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

[NicoV]

The reason this particular effort can be trusted for the time being is that Tesla appears to be generating order invoice IDs sequentially. This makes the "highest reported ID" conservatively approximate number of Model 3 orders in every EU country.

(The bad number from Finland was probably found by seeing a too large gap in the sampling.)

It doesn't have the uncertainties of VIN sampling.

And invoice numbers have to be sequential by law.

 

[Fact Checking]

when I saw this story, my immediate thought was that Exxon will just destroy the most incriminating stuff, but I don't see anyone else raising this as a concern. what are the impediments to Exxon doing this? I get that destroying evidence can be a crime... but, the courts would have to know that the evidence had existed to know it was destroyed.

It's very stupid to destroy evidence even if you have no moral compass whatsoever:

• The most incriminating bits were possibly already part of various lawsuits, just settled quietly or seal-protected.
• It's never completely clear who has various pieces of electronic documents. If just a single copy remained in an old archive, or if one of the employees brought it home and has a copy (because he created the document and worked on it at home), it provides a trail that the evidence exists and proves destruction of evidence.
  
• Whoever does the destroying is facing serious jail time and financial penalties. It's not just 'can be a crime', it IS a crime, and viewed very dimly by the courts, because integrity of evidence is so central to the judicial process.
• Whoever does the destroying would do it to cover up for crimes committed by others decades ago. What's the motivation?
• There's a legal concept that holds that if evidence was destroyed and there's a factual question in a lawsuit that could have been answered by that evidence through discovery, the court will assume by default that the evidence went against the defendant. I.e. destroying evidence can make the outcome of the lawsuit worse.
• Criminal activity like destruction of evidence can also trigger RICO charges, punitive damages and tripling of damages.
• The evidence is likely voluminous and was spread out across the company. Someone would have to order the destruction of evidence. Defying that order and making a copy would be the ultimate "get out of jail free card" to whoever does the destruction...
• Finally, such sensitive type of evidence is usually already handled by lawyers - who know all the above perfectly well, and who'd not just face jail time but also disbarment. They'll try to hide it via legal means, but they'll also be one of the fiercest defenders of that evidence's integrity.

So no, destruction of evidence in the U.S. is a really, really bad idea for large corporate entities. It's rarely done, and when it's done it's either done out of stupidity, or to mask even worse crimes than destruction of evidence: such as murder or actual racketeering, etc.

 

[SpudLime]

Well yea, of course. There have been hundreds before this one, and there'll be hundreds more. The "climax" of the Tesla story never climaxes - it just keeps getting better!

I’ve tried convincing my wife this many times.

Good night folks, going to bed...er.. couch.

 

[engle]

Well I guesse when Fox News tell the potus to end the shutdown he will.

Do you think it will be told to him from “The five”...or “Fox and friends”?

The Donald is special, so I refer to him as IOTUS [I=Idiot], pronounced: eye-OH-tus. Let's get that into the dictionary this year!

 

[SteveG3]

It's very stupid to destroy evidence even if you have no moral compass whatsoever:

• The most incriminating bits were possibly already part of various lawsuits, just settled quietly or seal-protected.
• It's never completely clear who has various pieces of electronic documents. If just a single copy remained in an old archive, or if one of the employees brought it home and has a copy (because he created the document and worked on it at home), it provides a trail that the evidence exists and proves destruction of evidence.
  
• Whoever does the destroying is facing serious jail time and financial penalties. It's not just 'can be a crime', it IS a crime, and viewed very dimly by the courts, because integrity of evidence is so central to the judicial process.
• Whoever does the destroying would do it to cover up for crimes committed by others decades ago. What's the motivation?
• There's a legal concept that holds that if evidence was destroyed and there's a factual question in a lawsuit that could have been answered by that evidence through discovery, the court will assume by default that the evidence went against the defendant. I.e. destroying evidence can make the outcome of the lawsuit worse.
• Criminal activity like destruction of evidence can also trigger RICO charges, punitive damages and tripling of damages.
• The evidence is likely voluminous and was spread out across the company. Someone would have to order the destruction of evidence. Defying that order and making a copy would be the ultimate "get out of jail free card" to whoever does the destruction...
• Finally, such sensitive type of evidence is usually already handled by lawyers - who know all the above perfectly well, and who'd not just face jail time but also disbarment. They'll try to hide it via legal means, but they'll also be one of the fiercest defenders of that evidence's integrity.

