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

[Skryll]

To me, it sounds even worse - that is, even sketchier reasoning. The Reuters line refers not to US sourcing costs, but that the RULES "would hike the costs". The collective understanding is that any rules emplaced since CATL's decision to build here have lowered, not hiked, the costs.

Maybe competing for raw materials will be harder with more subsidized players

 

[Artful Dodger]

He would probably have bought is initial 9% from this but waht else does he have?

Elon almost certainly paid off all his margin loans from the early Nov '21 equity proceeds. Call that -$500M (and like everything else we have no certainy, only crude estimates).

 

[Artful Dodger]

That seems incredibly short sighted on the part of the Austin plant. New colors to the US would be extremely welcome.

Joe Tegmayer showed drone video this week of two large (trailer-sized) green-wrapped crates being delivered to the North end of Giga Texas, need the Paint Shop. I suspect those are for new lines. So there may be some extra features coming, but again, Osborne is a thing...

 

[wipster]

To me, it sounds even worse - that is, even sketchier reasoning. The Reuters line refers not to US sourcing costs, but that the RULES "would hike the costs". The collective understanding is that any rules emplaced since CATL's decision to build here have lowered, not hiked, the costs.

My apologies, should have included the comment that was before the article which stated it was the Chinese government that was forcing CATL to abandon those plants...

May just be Hamel's opinion, but it sounds viable.

 

[Xepa777]

Joe Tegmayer showed drone video this week of two large (trailer-sized) green-wrapped crates being delivered to the North end of Giga Texas, need the Paint Shop. I suspect those are for new lines. So there may be some extra features coming, but again, Osborne is a thing...

I feel this is why Elon keeps calling it a robotaxi. It’s really the model 2, but he jus r doesn’t want to osbourne sales right now. Depending on the progress of FSD beta they can choose to deploy the autonomous version of the car or passenger version, with so modular interchangeable parts.

Martin V sort of hinted at this in his analyst meetings and Elon all but spelled it out in the earnings call. Also let’s not forget China’s desire to not just be a production plant but to show off a Chinese-inspired/designed car of their own making for the world, an unfulfilled Elon goal (and a big reason why China will still be a backer of tesla for the forseeable future regardless of what FUD articles come out).

 

[Unpilot]

The shade thrown at institutional analysts (with the exception of a few such as Alex Potter and Pierre Ferragu) is completely fair, because their estimation methodology completely sucks. Personally, I will explicitly bash their estimates. They have been very wrong in general and will continue to be very wrong.

Over and over again, institutional analysts have underestimated Tesla's performance, especially over the long haul but also on a quarterly level, as shown in the following table and charts.

The quarterly data uses institutional analyst consensus EPS expectations compiled by Zack's for each quarter since 2019. The average quarterly EPS prediction error was -25% of the actual EPS, with a cumulative error of $1.49 since 2019. This track record is especially egregious considering that all of these estimates were published immediately before each earnings release, after the production and delivery numbers were known, which (if we give analysts the benefit of the doubt regarding their good intentions) can only indicate that analysts have been consistently failing to understand the factors affecting Tesla's revenue and cost.

The annual forward-looking estimates for both vehicle production and earnings are far worse. I tried to find compiled data taken in December for each subsequent year, but couldn't find it. Instead, I resorted to looking up a few reports published by individual analysts.

If analysts had a 50/50 chance of guessing quarterly numbers too high or too low, the odds of guessing low on at least 12 of the last 13 quarters due to bad luck is 1/binom.dist(1,13,true) --> 585 to 1. Moreover, for the single instance in Q4 2020 when they estimated too high, it was only by 6% and that was mainly due to a $100M interest expense penalty that Tesla voluntarily paid to settle their convertible notes early. Without this unexpected one-time expense, the analyst consensus would have once again been 20% too low. If we want to count that as another miss, then that's 13 consecutive quarters of underestimation which has a 1 in 8,192 chance of happening randomly.

These results, along with the context of all the misunderstandings that apparently led analysts to their erroneous estimates, conclusively and unambiguously indicate systematic bias.

