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

[sundaymorning]

Following was posted on this sub (last year's thread) a year ago (April 2018). Can't seem to find the original poster.
(I keep the good stuff posted here stored in a separate document)


6 prior TSLA dips of 20%+ over the past 3 years:

8/24/15 Dip of 25% over 4 trading days, recovered to 104% of prior peak over 19 trading days; only 3 trading days to recover to 89% of prior peak

10/20/15 Dip of 25.5% over 21 trading days, recovered to 86% of prior peak over 12 trading days

2/9/16 Dip of 42% over 27 trading days, recovered to 111% of prior peak over 40 trading days; only 24 trading days to recover to 89% of prior peak

5/12/16 Dip of 24.5% over 25 trading days, recovered to 89% of prior peak over 18 trading days

6/27/16 Dip of 22.1% over 13 trading days, recovered to 98% of prior peak over 24 trading days; only 4 trading days to 89% of prior peak

7/10/17 Dip of 21.5% over 10 trading days, recovered to 90% of prior peak over 13 trading days

4/2/18 Dip of 32.2% over 18 trading days. Recovery so far to 80% of prior peak of $360 over 2 trading days

Conclusion: expect a 24 day heavy rise before a new dip occurs

No credits for me, I'm not the original author of this

I believe @bdy0627 was the original poster.

 

[Zero CO2]

I am sure this has been posted already but if you are long TSLA...it is a must watch...all 2 + hours
Now I finally understand the technical long term competitive advantage in FSD that Tesla owns outright ...
they just needs to navigate the regulatory environment...proved to me why vision(CNN) is the only correct path..to FSD

i would be curious to hear the counter arguments to those who disagree...preferably those who actually understand physics and biology at a PHD level ...

 

[neroden]

OT

i would be curious to hear the counter arguments to those who disagree...preferably those who actually understand physics and biology at a PHD level ...

Oh, it is the only correct path. I just think it'll take them so long to get to marketable, approved driverless robotaxis that it's not an investable opportunity.

(FWIW, I do not have PhDs in biology and physics, but I am quite capable of reading the literature and I do so)

 

[Artful Dodger]

not that i saw

Rob Mauer has a bit on the investor call in his Tesla Daily Podcast:

Increased Capital Raise, Details from Investor Call (05.03.19)

 

[neroden]

These "investor calls" where Tesla management discloses material information to a select few investors and doesn't make it public are totally illegal. They violate Reg FD.

Where's the SEC?

(I know, in the pocket of Wall Street crooks.)


------
There were definitely material statements, including that Model Y location decision is next week; that if it's at Fremont, they'll expand the south paint shop; that Model S/X orders in Q1 were weak because news of the refresh leaked; that Panasonic has not reached full capacity yet; that they're lining up other cell suppliers in China.

And that's what we got from the notes from the short-seller. We have no idea if they made other material statements.

This is illegal.

 

[Remus]

OT


I believe the individual training cases for scenarios take 10x longer to collect as you go to the next set of less common scenarios, however. (Maybe only once you get to a certain point, where the scenarios only happen in certain weather, or in locations which have very few cars per day traversing them... but eventually you get to scenarios where it just takes time to collect the data and you can't speed it up.)

Apart from that: the real problem which slows development is that as you work your way to less common scenarios, you may discover that there was a piece of information you should have been tracking and extracting which you weren't, which requires major software architecture revisions, and in the worst case, hardware architecture revisions. That slows things down.

(For example, after training the system to recognize car-with-bike-on-back as one object, you realize you actually have to train it to recognize it as object-plus-object-which-might-fall-off, and you have to go back and redo great hunks of the system training.)

Software architecture changes all the time. This is part of "meta parameters" choosing game all AI teams keep playing all the time, which Karpathy touched upon during its presentation. Yes, you basically have to retrain from scratch. But it's fully automated with existing labeled data, which can be trivially accelerated by throwing more computers at it. We are talking weeks, not months or even years. And the developers can keep working during this time.

The example you used only requires adding one "feature" in the labels which I bet they change all the time.

I don't think you need to worry
Hardware architecture changes causing big software change. NN computation is basically multiplication of matrixes. The hardware architecture is basically how to make that and other frequent operations more efficient more efficient. That is not gonna change unless there is another break though from academic with a whole new paradigm of doing things. Then all hell break lose.

 

[Remus]

While I agree that their costs are carefully structured to not scale exponentially, I am not so sure about the time involved, which may still increase exponentially later on.

