "

Musk forced to halt Tesla assembly line in struggling Berlin plant for 2 weeks over production problems"​

Including this "German tabloid Bild reported on Monday the factory has been struggling with a trio of persistent problems: longer-than-anticipated manufacturing time, a lack of skilled workers and quality assurance issues that require expensive reworking carried out on vehicles that had already rolled off the assembly line."

Oh well. See everyone in the 500's.

Because I'm pretty sure Tesla has never listed monthly run rates in any financial reports. They just have always been intentionally opaque.

 

[Prunesquallor]

Holy hell Honda…

Nice to see McLaren dong well.

 

[Gigapress]

That is... a relatively recent change.... from James His June 2020 #s for example (so this was not pre-covid) were let us just say overly optimistic on production ramps... He had 2021 deliveries projected at ~1.1 million for example...including almost 50,000 Semis he expected to start shipping Q1 2021... and Cybertruck shipping Q3 2021. He had Austin and Berlin both shipping cars Q1 2021 with Berlin shipping almost 100k vehicles in 2021 and Austin doing 145k in 2021.

He even had the new compact model that still doesn't exist shipping by Q4 2021 out of Shanghai.


I appreciate he's dialed things back a bit since then, and I do find his stuff very interesting and useful to read these days.

Good point. I only started following James around 2020.

 

[Englishman]

Higher margins would multiply this result even higher. James' model is estimating just $3.7k increase in profit per car even with all the price hikes; more Model Y, S & X in the mix; front casting; less shipping and tariff expenses; newer, more efficient factory designs; structural pack and 4680s; dilution of Fremont contributions vs Shanghai/Berlin/Austin; and less struggling with chip problems. If we model with the new total delivery estimate just shy of 3M cars and add an extra $3k to gross profit per car, that's another $9B for profit, or maybe $8B after tax. Now the adjusted GAAP income estimate is $58 billion, which would be $49 GAAP earnings per share fully diluted 1.18 billion outstanding shares projected by the end of 2023.

$49 GAAP EPS. Roughly $70 annualized for Q4 '23. All of this excluding FSD, leasing, etc. Still with upside on deliveries and margins if the ramps go well.

We all know that there is no demand problem for Tesla. Loving the analysis above, however, just a small part of me worries... 3M Tesla's sold next year. I bloody hope so, but I guess I find it hard to quantify...

Rough figures (that I've guessed)
7,000,000,000 people in the world
3,000,000,000 of driving age
40% car ownership
1,200,000,000
Changing car every 10 years therefore /10
120,000,000

Got this far then Googled "Number of cars sold worldwide" Thanks Statista

So, in 2021 it was 66,700,000 (I was getting there!!)

3M is 4.5%

Starting to think, actually this isn't so crazy after all.

Does make you think about 2030 - 20m. However, if Tesla goes the way of iPhone (which I think is quite likely), you will either own a Tesla/another. Robotaxi etc could completely disrupt car ownership however TaaS (transport) revenue/margins would more then compensate a loss of car sales.

 

[Gigapress]

We all know that there is no demand problem for Tesla. Loving the analysis above, however, just a small part of me worries... 3M Tesla's sold next year. I bloody hope so, but I guess I find it hard to quantify...

Rough figures (that I've guessed)
7,000,000,000 people in the world
3,000,000,000 of driving age
40% car ownership
1,200,000,000
Changing car every 10 years therefore /10
120,000,000

Got this far then Googled "Number of cars sold worldwide" Thanks Statista

So, in 2021 it was 66,700,000 (I was getting there!!)

3M is 4.5%

Starting to think, actually this isn't so crazy after all.

Does make you think about 2030 - 20m. However, if Tesla goes the way of iPhone (which I think is quite likely), you will either own a Tesla/another. Robotaxi etc could completely disrupt car ownership however TaaS (transport) revenue/margins would more then compensate a loss of car sales.

3 million deliveries is pretty reasonable. It's still less than most of the bigger car companies.

For comparison, in 2019 before the supply chain issues hit:

Daimler: 3.3 million​

Fiat Chrysler: 4.4​

Hyundai: 4.4​

Honda: 5.3​

Ford: 5.4​

Nissan: 5.5​

General Motors (incl. SAIC): 7.7​

Toyota: 10.7​

Volkswagen: 11​

 

[bkp_duke]

3 million deliveries is pretty reasonable. It's still less than most of the bigger car companies.