I would bet my life savings that their estimates for 2023, 2024 and beyond also are much too low just like all the previous estimates. Oh wait...I have bet my life savings on that prediction.

Quarter	Non-GAAP Earnings per Share	Estimated	Error	% Error
Q1 2019	-$0.19	-$0.08	$0.11	58%
Q2 2019	-$0.07	-$0.04	$0.04	52%
Q3 2019	$0.12	-$0.01	-$0.13	-108%
Q4 2019	$0.14	$0.11	-$0.03	-24%
Q1 2020	$0.08	-$0.01	-$0.10	-118%
Q2 2020	$0.15	-$0.04	-$0.19	-129%
Q3 2020	$0.25	$0.18	-$0.07	-29%
Q4 2020	$0.27	$0.28	$0.02	6%
Q1 2021	$0.31	$0.26	-$0.05	-15%
Q2 2021	$0.48	$0.30	-$0.18	-38%
Q3 2021	$0.62	$0.45	-$0.17	-28%
Q4 2021	$0.85	$0.67	-$0.17	-20%
Q1 2022	$1.08	$0.72	-$0.36	-33%
Q2 2022	$0.76	$0.64	-$0.12	-16%
Q3 2022	$1.05	$0.97	-$0.08	-8%
Totals / % Error Average	$5.89	$4.40	-$1.49	-25%

View attachment 866224View attachment 866225

View attachment 866239
There were 1.123 billion outstanding shares (diluted) at the time, so this translates to non-GAAP net income estimates of $10.0B and $11.8B in '22 and '23 respectively. '22 actual is already $9.3B just in the first 3 quarters and it's likely to be around $15-16B after Q4 gets added. This is tracking for a 50-60% beat despite Shanghai shutdowns and a highly unanticipated war in Europe. $11.8B for '23 is just...I have no words. Even Fremont and Shanghai by themselves will probably earn double or triple that in '23.

For another example, in May 2019 Joseph Spak from RBC Capital provided a price target equivalent to a ~$50B market cap and:

Actual non-GAAP net income results:
'19: $0.04B
'20: $2.46B
'21: $7.64B

Let's bear in mind this estimate was published before anyone had heard of COVID-19, so nobody was anticipating factory lockdowns and major supply chain chaos.

There are only two respectable ways for someone to respond to inaccurate estimation:

1. Acknowledge the problem, take responsibility and learn how to fix the model
   • (Note: this should be happening every time because all of our estimates are wrong and all we can hope for is to make our models less wrong over time)
2. Cease publication of authoritative-sounding "expert" estimates until such time that #1 has been accomplished

I can't help but notice that institutional analysts have selected neither of these options, which means that either they are incompetent at this particular task or are corrupt liars, or both.

I love you

 

[betstarship]

Wow, things are spicy today

I double down on this comment after that last post and say things are really spicy today.

 

[Artful Dodger]

As the number of models grow, perhaps iM2S3XY4...

... my Cat. Obviously.



Cheers!

 

[Singuy]

The shade thrown at institutional analysts (with the exception of a few such as Alex Potter and Pierre Ferragu) is completely fair, because their estimation methodology completely sucks. Personally, I will explicitly bash their estimates. They have been very wrong in general and will continue to be very wrong.

Over and over again, institutional analysts have underestimated Tesla's performance, especially over the long haul but also on a quarterly level, as shown in the following table and charts.

The quarterly data uses institutional analyst consensus EPS expectations compiled by Zack's for each quarter since 2019. The average quarterly EPS prediction error was -25% of the actual EPS, with a cumulative error of $1.49 since 2019. This track record is especially egregious considering that all of these estimates were published immediately before each earnings release, after the production and delivery numbers were known, which (if we give analysts the benefit of the doubt regarding their good intentions) can only indicate that analysts have been consistently failing to understand the factors affecting Tesla's revenue and cost.

The annual forward-looking estimates for both vehicle production and earnings are far worse. I tried to find compiled data taken in December for each subsequent year, but couldn't find it. Instead, I resorted to looking up a few reports published by individual analysts.

