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

[ggr]

I disagree, let me explain:

You just forgot the very first wave:

Buyers at IPO, us first gen Roadster owners. If you go back into the history of TMC, you will see that a lot of Roadster owners accepted the offer from Tesla to participate at the IPO at a price of 17 USD. At that time that was super risky move for a small investor. I got 500 shares allocated and I was super happy then, even more so now. The gains we can reap in the future is proportional to the risk we took back then. So please don‘t omit Roadster owners in the history books

Yep, I was another of them. My big regret is that I gave away to charity all of those initial shares I bought, to get the very significant tax advantage. I wish I had kept a couple just to prove that I had them. I still have 33 shares at $33.76 cost basis from December 2012.

 

[ItsNotAboutTheMoney]

Wow, that’s a profound discussion. Highly recommended.

I suggesting watching the full discussion, rather than just the Autopilot "clip":

 

[HG Wells]

Fox con blocked from restarting production on Feb 10th.
My Apple puts should do well.

"Chinese authorities have blocked Apple supplier Foxconn's plan to resume production in China from Feb. 10"

China stops Foxconn's plan to resume production on Feb 10 on coronavirus worries - Nikkei

EDIT I was amazed how thinly traded 290 puts are.

 

[StealthP3D]

Yeah, I don't know how often we have to talk about this:

TMC service announcement 1: Please ignore every article about Tesla's "falling demand" during the first month of every quarter as all those deliveries are simply leftover from the last end of quarter rush. Doesn't mean a thing as cars produced in any given quarter will only start deliveries ~mid quarter. Rule of thumb: month 3 deliveries are 50%+ of the quarter.

TMC service announcement 2: Please ignore every article about Tesla's "falling demand" related to comparing pre-Model 3 export and post Model 3 export era US numbers.

TMC service announcement 3: Please ignore every article about Tesla sales "soaring" related to comparing pre-Model 3 export and post Model 3 export era EU/China numbers.

PS: These may change as GF3 and GF4 production picks up, but may not completely disappear unless the full range of products is produced for each continent within that continent. (i.e. probably never)

I would go a step beyond 1,2, and 3 and ignore any article focussing on sales numbers of Model 3 or Model Y because Tesla can sell every one they can make. All that matters as far as the eye can see is how many they can manufacture. And that has been capped by how many batteries they can get their hands on. This is a production story, not a demand story. I would say, "Keep your eye on the ball" but it would be more accurate to say "Keep your eyes on the batteries".

I've been pounding my fist on the table for over a year about this while the short-sellers and media pundits continue to question demand. TSLAQ focussed on trying to find the secret parking lots full of unsold cars that proved there was no demand. Almost every Tesla "news" story put the question of "demand" for Model 3 front and center. Analysts continued to question demand. The only cars that Tesla makes that are demand limited cost over $70,000. That is a small market. Model 3 and Model Y have price points that make them accessible to most new car buyers. Demand is not, will not and has never been the issue. And as production volumes continue to increase, production costs continue to drop, giving Tesla powerful levers to pull to match demand to ever-growing production.

This disparity between perception (of limited demand) and the reality (of being production constrained) is a primary underlying reason why TSLA has been so profitable for so many of us. All of my most profitable investments over the years have featured a large disparity between perception and reality.

Now that more people are finally understanding the fact that the limits of demand have yet to been found, I fully expect the focus to start to move towards doubts about being able to scale production to match valuation. And Tesla will prove the doubters wrong. Elon told us about a year ago that it was all about the batteries (and autonomy). But the pundits didn't believe him as they continued to focus on sales. Sales only matter as an indicator of production volume. This means sales comparisons in various locals are virtually meaningless. As an investor, you need to remember one little thing: Autonomy might blow this thing out of the water but, until then, "It's the batteries, stupid!".

 

[Prunesquallor]

...

I put my money on the Houston area, specially near the port for logistics.

It would be kick-a$$ if this were true. Being one of your men-on-the-ground, I wonder what the best sources to pick up rumors and tidbits would be? I have heard zero scuttle-but about this so far.

Also, maybe I’d better train as a drone-pilot...
(Actually, if the location was in the industrial or Port of Houston area, dronage would be VERY frowned upon.)

