TSLA Market Action: 2018 Investor Roundtable

[Fact Checking]

Most military vehicles occasionally need to remain stationary in the enemy's (potential) line of fire.

Also, very importantly, most military conflicts, even short term ones, are won through logistics and most vehicles are used for logistics

So from an economics POV sales of a couple of thousand units of a light vehicle that is reasonably protected against casual attacks and which platform can also double for special ops use is IMHO a lot more valuable economically than any of the more specialized vehicles intended for open conflict - simply because a lot more of them are required and actually used as well. Bullets flying is basically just 0.01% of the time.

 

[Ugliest1]

It is sad that so many will be affected by the auto industry shift that’s in progress and picking up steam. Here’s the thing, though, if you’re in the industry you should be aware of what’s happening and be taking steps to align yourself for the future. Yes, even 30/40 year career workers. It’s never too late to reinvent yourself. I’ll have little sympathy for those too stubborn to change and move forward or those that keep themselves willfully ignorant.

I agree. It's very difficult, especially for those past middle age, but it's reality. And the time to start is not when it's completely obvious one is in a dying industry. (Some here would argue it's already completely obvious...)

Another example: I'm pissed at the Canadian federal government and the Alberta province for STILL not seeing the alternative energy writing on the wall and mainly harping on how to prop up the oil and gas industry. There are private companies and at least one non-profit in Alberta who are trying to provide a shift ladder (see Iron & Earth - http://www.ironandearth.org/) but they're not getting into the news much, that I've seen anyway.

 

[Krugerrand]

Oh the PC police is out defending Trillions in war spending. You can read for yourself what people have said about unions. Don't want to post OT stuff during market hours.

Tesla now trading 1.5% below NASDAQ.

I didn’t comment about war spending so you can stop that. I did respond to your wrong assumption about me being white collar worker and armchair quarterbacking.

You don’t have to like my real life experiences with unions, but it doesn’t make the experiences less true. Go bark up another tree.

Same ole, same ole TSLA SP movement: up and down, up and down.

 

[tivoboy]

That statement was certainly what you implied in your first comment, not sure why you are denying it now. Here's the comment chain, as a reminder, with your initial response in bold:







What you wrote is simply wrong: the E-Tron can only "take" Tesla Model S sales if Tesla is demand constrained, which it doesn't appear to be.

A very quick example of this "demand > supply" mechanism in action: Tesla was able to redirect dropping sales from China (where they were passing through the up to 65% of tariff overhead to customers) to other regions, without any drop in deliveries - in fact Model S/X deliveries rose and might hit new records in Q4. I.e. the extremely steep China tariffs didn't crowd out Tesla's demand to push sales below available supply of new vehicles.

Audi will only "take" Model S sales if Tesla becomes demand constrained, i.e. if demand drops below available supply. Which is a possibility, but you falsely portrayed it as a certainty resulting from E-Tron sales alone, which is a fundamental misunderstanding (or a deliberate misrepresentation) of how supply & demand works.

The more probable outcome right now is that both the Model S/X and the E-Tron will "take" ICE sales: in particular the E-Tron might be cannibalizing part of their own ICE sales, because existing Audi ICE car owners are the most likely ones to consider another Audi model first. This kind of cannibalization is certainly not what Audi wants, but they'll have to muddle their way through that phase and get to much higher EV production rates if they want any hope of surviving the EV transition.

While what you wrote might not have been FUD, it's pretty obtuse and a pretty boneheaded style of discussion to intentionally not recognize people's very valid counter-arguments ...

Sorry, I never said anything about supply constraint.. I simply said that I'd take the bet that said that a new audi e-tron would have NO IMPACT on tesla model S sales at all, once it the e-tron is out. That's it. The supply constraint argument was brought up by others who were trying to support the S/D argument as opposed to the - would a possible Model S buyer, actually buy the e-tron instead. Which they will.

But, you're all right. As usual. And I'm all wrong, as usual. And as usual, I'm off to the bank. $$$

And, I'm FAR from the person who isn't recognizing other peoples counter arguments. By a county mile.