If analysts had a 50/50 chance of guessing quarterly numbers too high or too low, the odds of guessing low on at least 12 of the last 13 quarters due to bad luck is 1/binom.dist(1,13,true) --> 585 to 1. Moreover, for the single instance in Q4 2020 when they estimated too high, it was only by 6% and that was mainly due to a $100M interest expense penalty that Tesla voluntarily paid to settle their convertible notes early. Without this unexpected one-time expense, the analyst consensus would have once again been 20% too low. If we want to count that as another miss, then that's 13 consecutive quarters of underestimation which has a 1 in 8,192 chance of happening randomly.

These results, along with the context of all the misunderstandings that apparently led analysts to their erroneous estimates, conclusively and unambiguously indicate systematic bias.

I would bet my life savings that their estimates for 2023, 2024 and beyond also are much too low just like all the previous estimates. Oh wait...I have bet my life savings on that prediction.

Quarter	Non-GAAP Earnings per Share	Estimated	Error	% Error
Q1 2019	-$0.19	-$0.08	$0.11	58%
Q2 2019	-$0.07	-$0.04	$0.04	52%
Q3 2019	$0.12	-$0.01	-$0.13	-108%
Q4 2019	$0.14	$0.11	-$0.03	-24%
Q1 2020	$0.08	-$0.01	-$0.10	-118%
Q2 2020	$0.15	-$0.04	-$0.19	-129%
Q3 2020	$0.25	$0.18	-$0.07	-29%
Q4 2020	$0.27	$0.28	$0.02	6%
Q1 2021	$0.31	$0.26	-$0.05	-15%
Q2 2021	$0.48	$0.30	-$0.18	-38%
Q3 2021	$0.62	$0.45	-$0.17	-28%
Q4 2021	$0.85	$0.67	-$0.17	-20%
Q1 2022	$1.08	$0.72	-$0.36	-33%
Q2 2022	$0.76	$0.64	-$0.12	-16%
Q3 2022	$1.05	$0.97	-$0.08	-8%
Totals / % Error Average	$5.89	$4.40	-$1.49	-25%

View attachment 866224View attachment 866225

View attachment 866239
There were 1.123 billion outstanding shares (diluted) at the time, so this translates to non-GAAP net income estimates of $10.0B and $11.8B in '22 and '23 respectively. '22 actual is already $9.3B just in the first 3 quarters and it's likely to be around $15-16B after Q4 gets added. This is tracking for a 50-60% beat despite Shanghai shutdowns and a highly unanticipated war in Europe. $11.8B for '23 is just...I have no words. Even Fremont and Shanghai by themselves will probably earn double or triple that in '23.

For another example, in May 2019 Joseph Spak from RBC Capital provided a price target equivalent to a ~$50B market cap and:

Actual non-GAAP net income results:
'19: $0.04B
'20: $2.46B
'21: $7.64B

Let's bear in mind this estimate was published before anyone had heard of COVID-19, so nobody was anticipating factory lockdowns and major supply chain chaos.

There are only two respectable ways for someone to respond to inaccurate estimation:

1. Acknowledge the problem, take responsibility and learn how to fix the model
   • (Note: this should be happening every time because all of our estimates are wrong and all we can hope for is to make our models less wrong over time)
2. Cease publication of authoritative-sounding "expert" estimates until such time that #1 has been accomplished

I can't help but notice that institutional analysts have selected neither of these options, which means that either they are incompetent at this particular task or are corrupt liars, or both.

Ha, ironically if there's one thing you and Gordon can agree on is this. In his mind, he feels that analysts are purposely estimating too low to generate an artificial beat for Tesla.

 

[Unpilot]

My thoughts on analysts and being wrong. (not that anyone cares ...and I have had a few mid-strength beers)....however...

Being wrong is part of the job...no one is always right....however being consistently wrong (in only one direction)to the extent @Gigapress has pointed out seems....suspect. Throw in the belief of many here (certainly mine) that Wall Street is a den of thieves, and one gets the distinct odor of crooks at work.

I am a big fan of Dan Carlin Product categories Common Sense with Dan Carlin

He had an episode on the experts that the MSM trots out whenever there is a war and wonders why they keep getting airtime when they have been so wrong over and over again.

