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

[Sean Wagner]

I believe Jeff "Who" Bezos could personally finance a project as expensive as the Apollo one.

In comparable outcome maybe, which is amazing.
In human effort - if I remember correctly, at the peak of Apollo roughly 500'000 people [!] were involved in the program.

 

[mad474]

Quite happy for Elon to follow the Ferengi rules of acquisition. Rule number 1:

Once you have their money, you never give it back.

$2Bn is just for starters.

Rule 45:

Expand or die

Rule 95

Expand or die

So good, they said it twice.

Yeah, and not to forget Rule #5 :
Always exaggerate your estimates.

(running for cover)

PS: Ferengi Rules of Acquisition

 

[Sean Wagner]

One more thing before I forget. The discussion about battery cell supplies is important, and I liked the limited information regarding GF3 - rigorous qualification of supplied cells, and multiple suppliers. Best way to minimize hitches and encourage competitive bids.

I thought it was clear that getting sufficient cells in China would be a challenge, what with government mandates and the flood of EVs required. See here, though:

China's BYD halts work at electric bus factory
Shutdown at country's largest EV maker follows big subsidy rollback
TAKASHI KAWAKAMI | MARCH 13, 2019
China's BYD halts work at electric bus factory

BYD also manufactures a shedload of cells. But not sure they're of a chemistry useful to Tesla.

 

[KarenRei]

Also worth considering is how much more wealthy the World as a whole and a few single individuals have become.

I believe Jeff "Who" Bezos could personally finance a project as expensive as the Apollo one.

Well... it was $25,2B, but that was in 1960s dollars. Using the CPI inflation index it's $112B in 2018 dollars, but CPI isn't the inflation index you use for space cost estimation - it's NNSI (Nasa New Start Index), which is a much higher inflation rate (CPI is based on consumer goods, whose prices have benefited from globalization and mass manufacturing, but thusfar, rocketry hasn't... rockets are still largely designed and built by hand by highly trained, skilled domestic labour). Jeff could not afford that.

Thankfully, today we stand on the shoulders of giants, so a new Apollo programme would be vastly cheaper than the original. If managed properly.

Anyway, I've noticed a not-so-subtle mod hint that we're supposed to be refraining from rocketry discussions, so I'll duck out of this conversation. (The conversation seems to be approaching its periapsis, so if I want to escape it, now would be the best time if I want to maximize the Oberth effect )

—

What goes for ICBM’s goes for Apollo rockets too.

Mod

 

[lklundin]

it's twitter. who cares.

this is a company whose cars feature a "fart mode", And you're mad that their Twitter handle is being silly?

Actually, in a nod to the incumbents it is "Emissions Testing Mode".

 

[Pras]

Tesla tweets
———————
As if there was any doubt, the late off hour tweets and the style confirms, at least to me, that it is EM.

The question is what it means.

1) hopefully it bridges the near 7 fold gap between EM twitter followership vs Tesla’s. So this could merely be a push from EM to make Tesla twitter more visible. The idea is, if you want a Tesla news or have a question, go to @tesla, not @EM (as has been the case for most Tesla complaints and comments).

2) it is a bait trap for SEC to poke into Tesla and media to write about the Tesla tweets. The former is a risky proposition but at some point the SEC can’t ban the CEO or any employee to “take up” the comms role within Tesla. A comms person is regularly certified, I guess, to talk freely as Tesla representative. It likely also bypasses the settlement signed (would like to know any legal expert’s opinion on this).

3) it could also be a shared responsibility within Tesla, e.g. there is a permanent person sitting next to EM or others who can tweet, effectively a sign off process.

Let’s see how it evolves.
******************************


FSD
————-
On the FSD, I hope that Tesla keeps posting updates more regularly (via twitter!?). Among other things, it should announce a pilot location where it can deploy test cars using development version softwares, hire a few test drivers (Waymo type), and launch the TN / app soon (hopefully).

 

[lklundin]

Tow hitch option added in Europe. About $1k.

This is going to be very popular - and that $1k is surely overwhelmingly profit. Might add $200 or so to European ASPs? With that I estimate a 2-3% increase in Model 3 profits in Europe and a 0,15-0,2% increase in European Model 3 margins.

I for one will be getting it Need to downsize my trailer, though, but I had planned on that regardless.

The offer also applies to orders for which delivery is still pending.

As such I added this option to my order, for Germany the cost is 1060 €.

PS. The tow bar adapter is (naturally) removable.
PPS. For comparison, some years ago I had a (removable) tow bar - beautifully made in aluminium - retrofitted to my Audi A8 for about 2k €.

 

[lklundin]

And there’s the funny thing. We know all about the march of nines. Regardless of readiness for the cars to ‘go it alone’, a year from now the cars will have an amazing array of party tricks. e.g. 4000 cars driving on and off boats, cars coming off the assembly line and performing their own stage one track test, cars in open spaces doing weaving manoeuvres, etc. Some vids will go viral.

Joe Average (who may not know the diff between mean and median) will be convinced that FSD is imminent. Neroden will be adamant it’s still years away and I imagine he’s right. What I can’t imagine is Joe Average reading or accepting anything written by Neroden.