For comparison, in 2019 before the supply chain issues hit:

Daimler: 3.3 million​

Fiat Chrysler: 4.4​

Hyundai: 4.4​

Honda: 5.3​

Ford: 5.4​

Nissan: 5.5​

General Motors (incl. SAIC): 7.7​

Toyota: 10.7​

Volkswagen: 11​

I salivate at the prospect of Tesla being in the Toyota/VW range . . . huge $$$$.

 

[Gigapress]

I salivate at the prospect of Tesla being in the Toyota/VW range . . . huge $$$$.

Exactly. We're talking about smashing all-time record car company profits next year for Tesla, yet with global unit sales volume that puts them in 10th place.

 

[oldTAVguy]

He got this from this source :View attachment 824724

With Q1 at 310 048 deliveries and Q2 at 254 695. If we substract to Q1 + Q2 deliveries the sales from this source of Janv - May, we get june deliveries of 157 874.
But it doesn't mean that was the production rate of june, cars may have been produced earlier in the quarter.

Found this in the comments from Disqus Profile - flipsivad in the article: US: GM Delivered Over 7,000 Plug-In Vehicles In Q2 2022

"The Hong Guang is already selling at a rate of 200k as of now so that by 2025 it might be almost half of GM's global target.

"One of the biggest appeals of the little EV is its price, as it sells for the equivalent of just over $4,000 USD. This, however, means that there isn’t much room for profit. In fact, China’s Xcar reports that GM and Wuling make a profit of 89 Yuan on each example they sell. At today’s exchange rates, that’s the equivalent of just $13.73 USD."

So GM is pocketing $6-$7 per unit sold! LOL. If they can increase sales to 500,000 Hong Guangs by 2025 then GM will pocket a cool $3 mil on the entire JV! That might be enough to pay for at least Mary's base salary."

How much truth is there in the $6-$7 per unit profit for each car for GM?

 

[The Accountant]

Whenever I see this guy on a YouTube video or a tweet of his shared, whether it’s P&D or quarterly earnings, it’s always overly optimistic.

Does anyone have a track record of this guys accuracy?

Here are the accuracy numbers for the forecasters who back up their estimates with details.
I have a very small sample size here; just including three quarters is unfair but I didn't want to do the work to go back further in time.
Over the past 3 Qtrs, Dave Lee has the best forecasts with an average variance of 7.9% (helped by that perfect Q3 2021 prediction).
James has the highest variance at 15% but his Q4 2021 forecast hurts his overall accuracy variance. He is not always bullish. He came in low twice and too high once.

 

[insaneoctane]

I salivate at the prospect of Tesla being in the Toyota/VW range . . . huge $$$$.

Yes, but to be fair, Toyota /VW are not exclusively selling $60K+ vehicles...

 

[thesmokingman]

Cathie Wood offers some insight on what's to come. She put forth a well reasoned argument for the economy being at a reversal point.

I particularly like the way she presents a plethora of indicators and then explains their interactions and what this means for the future. Cathie does a good job of analysis of a very complex system and this presentation left me feeling pretty good about things economic going forward.

Interesting, even Cathie is citing the chips sector ala Nvidia GPU pricing in the secondary markets and primary market with msrp becoming a reality after two years. It echos what sentiment in the chips sector a few days back.

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

In one of his videos, Ingineerix claimed that V2G ain't happening based off deliberate designs of the components. Later, Phil commented that this is on purpose as Tesla wants to sell Power Walls. Yeah, that's based on using in-car inverters.... I agree based on Phil's tear-down and analysis...

teslamotorsclub.com

 

[Knightshade]

Here are the accuracy numbers for the forecasters who back up their estimates with details.
I have a very small sample size here; just including three quarters is unfair but I didn't want to do the work to go back further in time.
Over the past 3 Qtrs, Dave Lee has the best forecasts with an average variance of 7.9% (helped by that perfect Q3 2021 prediction).
James has the highest variance at 15% but his Q4 2021 forecast hurts his overall accuracy variance. He is not always bullish. He came in low twice and too high once.

View attachment 824763

FWIW I think if you're asking "how bullish or not" you have to narrow your criteria... yours here is EPS, which is certainly a pretty good one... but the way they get it wrong can matter too depending how you're trying to use their projections.

For example James seems to consistently be overly optimistic on deliveries (and new model/factory ramps) for following years- which he then ends up having to dial back every June as we're halfway through the year and clearly not reaching the targets he predicted-- while being more conservative on say profit margins.