Not if the size of the fleet increase fast enough

 

[Tslynk67]

Total ICE bans in European cities under serious consideration

The Brussels Times - Amsterdam plans to ban all fossil fuel vehicles, Brussels looks to follow suit

The Amsterdam city council has approved a ‘clean air action plan’ to ban all fossil fuel vehicles by 2030.

The city assembly still needs to approve the plan, although it is likely to take effect at the end of this year. The council hopes for Amsterdam to become the world’s ‘capital of emissions free mobility’. The proposed plan would also aim to save inhabitants’ lives by reducing air pollution. At the moment, Amsterdam residents lose on average lose a year of life expectancy because of it, according to data from the Global Burden of Disease.

In Brussels, a low-emissions zone was introduced in 2018, within which the most polluting vehicles are banned. Failure to comply with the criteria for access to the zone results in a 350 euros fine.

The Brussels government targets a complete ban of all diesel vehicles within the region by 2030, while in the longer term "the next step" would be a ban on all petrol vehicles, according to Brussels minister for environment Céline Frémault.

"Diesel vehicles must be banned at the latest by 2025 in Brussels and cars running on petrol or natural gas in 2030," commented Joeri Thijs from Greenpeace Belgian office last week.

"More than half of Brussels families do not have a car, but they breathe the worst air in the country and this must change," said Brussels mobility minister Pascal Smet, one of the initiators and proponents of Villo, Brussels public bicycle sharing network. Brussels' Villo network has now passed 12 million trips since its inception, and plans the rollout of electric bikes this Summer.

Amsterdam will execute its plan in different stages. Starting next year, diesel cars older than 15 years are banned from entering the inner city. From 2022, buses running on petrol or diesel will follow.

The big change will come in 2025 when the zone’s limits are expanded to the area inside the A10 ring road. At the same time, the number of banned vehicles will be expanded to include canal ferries, tour boats, trucks, vans, taxi’s and buses, as well as mopeds and scooters. By 2030, all vehicles entering the city – including cars – will need to be emissions free.

Amsterdam city counselor for transport, Sharon Dijksma (PvdA), promises subsidies for people who have their old vehicle scrapped or replace it with an electric one. The city will up the number of charging points from 3,000 now to 23,000 in 2025, including more fast chargers.

Emissions-free vehicles will get privileges like parking licenses. Norway is a leading example in the field of such privileges. The country grants electric cars access to bus lanes, more parking spaces and lower parking tariffs.

EVs becoming cheaper

“This is a good approach for better air quality. Including existing vehicles rather than only focus on cars bought new now is daring,” said Maarten Steinbuch professor in automotive technology at the University of Eindhoven.

Will the plan work and get all inhabitants to convert to electric cars? Steinbuch thinks it will. “In five years time, fossil fuel cars and electric cars will cost about the same. Electric cars will eventually even become cheaperbecause their engine is much less complicated.” Steinbuch expects tourists and commuters to transfer to e-mobility at the city limits.

On average, people of Amsterdam have fewer cars than the rest of Dutch inhabitants, although there are still 200.000. The city will not get 200.000 charging points, which is not a problem according to Professor Steinbuch.

“The average daily drive is 19 kilometers. Batteries are getting larger and offer a range of about 350 kilometers. You will only need to charge a few times a week or use a fast charger for fifteen minutes.”

Exemplary role

City Counselor Dijksma feels the city’s plan fits within the uniform emission zone regulations the Dutch government wants to implement.

“Our integral approach is the most ambitious approach achievable within those regulations.” Later this year, Amsterdam will already go for fewer cars in the inner city, removing 10.000 parking spots to create more room for cyclists.

Will the Dutch government accept Amsterdam’s radical approach? Steinbuch thinks it should. “There is a lot of criticism on subsidies for electric cars, but at least they make people invest in them,” pointing out that Tesla has its European HQ in Amsterdam and electric buses are produced in Valkenswaard.

 

[Artful Dodger]

After listening to it four times I think it's "SEC".

Great. Now you're just the one to explain how the SEC would have approval authority over a merger, rather than the FTC. Thanks in advance.

Merger Review | Federal Trade Commission

Merger Review

How Mergers are Reviewed

"Among the key provisions in U.S. antitrust law is one designed to prevent anticompetitive mergers or acquisitions. Under the Hart-Scott-Rodino Act, the FTC and the Department of Justice review most of the proposed transactions that affect commerce in the United States and are over a certain size, and either agency can take legal action to block deals that it believes would “substantially lessen competition.” Although there are some exemptions, for the most part current law requires companies to report any deal that is valued at more than $90 million to the agencies so they can be reviewed."​

 

[neroden]

Great. Now you're just the one to explain how the SEC would have approval authority over a merger, rather than the FTC.