So no, destruction of evidence in the U.S. is a really, really bad idea for large corporate entities. It's rarely done, and when it's done it's either done out of stupidity, or to mask even worse crimes than destruction of evidence: such as murder or actual racketeering, etc.

appreciate the thorough response Fact Checking

 

[Singer3000]

Still hard to see how Brexit really impacts Tesla or TSLA. But since you raised it and with an underlined Off Topic applied...

FC is correct - there is currently no parliamentary majority for a second referendum and it's increasingly tough to see one arising. Even if there were one, none of the the UK government, the official Opposition, nor any likely new leader of the Conservative party supports a second referendum. The Prime Minister has wedded her legacy to delivering Brexit and the leader of the Opposition is a hard socialist / communist, who almost certainly secretly voted for Brexit - EU membership is incompatible with key planks of his policy agenda on state intervention.

Unless the Executive is prepared to back an Act of Parliament necessary to legislate for a second referendum, there is no procedural mechanism for one to occur. If I am wrong about the government's intentions and at the last minute it does implement a 12-month delay to Brexit pending a repeat referendum, my view of the British social psyche is that there would be a bigger majority for Brexit than in 2016. That'd be a question for 2020 however.

FC is also correct that a WTO Brexit (what is termed by many as a "No deal Brexit") is the default legal position. Unless a new Act of Parliament is passed stipulating something else, this is what will happen, as currently scheduled on 29 March 2019. Whether this is the most likely outcome at this point is hard to say. There is brinkmanship on all sides, both within the UK Cabinet, parliament and at the European level. The "deal" on the table will almost certainly not pass the initial parliamentary vote on c. 15th Jan, with a massive rather than narrow defeat for the government quite likely.

What happens next largely depends upon whether the EU side are willing to re-open certain terms for discussion to prevent economic fallout for their own citizens, or would rather set an example of the UK. I say this because within the UK, the parliamentary mathematics suggest that the government is perversely more likely to fall if the deal is passed than if it's rejected - the govt relies upon the Northern Irish DUP for its majority. The DUP are to their core, ideologically opposed to the existing deal and would bring the govt down rather than see it passed in its current form. The vote on the 15th Jan is only an indicative vote and an Act of Parliament is still required to implement it, so plenty of time for them to cause political mischief even if the deal is passed in the initial vote on 15th Jan.

It's very hard to say with confidence what might be the most likely outcome at this point and anyone that says they know is lying. If you put a gun to my head, I'd wager that there may be a delay to Brexit of a few months, followed by a negotiated WTO exit. So a "No deal" on trade but with a series of Deals on various policy areas that matter most to both sides (e.g. airspace / overflight rights, passage of radioactive material, criminal record sharing etc...), with the further time used to implement larger customs facilities at the key ports. Needless to say, UK politics (and consumer demand / business sentiment) would be pretty turbulent during this period but life would go on.

The bizarre thing in all this is that even if the UK brexits without a trade or customs agreement, both the UK and Irish have said they will not impose a visible border between Eire and Northern Ireland. And there's good reason for this - there would certainly not be any political or popular appetite to deploy the troops necessary to enforce one. But yet the reason why the Withdrawal Agreement is unlikely to pass the UK parliament is almost squarely due to the terms designed to prevent the need for a visible border on the EU's Irish frontier (the much discussed "Irish Backstop").

If you're not going to implement a visible border anyway, then it would be in both sides' rational interests to sign a mutually beneficial trade deal and manage the smuggling risk of third country of origin goods in a sensible rather than dogmatic way. But politics is not always rational and a "hard Brexit" may be a necessary precursor to the trade deal if the brinkmanship on either side oversteps the line.

Anyway, go Tesla. An accelerated construction timetable for a 100%-owned plant in the biggest car market in the world. Far more important than what happens with Brexit.

The UK is not the decision maker and has no power to extend the clock: it's the EU, and the EU made it clear they are not renegotiating, nor waiting for a referendum with uncertain wording and uncertain outcome. (Which is the right EU decision.)



These UK political figures still talking about a new referendum are disinformed, delusional or are grandstanding.

It's similar to the Greek EU referendum that was mooted by the time it was held.

BTW., the January 15 vote will very likely go against the UK government, which makes a no-deal Brexit the most likely outcome currently.

Numerous major political figures in the UK have publicly disagreed with this sentiment. Over 100 MPs have already signed on to a policy of supporting a second referendum. Far from a majority, but of course the vast majority have not committed themselves either way in advance of voting on May's proposal.

FT's analysis of timing is:

"Would there be time to hold a second referendum before exit day in March 2019? The UK would have to seek an extension of the Article 50 process from Brussels. Many EU experts believe Brussels would agree to this. However, a referendum to resolve the UK’s status vis-à-vis the EU would almost certainly have to be held before the European elections in late May next year."