With respect to the value of FSD and its effect on TSLA, perception is reality.

On the topic of Joe Average, it is becoming increasingly clear to me that one major FSD task still remains,
namely handling of the handover of the driving between the human driver and the FSD computer.

Consider the scenario where Joe Average has enabled FSD and is dutifully monitoring the car's driving with his hands on the wheel, when the car initiates a maneuver that Joe Average did not expect.

Two things can be happening:
1) The car is making a mistake (e.g. has misidentified the lane markings) and is about to cause an accident.
The action required by Joe Average (potentially in order to save his life), is to take control of the vehicle
- and to do so on basically split-second notice.
2) The car has detected that an accident is about to happen and is maneuvering (in potentially the only possible way) to avoid this accident.
The action required by Joe Average (potentially in order to save his life), is to _not_ take control of the vehicle
- and to do so on basically split-second notice.

I have to wonder what Tesla is doing (and what they can do), in order to reduce the risk that Joe Average gets into an accident due to false negatives and false positives in the above two alternative scenarios.

I believe this is a non-trivial task, with the recent crashes of the Boeing 737 coming to mind.

 

[Carl Raymond]

On the topic of Joe Average, it is becoming increasingly clear to me that one major FSD task still remains,
namely handling of the handover of the driving between the human driver and the FSD computer.

Consider the scenario where Joe Average has enabled FSD and is dutifully monitoring the car's driving with his hands on the wheel, when the car initiates a maneuver that Joe Average did not expect.

Two things can be happening:
1) The car is making a mistake (e.g. has misidentified the lane markings) and is about to cause an accident.
The action required by Joe Average (potentially in order to save his life), is to take control of the vehicle
- and to do so on basically split-second notice.
2) The car has detected that an accident is about to happen and is maneuvering (in potentially the only possible way) to avoid this accident.
The action required by Joe Average (potentially in order to save his life), is to _not_ take control of the vehicle
- and to do so on basically split-second notice.

I have to wonder what Tesla is doing (and what they can do), in order to reduce the risk that Joe Average gets into an accident due to false negatives and false positives in the above two alternative scenarios.

I believe this is a non-trivial task, with the recent crashes of the Boeing 737 coming to mind.

We have to trust that FSD will not initiate any sudden manoeuvres unless it’s certain the manoeuvre is required. It is likely the manoeuvre will be complete before Joe even begins to react.

Trust in new machines is not unprecedented. Lifts, traffic lights, aeroplanes, etc, would all have been quite scary for early adopters.

 

[Sancho]

And there’s the funny thing. We know all about the march of nines. Regardless of readiness for the cars to ‘go it alone’, a year from now the cars will have an amazing array of party tricks. e.g. 4000 cars driving on and off boats, cars coming off the assembly line and performing their own stage one track test, cars in open spaces doing weaving manoeuvres, etc. Some vids will go viral.

Joe Average (who may not know the diff between mean and median) will be convinced that FSD is imminent. Neroden will be adamant it’s still years away and I imagine he’s right. What I can’t imagine is Joe Average reading or accepting anything written by Neroden.

With respect to the value of FSD and its effect on TSLA, perception is reality.

It’s going to get fairly weird when the average perception of the value of TN rights to a Tesla car is, say, $10,000. At that point, either (1) Tesla will greatly increase the cost of the FSD option, generating unprecedentedly high margins, or (2) the effective price of a new Tesla car net the TN value becomes unbelievably attractive vs. all alternatives.

But that “fairly weird” scenario doesn’t seem to be that far off. Given the very high projected value of TN per car, $10,000 represents something less than a 10% chance of it happening. How much more FSD functionality do we need Tesla to demonstrate before we give Tesla Network a 10% chance of it happening in a few years’ timeframe?

 

[Dan Detweiler]

...and in related news. The competition for ride share services seems to be getting a little testy.

Uber and Lyft drivers plan to strike to protest low wages, job insecurity

Dan

 

[Carl Raymond]

It’s going to get fairly weird when the average perception of the value of TN rights to a Tesla car is, say, $10,000. At that point, either (1) Tesla will greatly increase the cost of the FSD option, generating unprecedentedly high margins, or (2) the effective price of a new Tesla car net the TN value becomes unbelievably attractive vs. all alternatives.

But that “fairly weird” scenario doesn’t seem to be that far off. Given the very high projected value of TN per car, $10,000 represents something less than a 10% chance of it happening. How much more FSD functionality do we need Tesla to demonstrate before we give Tesla Network a 10% chance of it happening in a few years’ timeframe?

FSD will hopefully never (pre Tesla driverless network) be priced higher than an optimistic evaluation of what it is worth to buyers at the time of purchase. If Tesla were to get greedy and price in future_value x probability, there is always the risk the dice don’t roll their way. This would leave a lot of sour grapes. The damage to reputation make it not worth the risk. We can be happy enough that the FSD revenue will continue to climb based on driver assisting utility alone.

We can also be happy knowing that Mr Market determines the price of TSLA. The effect of greed here will not be in within Tesla’s power to restrain.

 

[Cobbler]

All those experts commenting on AI/Neural Nets/... posting here instead somewhere else.