The SEC has approval authority over the stock swap, bond indentures and similar *securities related* changes associated with a merger. It's usually rubber stamp, but Tesla did choose a weird structure for the merger, and the SEC is being stupid about Tesla, so it could have caused a review delay.

 

[neroden]

Not if the size of the fleet increase fast enough

Fair point, but the fleet only doubles every two years.

 

[goinfraftw]

Had a chat with a mentor and friend about Tesla today and he asked about the rising risk profile on Tesla and concerns about the stock overall. These were my thoughts:

The current risks (that he mentioned):
1. Departing Board Members
2. Outside looking view of change in Autonomous Driving (or "Robotaxis")
3. Slumping sales
4. Federal Tax Credit Ending
5. Increasing Competition
6. Quality Issues
7. Support Problems
8. Logistic Issues
9. Convertible Bonds raise of $2.7B

Other risks (I mention):
10. Macro factors: recession looming
11. Side effects on Automotive market due to Tesla effect
12. Labor effects on Automotive market and negative sentiment effect
13. No effect on global temp change due to Tesla push
14. Resource crunch due to hindrance by negative forces

Headwinds (2019):
1. China Gigafactory Completion in May
2. Buffalo Gigafactory Ramp-up in Q2
3. Reduced cost to operate for Automotive
4. Increased sales of FSD (baked in to Automotive Sale)
5. New paid software features
6. Multiple quarters of positive new cash flow (thus, incorporation of Tesla into S&P 500)

Considerations:
* Is climate change real? If agreed upon yes, then entire entire energy consumption to usage lifecycle needs to be rethought towards renewables
* How much is Automotive and Energy Consumption+Usage Utility Market worth? Both are $3T markets (Tesla is tackling both)
* Is Tesla the primary leader in the new market? I'd say yes based on demand and execution.
* Will Tesla continue to be the primary leader going forward? I'd say yes because they have the best lead and they're moving faster. With even better products likely to be built off of great design and data.
* What does a market leader generally comprise of in market share? Anywhere between 10-50%.

With all of those laid out (I'm sure I missed some), the simple investment question is, "do you believe climate change is real?". I've believed yes since 2009 and wanted to see someone executing on fixing the problem. I've learned that it's not just a problem to fix, but an innovation cycle that will make things better. Tesla is a safe bet, as long as it has cash, to become gigantic...anywhere from a $600B -> $3T market cap even though it's sitting on ~$45B today. If you take a 5-10 year view out and look at the situation, I'd say Tesla is in the 3rd-4th inning of a fight to completely replace our energy consumption and usage + transportation lifecycle [ECUT Lifecycle] against the incumbents that have spent the better part of 100+ years owning it and established companies, organizations, and national powers around it. This all likely affects a lot of people...

...so, you're essentially betting that Elon Musk and Tesla (and everyone that works and/or roots with/for them) is there in order to fix one question: is climate change real? If so, then that means the likelihood of the entire human civilization being wiped out is possible. It's very ...extraordinary and the stakes sound really dire and out of a comic book. Though, that's the situation that we seemingly seem to be in. From the risks you outline, the Tesla strategy is to replace the world's ECUT Lifecycle and needs to use the current methodology to do so in order to deliver cars and new utilities in a situation where the weather is against you (e.g. Puerto Rico). So, slumping sales, volatility of board members, "quality issues", and "support issues" can be characterized by worldwide strategy being implemented to ECUT products around the world from only a 42B company. Also, this is undercutting sales of other very large companies as we see in US April auto sales (most unsold ever). How do you deal with this situation other than keeping on pushing forward in the 3rd-4th inning rather than throwing up your hands and saying, "oh *sugar*, this is too much and too many problems to deal with!". I think that's where Robotaxis have come in.

Tesla's vision since 2015 has incorporated this concept due to huge under-utilization of the concept of the car of 5% of time being used for driving and the rest just sitting around doing nothing in a parked state. I believe Tesla's going to find out just how much that utilization rate can be increased by deploying the Robotaxis (maybe 1st in Norway) and seeing how profit margins can increase. It's incorporated into the design of the car/vehicle itself with 1M miles as a design constraint. With every % increase in utilization, the less cars need to be sold (out of the 98M annual car worldwide sales market). We're in this massive innovation change and we're all a part of it at this point to make sure climate change is averted now that most of the world (and hopefully the entire world) understands that it's real.

Anyways, probably missed some points, but that's what I can come up with off the top of my head. Am I missing anything?