I wonder the same thing about these analysts.

 

[ZeApelido]

https://twitter.com/x/status/1583322701938053120

Why is the floor elevated off the ground? Was it pre-fabed then dropped in place same as they do the charging stalls and cabinets?

Since the homeless of Santa Monica will undoubtedly inhabit this place, it's so the urine doesn't leak in.

First principles.

 

[wipster]

... my Cat. Obviously.

View attachment 866367

Cheers!

Or the name of Elon's next child...

 

[MitchMitch]

Elaine?

Nah, Doris! (Saphron)

And for those playing 4D, I Bet (tte Midler) the Get Shorty (ies) reference is what make it topical??

 

[MP3Mike]

To me, it sounds even worse - that is, even sketchier reasoning. The Reuters line refers not to US sourcing costs, but that the RULES "would hike the costs". The collective understanding is that any rules emplaced since CATL's decision to build here have lowered, not hiked, the costs.

Yep. It might make it more expensive if you have to compete for limited local minerals, but it wouldn't stop you from importing them as was probably planned. It might even make them cheaper to import as demand is reduced from other places going for local minerals.

 

[Artful Dodger]

I don't think the color change or option is a demand lever. It was planned and I believe was mentioned years ago when GF Berlin was going up.

Yeah, not a "lever". Its a "ramp".

 

[Ogre]

Yep. It might make it more expensive if you have to compete for limited local minerals, but it wouldn't stop you from importing them as was probably planned. It might even make them cheaper to import as demand is reduced from other places going for local minerals.

They would have to build out a parallel supply chain to qualify for any US incentives.

I wonder how this is going to impact GM/ Ford who rely on their Asian partners for all the heavy lifting.

 

[Artful Dodger]

@Mengy described Gary Black as a mild bull. Thought I would have a go at creating some levels:

TSLA Bull - 2-10% of portfolio (S&P ETFs etc. excluded)
TSLA Big Bull - 10-100% of portfolio but doesn't believe FSD/RT will contribute more than 10% of market cap before 2030
TSLA Super Bull - 50-200% of non margin portfolio but doesn't believe Optimus will contribute more than 10% of market cap before 2030
TSLA Hyper Bull - 80%+ of non margin portfolio and believes market cap will exceed $20T before 2030

No bull *sugar*?

 

[elasalle]

Yep. It might make it more expensive if you have to compete for limited local minerals, but it wouldn't stop you from importing them as was probably planned. It might even make them cheaper to import as demand is reduced from other places going for local minerals.

For same money, they can produce more in other locations and Tesla will buy them all.

Since CATL don't have end product(car), what advantage do they have for building with everything local in US .. where they don't even have local sourcing and all their rare minerals/supply chain is in China ... why, just to make GM, Ford happy

Also for the batteries they build here, TSLA likely gonna be direct competition with a 65% cost reduction. $75 kWh state side by 2026 ... no way

Also someone might have said " No Shipping"

 

[AudubonB]

Nah, Doris! (Saphron)

And for those playing 4D, I Bet (tte Midler) the Get Shorty (ies) reference is what make it topical??

Bingo!

And I'm going to have to up the pot to 420 TeslaPoints for not allowing me the punchline

 

[Favguy]

I feel this is why Elon keeps calling it a robotaxi. It’s really the model 2, but he jus r doesn’t want to osbourne sales right now. Depending on the progress of FSD beta they can choose to deploy the autonomous version of the car or passenger version, with so modular interchangeable parts.

Martin V sort of hinted at this in his analyst meetings and Elon all but spelled it out in the earnings call. Also let’s not forget China’s desire to not just be a production plant but to show off a Chinese-inspired/designed car of their own making for the world, an unfulfilled Elon goal (and a big reason why China will still be a backer of tesla for the forseeable future regardless of what FUD articles come out).

Unless the Taiwan situation develops badly and a tit for tat sanctions escalation follows between the US & China. I'm hoping the Berlin & Austin ramps happen rapidly as I'm factoring in China hostilities as at least a none negligible risk to our TSLA derived wealth.

PS, I look forward to anyone convincing me as to why I may be very wrong on this worry...