Maybe they can create a Neural Net to decide if the post is market related or not?
Plenty of data here to teach it!

When can we expect Maxwell acquisition news?

 

[saniflash]

round closed

https://ir.tesla.com/node/19861/html

 

[Sancho]

FSD will hopefully never (pre Tesla driverless network) be priced higher than an optimistic evaluation of what it is worth to buyers at the time of purchase. If Tesla were to get greedy and price in future_value x probability, there is always the risk the dice don’t roll their way. This would leave a lot of sour grapes. The damage to reputation make it not worth the risk. We can be happy enough that the FSD revenue will continue to climb based on driver assisting utility alone.

We can also be happy knowing that Mr Market determines the price of TSLA. The effect of greed here will not be in within Tesla’s power to restrain.

Agreed: Tesla is unlikely to price FSD more than current actual value. But that is why Musk may be right in his assertion that a person is “absolutely insane” to buy any car now other than a Tesla.

Can we also agree that most people are not insane? The only reason that anyone is still buying a non-Tesla car is that they either haven’t yet heard of TN, or they give it a very remote chance of happening.

What happens if Musk fulfills his projection that FSD is feature complete by the end of the year, and demonstrates that to experts and the public? Many people will still be unconvinced (not just Neroden), but many other people will start to give TN a real chance of happening. THEN things start to get weird (in a good way for Tesla and hopefully TSLA).

My point is, we don’t need TN to actually happen to win. We win when TN is perceived as having say, a 25% chance of happening. Because given how absurdly valuable TN rights could be, people will then really be insane to buy any other car.

 

[TheMrX]

The offer also applies to orders for which delivery is still pending.

As such I added this option to my order, for Germany the cost is 1060 €.

PS. The tow bar adapter is (naturally) removable.
PPS. For comparison, some years ago I had a (removable) tow bar - beautifully made in aluminium - retrofitted to my Audi A8 for about 2k €.

As much as I'm happy for you, I'm also really jealous.

I think I'm about to cancel my order and tell them that I need an european model 3 : CCS (I know, not the same in NA) and towing should be available everywhere.

 

[mutle]

VW is currently live streaming their presentation of the new ID cars: The all-new ID. Pre-booking Launch Press-Conference

 

[KarenRei]

VW is currently live streaming their presentation of the new ID cars: The all-new ID. Pre-booking Launch Press-Conference

Will compare with Model 3 SR+ (in brackets) in a second.

• First version: "I.D. 3 1st":
• 58kWh [vs. ~50kWh]
• 420km WLTP (ad later on says "drive up to 550km on one charge" - NEDC? Later variant?) [vs. 415km WLTP. So about 20% higher energy consumption per unit distance for the I.D., which is about what I'd expect from that shape]
• <€40k [vs. €44,5k w/German VAT - assuming that ID is priced with German VAT]
• 30k of the first version being made (unclear of what timeframe - sounds like perhaps over the course of the first year of production?)
• Production start: end of this year.
• "More than 100k per year" will be achieved "over the coming years". [vs. 500k target over the next year, starting this quarter]. 100k/yr "In a running year, not in 2020", although a new car to be introduced next year.
• At plant completion: "Plant has a capacity of 300k cars" (multiple models combined).
• I didn't catch charge rates, just that it'll work with Ionity.
• He refused to answer a reporter question about performance.

Not mentioned, but we know from previous reporting: they lose ~€3k per vehicle, instead of making several thousand € like the Model 3 SR+ And the latter is the price with EAP included. Doing a bit of math, that suggests that their non-VAT production cost is around €36,3k / $40,7k USD (assuming that "under €40k" means "right about €40k", as it usually does)

One big point of kudos to Jürgen: he took the time to diss fuel cells in response to a reporter question Pointed out that given the state of the technology, they simply cannot make a big enough impact in the coming decade, whereas VW wants to make a big impact now.

 

[elasalle]

 

[T@oo]

Will compare with Model 3 SR+ (in brackets) in a second.

• First version: I.D. 3 1st:
• 58kWh [vs. ~50kWh]
• 420km WLTP (ad later on says "drive up to 550km on one charge" - NEDC? Later variant?) [vs. 415km WLTP. So about 20% higher energy consumption per unit distance, which is about what I'd expect from that shape]
• <€40k [vs. €44,5k w/German VAT - assuming that ID is priced with German VAT]
• 30k of the first version being made (unclear of what timeframe)
• Production start: end of this year.
• "More than 100k per year" will be achieved "over the coming years". [vs. 500k target over the next year, starting this quarter]
• I didn't catch charge rates, just that it'll work with Ionity.
• I didn't catch performance specs.

Not mentioned, but we know from previous reporting: they lose ~€3k per vehicle, instead of making several thousand € like the Model 3

• 420km WLPT is for the 58kWh variant
• 550km WLTP for the 77kWh
• Charge rates up to 120kW
• Augmentented Reality HUD (optional)
  
• Voice Commands in natural language
• Matrix LED headlamps
• Orders start in April 2020
• Deliveries Summer 2020
• no tow hitch mentioned
• size of a VW Golf outside, inside larger
• glas roof (full size?)