 

[Fact Checking]

That's still Tesla/Musk making the claim, just like I said. Bannon is part of Tesla, not so?

Even if we take the 10x speed up at face value, that still leaves the much more important question : is it enough? And if it is enough, is their software stack able to take advantage of that speed up in a way that brings material progress towards FSD?

Those are all fair points, and I have attempted to reply to these questions in my reply to @humbaba.

TL;DR: I believe there's a big difference between Elon's old FSD statements and the latest very specific statements based on actual HW3 test units and actual real HW3 neural networks, and I find the latter very credible. Elon's a hopeless optimist when it comes to judging the progress of future complex R&D, but when he says a specific figure of "about 10-fold speedup" based on field test units I've yet to see him actively lie or mislead.

Yes, the sources I quote are from Tesla and I'm speculating a bit on top of that - but given how relevant FSD is to Tesla's valuation I believe it would be irresponsible not to speculate.

 

[jhm]

Also, very importantly, most military conflicts, even short term ones, are won through logistics and most vehicles are used for logistics

So from an economics POV sales of a couple of thousand units of a light vehicle that is reasonably protected against casual attacks and which platform can also double for special ops use is IMHO a lot more valuable economically than any of the more specialized vehicles intended for open conflict - simply because a lot more of them are required and actually used as well. Bullets flying is basically just 0.01% of the time.

Additionally, transport of fuel is one of the most dangerous duties in battle. The logistical requirements for fuel are quite substantial and costly in life and treasure.

 

[Joe F]

Well it's 8AM here right now, still at home and just done a Twitter pass, as you're keeping notes...

Not keeping notes. It's 11:35 here and just getting caught up with the overnight posts.

I kid as I'm retired so TMC is always on here, across multiple devices. Twitter too. I don't post much, but try to keep on top of most topics. Things get too serious at times and I find if one can't make someone laugh or at least smile daily, one wastes one of, if not the single most important aspect of human interaction.

TSLA down but mirroring NASDAQ, and flirting with the upper BB.

 

[Unpilot]

Sorry, I never said anything about supply constraint.. I simply said that I'd take the bet that said that a new audi e-tron would have NO IMPACT on tesla model S sales at all, once it the e-tron is out. That's it. The supply constraint argument was brought up by others who were trying to support the S/D argument as opposed to the - would a possible Model S buyer, actually buy the e-tron instead. Which they will.

But, you're all right. As usual. And I'm all wrong, as usual. And as usual, I'm off to the bank. $$$

And, I'm FAR from the person who isn't recognizing other peoples counter arguments. By a county mile.

That's a very trumpian thing to say.

 

[UncaNed]

Don’t be dramatic, passive aggressive and put words in people’s mouths. Nobody said all union workers are lazy.

The issue should be about union's priorities; are they looking out for themselves (what's good for unions) - or are they looking out for the actual WORKERS (unions, IMHO, often DO NOT). Those who disagree need only use the Googles to learn of pensions dissappeared at the hands of union management, workers trampled on by unions in so many ways... Not saying unions aren't needed in some cases, just that I see no use for unions in their present form in the US. I'm told that unions are very different in countries like Germany, that kind of union wouldn't hurt Tesla so much, but I don't believe that's the kind of union we'd have anywhere in the US.

 

[Fact Checking]

would a possible Model S buyer, actually buy the e-tron instead. Which they will.

Again, and about half a dozen people tried to explain this to you as well, a customer deciding to buy an E-Tron instead of buying a Model S is not "taking" sales from Tesla and is not reducing Tesla sales as long as Tesla has higher demand than supply.

So your original statement of "it'll take some of the Tesla S sales for sure and it won't be 1-3%" is an active misconstruing of the underlying mechanism.

There's a huge difference between the E-Tron and Tesla splitting up demand among themselves and still not being able to supply enough units and still not being demand constrained, versus demand constrained competitors taking sales from each other in a zero-sum competitive environment.

This isn't a difficult concept, will you finally concede this very simple argument?

 

[humbaba]

Those are all fair points, and I have attempted to reply to these questions in my reply to @humbaba.