I agree how close on EPS is probably the gold standard if you're gonna pick a single thing to check-- but if I were less lazy I'd be interested in if there's a consistency to how they miss.... (and even in the EPS data there's something interesting... for example Gary Black from the limited sample has never really missed too high.... everyone else appears to have been too low for Q1 22, too high for Q1 21 (Except Gary which I'll call 0.01 off a wash), and then Q3 21 2 people high, 2 people low, 1 dead on.

3 million deliveries is pretty reasonable. It's still less than most of the bigger car companies.

I find 3 million in 2023, with existing models plus CT delivering in some fair #, to be perfectly reasonable... but after seeing how S/X refresh went I don't think a fast CT ramp is a guaranteed slam dunk until I see it happen.

That said, the only way they get to 20 million (or even 10) is either RT, or more models than are currently announced- and yeah, in particular, a smaller/cheaper one. I still don't love how heavily Elons thumb seems to be on the scale of that choice toward RT.

Interesting, even Cathie is citing the chips sector ala Nvidia GPU pricing in the secondary markets and primary market with msrp becoming a reality after two years. It echos what sentiment in the chips sector a few days back.

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

In one of his videos, Ingineerix claimed that V2G ain't happening based off deliberate designs of the components. Later, Phil commented that this is on purpose as Tesla wants to sell Power Walls. Yeah, that's based on using in-car inverters.... I agree based on Phil's tear-down and analysis...

teslamotorsclub.com

Without getting in to the Cathie of it all-- there's also the fact the secondary market is being flooded with used beat-to-heck GPUs from miners who can no longer use them profitably because of the crypto crash .... and the primary market by both miners no longer scooping up normal supply and vendors realizing next-gen cards come out in a couple months and they don't want to be stuck with inventory of the old stuff.

So I don't find that bit nearly as compelling evidence of "CHIP SHORTAGE OVER" as she might.

 

[The Accountant]

FWIW I think if you're asking "how bullish or not" you have to narrow your criteria... yours here is EPS, which is certainly a pretty good one... but the way they get it wrong can matter too depending how you're trying to use their projections.

For example James seems to consistently be overly optimistic on deliveries (and new model/factory ramps) for following years- which he then ends up having to dial back every June as we're halfway through the year and clearly not reaching the targets he predicted-- while being more conservative on say profit margins.

Fair points. I would also add that I believe James tends to be more bullish on annual and multi-year forecasts than his quarterly forecasts where he does not have to look out too far.

 

[mongo]

James with $32 non-GAAP EPS for 2023. Oh My!

imo, James Stephenson is one of the best (maybe the best) Tesla forecasters on social media.
He shared his Q2 numbers and in doing so, gave us a peak at his 2023 numbers. You have to add up the 2023 qtrs . . .and when you do, it's $32.

https://twitter.com/x/status/1543672493403250688

Any idea what the Q4 2023 tax item is?

 

[Gigapress]

To put in context more just how low 2.2 million deliveries for 2023 would be, also remember that Elon guided a few months ago for 70-80% growth annually and maybe higher than that in the future.

2021 was 0.95M deliveries.

Two years of 70% growth would be 2.75M in 2023.

Two years of 80% growth from 0.95M gets us to 3.1M.

The actual recent trend has been 90% annual growth.

 

[Stretch2727]

Found this in the comments from Disqus Profile - flipsivad in the article: US: GM Delivered Over 7,000 Plug-In Vehicles In Q2 2022

"The Hong Guang is already selling at a rate of 200k as of now so that by 2025 it might be almost half of GM's global target.

"One of the biggest appeals of the little EV is its price, as it sells for the equivalent of just over $4,000 USD. This, however, means that there isn’t much room for profit. In fact, China’s Xcar reports that GM and Wuling make a profit of 89 Yuan on each example they sell. At today’s exchange rates, that’s the equivalent of just $13.73 USD."

So GM is pocketing $6-$7 per unit sold! LOL. If they can increase sales to 500,000 Hong Guangs by 2025 then GM will pocket a cool $3 mil on the entire JV! That might be enough to pay for at least Mary's base salary."

How much truth is there in the $6-$7 per unit profit for each car for GM?

This is why many people here have brought up that this car just should not count. But it is an EV.

I am not sure why GM gets credit for it when they only own 44% of the joint venture.

Maybe a better way to look at market share is to look at $$ rather than units. I am sure Tesla is way above 17 percent.