 

[Skryll]

They land it right on $255 what are the odds? If that’s not manipulation, lol

So whats the number for next week

 

[dqd88]

The price of the FSD option will likely go up as Tesla keeps introducing new FSD technologies. I pointed out potential FSD price hikes last October.

But as the price goes up I'd expect there to be more levels to it: for example "base", "consumer", "professional" and "commercial". It would be a marketing/pricing mistake to sell any of the individual FSD components for much more than ~$5,000, as it wouldn't allow budget constrained customers to stretch as long as they can.

So I'd expect Tesla to accommodate regular car owners: if for example in two years if commercial level FSD costs $19,900 then Tesla should make $5k, $10k and $15k variants available as well - why lose a lower price sale if someone doesn't have budget (or interest) for the full $20k option?

Tesla might also offer revenue sharing tiers - for example with a $20k FSD option Tesla keeps 40% of the robotaxi revenue, but pay $50k upfront if Tesla only keeps 25%.

I.e. since Tesla FSD is going to be a software licensing plan, not a physical option, Tesla has literally an infinite amount of pricing options available to front-load revenue or earn it later on, if they so desire.

I also expect 'FSD jail-breaking' to be a very lucrative business - but since Tesla's new AI chip has built-in crypto it's probably a losing battle in the long run - so to run Tesla's FSD chip you'll need valid a license.

This is a very sound approach. Always impressed by how you think. I was wondering how they (tesla inc) might approach a skyrocketing model 3 demand if say entrepreneurs buy up fleets of them for the network. The tiers makes sense.

 

[Heimdallr]

I have some dry powder. Should I buy TSLA or wait for the uber IPO? I'm already kicking myself for missing Pinterest and BYND.

Thank you in advance for subsidizing my Uber rides in case you wait for the Uber IPO.

 

[printf42]

OT

I believe the individual training cases for scenarios take 10x longer to collect as you go to the next set of less common scenarios, however. (Maybe only once you get to a certain point, where the scenarios only happen in certain weather, or in locations which have very few cars per day traversing them... but eventually you get to scenarios where it just takes time to collect the data and you can't speed it up.)

Apart from that: the real problem which slows development is that as you work your way to less common scenarios, you may discover that there was a piece of information you should have been tracking and extracting which you weren't, which requires major software architecture revisions, and in the worst case, hardware architecture revisions. That slows things down.

(For example, after training the system to recognize car-with-bike-on-back as one object, you realize you actually have to train it to recognize it as object-plus-object-which-might-fall-off, and you have to go back and redo great hunks of the system training.)

This is not a problem for deep learning, and it’s actually the beauty of software 2.0.

So basically with deep learning once you discovered a new case you need to cover, you just add it to the dataset and retrain the network.

Sure when you added too much new stuff to the original dataset, you might need to tweak the network architecture, but that can be done fairly easily, and most people would start with a network that’s bigger than needed anyway.

So in short, deep learning engineers are practically retraining the whole network every day, so covering new cases is not that big of a hustle and won’t make the project scope grow exponentially in a traditional sense.
(Most of practical CNN training will converge within hours not days, don’t know exactly how Tesla’s network/dataset looks like but they have resources to massively parallel the process, so it shouldn’t be a big deal)

 

[dqd88]

Thank you in advance for subsidizing my Uber rides in case you wait for the Uber IPO.

haha.
I wouldn't stay in Uber long term but I think it will take off in the first week. Will be different than Lyft.
I think Tesla will put a major dent in Ubers long term plans, and possibly take it out of commission. But that's at least a few years away.

 

[RobStark]

U.S. rejects Tesla bid for tariff exemption for Model 3 'brain' - Reuters

 

[Right_Said_Fred]

There were definitely material statements, including that Model Y location decision is next week; that if it's at Fremont, they'll expand the south paint shop; that Model S/X orders in Q1 were weak because news of the refresh leaked; that Panasonic has not reached full capacity yet; that they're lining up other cell suppliers in China.

And that's what we got from the notes from the short-seller. We have no idea if they made other material statements.

This is illegal.

I don’t read much that wasn’t already known. And if there was some new information, it wasn’t very material. The only ‘news’ I see is that the location choice for Model Y will be made next week. But they already said it would be soon. So not exactly stock moving info.

 

[T@oo]

I don’t read much that wasn’t already known. And if there was some new information, it wasn’t very material. The only ‘news’ I see is that the location choice for Model Y will be made next week. But they already said it would be soon. So not exactly stock moving info.

... and btw, we move the production forward by 6 month... one can dream.