TL;DR: I believe there's a big difference between Elon's old FSD statements and the latest very specific statements based on actual HW3 test units and actual real HW3 neural networks, and I find the latter very credible. Elon's a hopeless optimist when it comes to judging the progress of future complex R&D, but when he says a specific figure of "about 10-fold speedup" based on field test units I've yet to see him actively lie or mislead.

Yes, the sources I quote are from Tesla and I'm speculating a bit on top of that - but given how relevant FSD is to Tesla's valuation I believe it would be irresponsible not to speculate.

I'm certainly not one to disagree with the 10x speed up. That's a statement of fact, not opinion, and I think Elon & other's at Tesla can be trusted for that. Its the utility of the speed up, and what that actually means. IOW, is that really enough on its own to enable FSD? Was the only thing holding them back the lack of hardware speed? I think not, but at the least I don't think its controversial to say there is a disconnect between "10x faster" and "capable of FSD". One does not grant the other.

 

[humbaba]

The volume today is so low that not even the shorts are doing anything -- the SP is a pretty close mirror to NASDAQ, just shifted down ~1%. Ho hum, were in a doldrum.

 

[gtrplyr1]

Audi's e-tron is 2 years away from being available and that's probably optimistic. 2 years in this fast developing industry may as well be a lifetime, so I'm not going to start measuring the potential impact on Tesla's sales just yet. I do think it's wise to keep up with what is going on with the competition but to compare Audi's offering to what Tesla offers RIGHT NOW is not fair. I have no doubt the engineers and designers at Tesla are hard at work improving and developing new and better ideas as I type this.

The other thing that get's lost in the shuffle is the fact that the market for EV's is growing rapidly and there is no way 1 manufacturer can supply all of the cars that will be needed .... there will be many winners in this space and I'm pretty sure Tesla will be among them.

Cheers to the longs

 

[Fact Checking]

IOW, is that really enough on its own to enable FSD? Was the only thing holding them back the lack of hardware speed?

That's why I quoted Andrej Karpathy whose statements gave me the impression that they already have their FSD neural networks trained and are only waiting for the new hardware to be available to customers:

Andrej Karpathy:

"So in other words, we are currently at a place where we trained large neural networks that work very well, but we are not able to deploy them to the fleet due to computational constraints. So, all of this will change with the next iteration of the hardware. And it's a massive step improvement in the compute capability. And the team is incredibly excited to get these networks out there."​

The tear-down of v9 by jimmy_d I linked to independently supports that view as well.

(It's understandable that Tesla is being a bit coy about this, they wouldn't want to overly impress people about HW3's capabilities, to not trigger deferred purchases until March when owners can get HW3 'for free'. So this could be one of the rare instances where Elon and Tesla is sandbagging the capabilities.)

 

[Open Interest]

It is sad that so many will be affected by the auto industry shift that’s in progress and picking up steam. Here’s the thing, though, if you’re in the industry you should be aware of what’s happening and be taking steps to align yourself for the future. Yes, even 30/40 year career workers. It’s never too late to reinvent yourself. I’ll have little sympathy for those too stubborn to change and move forward or those that keep themselves willfully ignorant.

I’ve always thought the same way about the coal industry. Instead of enabling workers to be victims of a dying industry why not help them train for the future? But entitlement runs deep, and change really is hard.

 

[Open Interest]

Why does the price fall like a rock? No more believe in Tesla? hm

I’ve seen Tessie fall like a rock. This is more like a feather in the wind.

 

[Menifeer]

Additionally, transport of fuel is one of the most dangerous duties in battle. The logistical requirements for fuel are quite substantial and costly in life and treasure.

How would you propose to recharge a military EV on the battle front?

 

[Fact Checking]

Instead of enabling workers to be victims of a dying industry why not help them train for the future? But entitlement runs deep, and change really is hard.

That's pretty close to the federal program that Hillary Clinton proposed as a presidential candidate (she's a policy nerd after all) - but obviously most of the coal mining communities are deep red, are biased against and suspicious of federal programs, and are also essentially in an echo chamber of conservative media and cut off from sensible arguments. Any such proposals will be presented by and seen through the distortion lens of Fox News and conservative talk radio.