Yes, those are the three options that would need to be voted on.

 

[Buckminster]

A new referendum is highly unlikely, because the Brexit clock will run out before it can be held: fastest estimate is 8 months, but 10 months is more realistic due to legal challenges. The EU has already made it clear they are not going to extend the clock for a new referendum.

The only 3 options possible for Britain are:

• The 1993 deal, (staying in the EU)
• The 2018 deal, (soft Brexit)
• No deal (hard Brexit - this is the default outcome if Britain does nothing)

One of the above three options is going to trigger on March 29.

Anything else, such as a new referendum or a "renegotiation" of the 2018 deal is not going to happen.

How about throwing us a backstop bone? A treat, nothing more. Woof. We promise not to bite the hand that offered it.

Much of the UK is still in denial about Brexit.

There is no UK anymore - we are reverting to the Scottish clan system. I am developing my tartan based on a bit of Arsenal red combined with Tesla red. Its a bit garish..

 

[Fact Checking]

My tin hat theory is that (a large enough portion of) the market realised that a transition phase was upon us. In the near term Tesla creates its own headwind. Ultimately, all macro stocks currently fossil fueled will be replaced by cleantech driven stocks and Tesla will be king of the hill.

I think the market in general is far, far more shortsighted than that, and anyone making such a realization will simply restructure their portfolio accordingly to be green-tech-proof or green-tech-weighted - which would drive $TSLA up, not down.

Starting some sort of global anti-Tesla conspiracy needs enormous resources and there's very few such entities with the capability (and near-criminal motivation) to execute on such a plan. Yes, there's a shared interest in driving anti-Tesla FUD, but there's very few mechanisms for shared action against Tesla, other than selling it short that is.

I think the macro headwinds are pretty clearly identified and are not directly related to Tesla - see this recent post of mine:

Macro is still a big factor I believe:

• Fed rate rise worries are easing after Powell clarified on Friday that they are data driven and if the data is worse than expected they won't rise rates (which was pretty clear from the recent FOMC statement to begin with but the markets can sometimes be rather obtuse). Not raising rates will also get the Trump propaganda machinery off the Fed's back - and not raising rates based on "data" is the perfect excuse: with the 2020 elections getting nearer the Fed will probably try to stay out of politics and keep rates constant.
  
• Brexit is still a cluster-sugar, and the British political elite still doesn't realize that they got outsmarted by EU competitors who will now take over Britain's (former) EU-internal markets. Brexit it is in March 29, or full capitulation via Article 50 withdrawal. Whether soft or hard Brexit probably doesn't matter much to the markets anymore: the EU appears to be ready to 'soften' any hard BRExit, i.e. airport closures and starving Brits lining up for food rations probably won't be in the news, for humanitarian and financial stability reasons. 99% of the bad part of Brexit will still hit Britain in the years to come, with no good part of Brexit to gain from.
• U.S. Government shutdown - the messaging war is still on. If a solid majority of U.S. voters keeps blaming Trump and Republicans for manufacturing this crisis as a distraction (which would be a factual description) then I'd expect it to end in some sort of compromise that Trump will be able to tout as a win on Fix News. It's unclear at this stage what the timetable is going to be. Weeks of uncertainty, and uncertainty can project to consumers, and consumers can trigger recessions ... at which point the Fed will probably step in so it's not all bad - but not all good either.
  
• China tariffs, another manufactured Trump crisis, are a wildcard. It's still unclear (to me) whether Bolton (the true foreign policy decision maker at the White House) wants to bleed China dry before China gets too strong. With China's economy hurting and the U.S. economy hurting as well there's strong incentives for both sides to make a deal to get back where they started. Assuming rational actors - but Trump & Bolton is all but rational.

The sum of these macro topics is still "worrying" and is projecting a lot of uncertainty, and some of these got a bit worse over the weekend (the government shutdown sabre rattling got more intense) - so barring any breaking news I'd expect some early trading rise of $TSLA on Shanghai Gigafactory hype, which might be corrected back if macro is weak.

U.S. futures indices are, after a +0.7% optimism in Asian markets, currently neutral - i.e. they gave back the Asia gains.

3 out of these 4 macro factors are still very negative and uncertain today.