 

[generalenthu]

I do not believe we are goin by vins in hw3 upgrades. My vin 41xx with full FSD paid at purchase still waiting in northern va market (huge number of tesla cars and most earlier vins in california)

Got a tire rotation appointment yesterday Tyco Rd for 20th. Guess who's getting a new FSD hardware upgrade. 19xxx Vin.

 

[Cherry Wine]

Fox con blocked from restarting production on Feb 10th.
My Apple puts should do well.

"Chinese authorities have blocked Apple supplier Foxconn's plan to resume production in China from Feb. 10"

China stops Foxconn's plan to resume production on Feb 10 on coronavirus worries - Nikkei

EDIT I was amazed how thinly traded 290 puts are.

The region has 4x the reported cases as compared to Shanghai, so not surprising in that respect. I am growing more concerned with the economic fallout (as is the Chinese government based on its actions). Hoping to see sustained signs of decreased transmission.

 

[sparcs]

"Chinese authorities have blocked Apple supplier Foxconn's plan to resume production in China from Feb. 10"

China stops Foxconn's plan to resume production on Feb 10 on coronavirus worries - Nikkei

the relevant quote about Foxconn:

Public health experts carried out inspections at Foxconn's factories in Shenzhen and told the company that its factories had a "high risk of coronavirus infection", making them unsuitable for a production restart

That may imply that Giga Shanghai passed these inspections. Edit: Just noticed Cherry Wine's added post about extra severity of coronavirus in that region where Foxconn factory is located.

 

[MarcusMaximus]

That makes you optimistic? Ugh, no. Never start a project in Python that it's obvious you're going to have to end up doing in C++. I've a number of times had the misfortune of starting projects in Python and then later discovering that I had to switch to C++, and it was a huge waste of time every single time.

Python = Best for simple projects where memory and/or performance are not critical
C++ = Best for complex projects where memory and/or performance are critical
Cython = Where you started a project in python and then discovered that memory and/or performance are critical and that you really screwed up and don't have the time to rewrite the whole thing

FWIW, this approach is common for large machine learning projects. Python is quicker and easier to experiment with(and, largely because of that, most of the most popular ML frameworks are in it), plus the people designing the nets aren’t typically expert software engineers. So it’s commonly used on the training side. On the inference side, you need everything optimized for feed-forward runtime on that hardware, so pre/post processing is converted to C++ implementation(the network itself is typically just a precompiled machine-code binary loaded onto the GPU/ASIC). I work in this industry, and this is basically the practice, across the board.

 

[Artful Dodger]

Dumb question that I couldn't find the answer to: What does SHO stand for?

Lol, never seen a definition for what SHO means, but the Backronym Generator offers up this fitting gem:

Spongy Horse-and-buggy Out-basket​

But seriously, read the entire Wikipedia article, especially the section on Regulation SHO.

Cheers!

 

[AlanSubie4Life]

Seems like a good plan. I am pretty sure my highway numbers in the spreadsheet for the Model Y will be close to the actual value from the test. Has to be!

In any case, Model Y looks like it will (predictably) have much worse range than Model 3 on the highway, under best case conditions at least (it is still fine though). I would guess long term they’ll want to cram more battery in there if they can.

@Words of HABIT any particular reason for disagreement with this?

Definitely will see what sort of numbers I can pull out of the detailed certificates that @bhtooefr has pointed out to see if I can track the efficiencies on these other cycles.

But the fact remains that the raw HWFET is a very decent (optimistic) predictor of best-case Highway range. And comparing that result to Model 3 should be safe. And it clearly (no surprise!) is a lot worse than Model 3.

@SageBrush yes I think that squared coefficient isn’t bad as a model of relative aero loads. When calculating to get predicted Wh/mi impact, though, you have to make some assumptions about drivetrain efficiency; it can’t be used directly. So you can’t necessarily use it to compare the efficiency of two different vehicles at highway speed, if the drivetrains are different. I’ve never tried to piece this all together though. I just have my own empirical physical model, which is good enough for who it is for.

 

[MarcusMaximus]

OT more FSD discussion:

I am very bullish that if Tesla can achieve accurate FSD, it will be very bullish on share price.

I am not confident at all that this is happening soon. Frankly, I have no idea how anyone here can have high confidence, unless you work in computer vision and are familiar with this specific type of problem.