Also, in practice such programs have to be pretty generous to make a real difference, and it will often leave out older people and forced change is hard to all of us.

It also doesn't help that many urban liberals have and often show open contempt for the 'hillbillies' and have absolutely no idea about how rural life looks like in such communities. They are quite correctly seen as out of touch.

This is why convicted felon coal baron who spent a year in prison for essentially murdering 29 miners because his company intentionally skipped mandatory safety equipment to increase profits (Don Blankenship) is still very popular with miners. Just being of the community and speaking their language cancelled out his awful, murderous track record.

It's very hard to break through that wall of ignorance and isolation, mixed with (justified!) economic desperation and distrust. Most of these are high unemployment regions with very low per capita GDP and a shrinking, aging population.

So it's not about 'entitlement' at all, people are quite literally trapped in those regions in several senses of the word.

Probably no easy solutions but to wait for time to solve it birth by birth and coffin by coffin I'm afraid.

 

[AWDtsla]

That's why I quoted Andrej Karpathy whose statements gave me the impression that they already have their FSD neural networks trained and are only waiting for the new hardware to be available to customers:

Andrej Karpathy:

"So in other words, we are currently at a place where we trained large neural networks that work very well, but we are not able to deploy them to the fleet due to computational constraints. So, all of this will change with the next iteration of the hardware. And it's a massive step improvement in the compute capability. And the team is incredibly excited to get these networks out there."​

The tear-down of v9 by jimmy_d I linked to independently supports that view as well.

(It's understandable that Tesla is being a bit coy about this, they wouldn't want to overly impress people about HW3's capabilities, to not trigger deferred purchases until March when owners can get HW3 'for free'. So this could be one of the rare instances where Elon and Tesla is sandbagging the capabilities.)

You're way too optimistic here. None of these statements implies FSD is imminent here, for any definition of "Full". My prediction is onramp to offramp will (someday) work reliably with HW3 (in mostly good weather). Beyond that is a huge question mark.

 

[humbaba]

That's why I quoted Andrej Karpathy whose statements gave me the impression that they already have their FSD neural networks trained and are only waiting for the new hardware to be available to customers:

Andrej Karpathy:

"So in other words, we are currently at a place where we trained large neural networks that work very well, but we are not able to deploy them to the fleet due to computational constraints. So, all of this will change with the next iteration of the hardware. And it's a massive step improvement in the compute capability. And the team is incredibly excited to get these networks out there."​

The tear-down of v9 by jimmy_d I linked to independently supports that view as well.

(It's understandable that Tesla is being a bit coy about this, they wouldn't want to overly impress people about HW3's capabilities, to not trigger deferred purchases until March when owners can get HW3 'for free'. So this could be one of the rare instances where Elon and Tesla is sandbagging the capabilities.)

See, that just says that they needed the hardware for the large neural networks, not that those neural networks would be sufficient in and of themselves for FSD. And contending that they are is ignoring many issues, not least of which FSD requires more than radar/lidar/sonar + steering/accelerating/braking. The neural networks are for vision, a necessary component and maybe the vision has gotten "good enough". But even if that is the case, it is not sufficient to enable FSD.

And I go back to my point about incremental deployment of FSD-like features. Which isn't even speculative -- that is exactly what Tesla has been doing. Each new feature is another component supporting an FSD solution.

For market purposes I think it is sufficient to identify that Tesla:

1. is actively working on FSD
2. that they have more instrumented miles driven
3. develop cutting edge hardware
4. prove out their features in a semi-public fashion

Waymo isn't really a competitor to Tesla, but in comparison all they have to show is #1, though they are making some progress on #4. Certainly any competitor could do #3, but no one comes even close on #2.

I saw mention recently of Waymo and it was bragging on their exponential growth of instrumented miles driven. Which is not only a minute speck of a fraction of Tesla's, but Tesla's is also growing exponentially. For me, #2 is probably the most persuasive reason for believing that Tesla will achieve FSD first.