 

[lklundin]

As a US citizen, the way the EU is trying to blackmail the U.K by getting them to pay billions is shameful. I won't blame the Brits (and would probably encourage them) to give the finger to Brussels and do a hard exit. The $billions they don't pay can be used to improve their own trade and country. Don't put up with Belgium's powerless threats. Just trade with US and China and everyone else and they'll be just fine

If the EU would not stick strictly to the treaty the entire union would quickly unravel.

You see, the EU is just a collection of individual countries each trying to get the most from the EU with the minimum contribution (and this works because of synergy effects).

So the EU's handling of the Brexit is really driven by its need to survive.

To sum up: David Cameron made a very bad bluff and got called on it and now the UK is screwed for at least a couple of generations.

 

[engle]

I doubt the source code will ever set foot in China, or in any car for that matter. Only the compiled code resides within the cars.

The car runs a version of Linux. Somebody published how to root the vehicle with physical access to Tesla's custom interface connector. From there, all of the compiled software binary code and dynamic linked libraries can be extracted for disassembly and analysis. I do agree with an early poster that without access to the machine learning in the cloud from 1 Billion + miles driven around the world, it is useless for FSD.

As an aside, I wonder if a Model 3 Dual Motor's performance can be "upgraded" to Model 3 Performance if rooted into factory mode by an owner? I recall some postings last year around end of Q3 about some Performance buyers delivered Dual Motor versions and it was fixed in software. True?

 

[lklundin]

How about throwing us a backstop bone? A treat, nothing more. Woof. We promise not to bite the hand that offered it.


There is no UK anymore - we are reverting to the Scottish clan system. I am developing my tartan based on a bit of Arsenal red combined with Tesla red. Its a bit garish..

Speaking of the Scottish, who were promised continued EU membership for not leaving the kingdom:

Couldn't they just dissolve the UK (by leaving) and then subsequently inform the EU that they are the successor state to the UK and take over the UK's EU offices ?

That would be epic.

 

[brian45011]

Tesla should be at least notionally pricing their leases at a margin over LIBOR to protect the profit however I haven't found any lease rate information in the Moody's reports for either TALT deal.

I can only offer one anecdotal data point about interest rates on Tesla directly financed leases. In May 2016 we were quoted a money factor of 0.0018 (4.32% ? ) on a 36 month lease. There were various other fees (acquisition,purchase, disposition etc.) but I never tried to calculate the APR% (nor was it disclosed in the quote and the FIT credit was supposedly added back to the market residual value)

I'm not sure, but if I remember it correctly, I think:
-Tesla re-paid in full the orgininal WH credit line in December 2015
-between December 2015 and early August 2016 it was using the Asset Backed Credit Line and/or in-house funds to finance direct leases.
-in August 2016 draws under the 2016 WH Agreement started to be used to fund direct leases.

Unfortunately, it's not TMI but NEI.

 

[SOULPEDL]

You have fundamental misunderstanding of how deep learning works. Even if they get all the "source code", without the data pipeline, the training and maintenance pipeline, it's completely useless

Actually, I was thinking the AI chip itself might be the hurdle. So who says they can't tap the data pipeline and reconstruct the net?
Not trying to sound paranoid, I'm just thinking it might be part of the picture here.

 

[Fact Checking]

Unless the Executive is prepared to back an Act of Parliament necessary to legislate for a second referendum, there is no procedural mechanism for one to occur. If I am wrong about the government's intentions and at the last minute it does implement a 12-month delay to Brexit pending a repeat referendum, my view of the British social psyche is that there would be a bigger majority for Brexit than in 2016. That'd be a question for 2020 however.

Note another procedural hurdle: the only mechanism through which the UK could create time for a new referendum would be by the UK government withdrawing Article 50 before March 29 (which they have the legal power to do) - and there's primary UK legislation in place that forces the government to not do that, and the May government has said that they will not withdraw Article 50.

Any other 'delay' to Brexit can only be done if the EU27 agree, which they indicated very clearly that this won't happen: they want Brexit to be over one way or another. In fact EU27 industry leaders have already indicated it to the EU commission that they are more worried about Brexit concessions than about the economic effects of a hard Brexit.

I.e. the UK's EU-internal competitors (financial institutions, automotive, large manufacturing companies, etc.) are completely happy about the UK unilaterally withdrawing from the EU single market and handing them juicy market share.

In case anyone wonders how a whole political elite is able to misunderstand and screw up the Brexit process should remember the Lehman bankruptcy of 2008: it was so catastrophic to financial markets primarily because the Bush administration didn't understand how damaging a "short-form bankruptcy filing" would be. Once it was filed there was no way back, various automatic legal mechanisms triggered that couldn't be undone despite the Bush administration having regrets immediately afterwards and throwing TARP and Fed trillions to prevent a Great Depression 2.0.