As background, I work in building machine learning / deep learning algorithms that end up running on memory constrained systems.

I was always skeptical about FSD on HW2 in 2016 because of computational limitations. Images weren't even proccessed at full resolution - they were downsampled. Not a good start.

Before HW3 came out, @jimmy_d was excited because some big neural net was uncovered that was used for all cameras and took not one, but 2 full scale images as input.

Exciting yes, but 2 full sized images is a far cry from 6 seconds of video (if say that is what Dojo is going to be training on).

My point being, I think we should be excited that Tesla has the data and possibly getting the computational resources to train a great FSD net. But I don't see any evidence that it would work on HW3.

I am curious about peoples thoughts on why they think differently, other than "Elon said so"

You don’t need all those frames at the same time, as long as you have some recurrent layers in the network. My guess would be that Dojo uses 6 seconds of video as the sample for training a recurrent net, vs passing all 180-360 frames as input for a single feed forward.

 

[rsm287]

Wow, that’s a profound discussion. Highly recommended.

Yes, totally agree. In my ignorance I didn't know who Jim Keller was prior to watching this, but I looked him up afterwards. It seems he's a bit of a rock-star when it comes to chip architecture/design. That someone who is held in such high esteem thinks so highly of Elon and the Tesla approach makes it even more profound.

Loved what he said about first principles thinking - here's a couple of quotes from the podcast:

You think you have an understanding about what the first principles of something is.....and then you talk to Elon about it.....and you didn't scratch the surface......no matter what you do, it's only a local maximum...

Imagine that 99% of your thought process is just protecting your self conceptions (assumptions).....and 98% of that is wrong

Pretty mind blowing stuff!

 

[Krugerrand]

I tried to find the source of your data - best I found is: February 2020 Reg SHO Daily Files where I had to extract one line from each daily file and it does not have the last column you showed - Please let us know if there is a more convenient way to get them?

I then looked up the data around the last two times I remember the uptick rule being in place for TSLA - April 4, 2019 and July 25, 2019. I also pulled out the data for last week for comparison:

View attachment 509217
It looks like the exempt interest goes always way up if the circuit breakers are tripped - makes sense now to me as it allows them to short sell ignoring the uptick rule.

It also looks like the exempt interest slowly was creeping up last week already.
Edit: It started the day of the earnings call.
Is that a sign that:
- the MM had problems with the increasing SP and/or volume?
- manipulations started last week already?

Might be interesting to watch in similar situations in the future as an early warning sign???
Like right now ...

More and more people speaking Greek in this thread. It’s like having dinner at the spouse’s Oma’s and everyone jabbering in German and me be . The best I can do is mimic facial and emotional expressions in that situation. And nod. A lot.

Here I have to interpret the written word, which is just . I’m extrapolating and believe my reaction to this post should be .

 

[StealthP3D]

Cathie Wood on Tesla’s meteoric rise: ‘We’ve been waiting for this’

I feel as though Cathie Wood’s presentation is getting better and better. Like a Tesla it just gets better and better.

Nice! One thing that might turn off semi-knowledgable people is her claim that "Tesla has the best battery technology". Because you might know that Tesla has Panasonic making their batteries in the USA for cars made in America and they will be buying batteries from LG for cars made in China. I wish she would clarify that "battery technology" extends far beyond the individual cells. A car "battery" is a collection of cells (much like an artillery "battery" is a cluster of armaments under centralized control).

Tesla's technology battery lead to date is primarily speaking to the pack level (the physical end electrical design and construction) and also the software control of the pack. It manages charging, discharging, pack cooling and heating to extend battery lifetime, etc. This is very valuable stuff that can't easily be copied without reverse-engineering the software and fully understanding what Tesla was trying to achieve. But people who only know enough to be dangerous to themselves constantly repeat that Tesla doesn't have a lead in batteries because they simply purchase their batteries from others.

 

[SageBrush]

the relevant quote about Foxconn:



That may imply that Giga Shanghai passed these inspections. Edit: Just noticed Cherry Wine's added post about extra severity of coronavirus in that region where Foxconn factory is located.