So we should not underestimate the ability of political elites to back themselves into a corner.

 

[Singer3000]

Speaking of the Scottish, who were promised continued EU membership for not leaving the kingdom:

Couldn't they just dissolve the UK (by leaving) and then subsequently inform the EU that they are the successor state to the UK and take over the UK's EU offices ?

That would be epic.

Maybe. But for that you would need 1) there to be a majority in the Scottish parliament for an independence referendum - there is not one; and 2) the UK government to grant the referendum with an Act of Parliament - highly unlikely. All this would need to occur before the UK brexits. So pretty unlikely...

 

[Fact Checking]

Actually, I was thinking the AI chip itself might be the hurdle. So who says they can't tap the data pipeline and reconstruct the net?
Not trying to sound paranoid, I'm just thinking it might be part of the picture here.

You are still misunderstanding how neural networks are structured and created: the FSD neural network itself can be reconstructed from the firmware images already - that's how recent tear-downs of HW 3.0 code were performed.

But that network is the result of a very complex process that involves training on a very powerful supercomputer, and a process of continuously maintaining a complex library of probably over a billion specially labeled training images. By having access to HW 3.0 or even HW 3.0 production facilities competitors are not closer to 'stealing' Tesla's FSD networks: roughly only 0.000001% of the data required to build the FSD network is actually present in the car.

 

[SOULPEDL]

You are still misunderstanding how neural networks are structured and created: the FSD neural network itself can be reconstructed from the firmware images already - that's how recent tear-downs of HW 3.0 code were performed.

But that network is the result of a very complex process that involves training on a very powerful supercomputer, and a process of continuously maintaining a complex library of probably over a billion specially labeled training images. By having access to HW 3.0 or even HW 3.0 production facilities competitors are not closer to 'stealing' Tesla's FSD networks: roughly only 0.000001% of the data required to build the FSD network is actually present in the car.

OK uncle. Tesla can never be hacked.

 

[Singer3000]

Note another procedural hurdle: the only mechanism through which the UK could create time for a new referendum would be by the UK government withdrawing Article 50 before March 29 (which they have the legal power to do) - and there's primary UK legislation in place that forces the government to not do that, and the May government has said that they will not withdraw Article 50.

I'm not a lawyer but there's a question mark on whether primary or secondary legislation would be needed to withdraw Article 50. However your point stands - if the government sought to do this without primary legislation (i.e. a new Act of Parliament) then this would no doubt be subject to legal challenge which would take some time to resolve.

Any other 'delay' to Brexit can only be done if the EU27 agree, which they indicated very clearly that this won't happen: they want Brexit to be over one way or another. In fact EU27 industry leaders have already indicated it to the EU commission that they are more worried about Brexit concessions than about the economic effects of a hard Brexit.

I.e. the UK's EU-internal competitors (financial institutions, automotive, large manufacturing companies, etc.) are completely happy about the UK unilaterally withdrawing from the EU single market and handing them juicy market share.

Maybe. I doubt this personally. If the UK said it wanted to delay to allow another referendum with Remain on the ballot, I expect the EU institutions would fall over themselves to facilitate it. Hard to see a substantial delay for any other reason. A procedural delay of a few months is no big deal and would be worth it to all parties - these nations are above all else strategic allies remember.

Your comments on UK-EU internal competitors are for my money rather simplistic. You may as well say that all multinational companies would rather operate in terms as close to autarky as possible. I don't buy this at all. Pretty much all of the large banks and industrial / manufacturing sector would rather have the flexibility to sell easily into the world's c. 6th largest economy and to easily/cheaply access its labour pool, capital markets and supply chain. The nuance is that they would not wish to pursue this goal to the detriment of the EU's Single Market project. It's this nuance that's been missed on the UK side.

 

[Remus]

OK uncle. Tesla can never be hacked.

Please stop making outrageous claims in the area you know nothing about. Hacked for what? Stealing a couple of trained models that may work for some areas for half year maybe? How is that useful for anyone?

You you mentioned data pipeline being tapped. From where?! Majority of the pipeline have human engineers in the loop involving data gathering, pruning, labelling, training and retraining. Tapped where? The worst case someone stole several hundred GB of image from thousands of TB data repository, how is that useful for anybody?

 

[Matias]

Shanghai's SR RWD Model 3 is going to be like the Volkswagen was in 1938. They are going to sell hundreds of thousands... millions even! Basic family transportation (for middle class I guess)

With Elon's new moustache, I don't think that that is the best possible comparison