The photos I have seen of Foxconn are workers pretty tightly packed, like people at a bar.
One step below sardine packaging

photos of foxconn worker conditions - Google Search:

 

[Stretch2727]

I tried to find the source of your data - best I found is: February 2020 Reg SHO Daily Files where I had to extract one line from each daily file and it does not have the last column you showed - Please let us know if there is a more convenient way to get them?

I then looked up the data around the last two times I remember the uptick rule being in place for TSLA - April 4, 2019 and July 25, 2019. I also pulled out the data for last week for comparison:

View attachment 509217
It looks like the exempt interest goes always way up if the circuit breakers are tripped - makes sense now to me as it allows them to short sell ignoring the uptick rule.

It also looks like the exempt interest slowly was creeping up last week already.
Edit: It started the day of the earnings call.
Is that a sign that:
- the MM had problems with the increasing SP and/or volume?
- manipulations started last week already?

Might be interesting to watch in similar situations in the future as an early warning sign???
Like right now ...

Confused on the short exempt classification. So they can sell shares without locating shares to borrow. I can understand how market makers would need this ability but eventually they need to buy the shares back or borrow them. If not it is like they are issuing stock which cannot be possible.

So in this case they heavily short and just buy back slowly at a lower price? No need to locate shares to borrow. Correct?

 

[dc_h]

Yeah, Austin isn't Mordor. That'd be Port Arthur.

Tesla needs to build a foundry to make their special Stainless steel alloy, then needs to deliver it to SpaceX for Starship production, and Tesla for Cybertruck.

One caveat here: if plans for a 2nd Gen Cybertruck acually do involve a giant casting machine for injection molding the entire hull of CT2, then that process will be done somewhere near the foundry, were they will have easy access to the still molten steel.

Port Arthur and the surrounding industrial area also makes large tanks for the oil industry. SpaceX is planning to build Starship like a water tower (tanks built from loops of steel).

I put my money on the Houston area, specially near the port for logistics.

It would be something to build a big 10,000 employee plant in Port Arthur. That is the hub of big oil and the chemical industry. It is also the soul of legacy space. The heart is in Alabama in a box on Richard Shelby's fireplace mantle. Politically it seems like it would make sense to have an AI and chip team in Austin and the manufacturing down in Houston. The size of the oil processing and Dow and other chemical sites is crazy. I would love to see Tesla go into the heart of big oil.
On a tangent, my colleagues in Dallas are really coming around on Tesla. I know some early fans, but others who were real haters and doubters. Seeing the cars and hearing about the warrantee and low cost of operating, and the pure beauty of the car is winning over some serious doubters.
I'm still serious about demolition derby with the Cybertruck. If Elon held an event with all the big trucks, even if Cyber lost, the marketing would be amazing and Elon would vow and deliver on improvements to win the next time. It would be a break the internet moment (again) for Tesla, probably 10x the model 3 reveal. Monster truck shows are invisible to most of us here, but they sell out 100,000 people in Cowboy Stadium (ATT Center) and it would be a demonstration of product parity and hopefully superiority. This would be Nurburgring for trucks, but with way better videos.

Sorry mods, but I think demo of the truck kicking butt will be relevant to the stock price. The only problem with the truck is making a plant fast enough to make the trucks and then spending 5 years catching up with demand.

 

[pz1975]

You mention me as a victim of this illegal manipulation, but it only gave me a bruised ego (and I actually made 15k during my involuntary absence). But people with huge option wins that evaporated, like @pz1975 who lost half a million or more compared to his planned exit point, really have a reason to be annoyed, to put it mildly. Maybe someone who ‘lost’ a lot should start a class action suit to go after this illegal naked short selling.

I actually lost just over $1 million in profit because of the dip (my short-term calls were up by $1.8 million when the SP was $960). I still made $800K profit on them by selling Wednesday morning so I am not complaining overall of course. But I wonder what could have been, especially if the SP had continued up above $1,000 like it looked like it would do. It may have ended up with $3-4 million profit.

 

[defc0n]

Wow, that’s a profound discussion. Highly recommended.

Funny to me how Jim just hand waves and says it’s a data problem He may be right but a lot of time and hard work from people still need to make it all come together. I’m sure he knows that too though.

Keep in mind neither he nor anyone is quite sure how good the computer is going to need to actually be. But he at least seems confident the hardware needed is pretty close to being there.