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

[KJD]

Except that given the opportunity, hardly anyone wanted the SR. So not drastic to remove it.

Yes I agree 100%. I would also add that we have seen this happen previously.

Remember when the Model S came out and we had the 40, the 60 and the 85. It was obvious that the Model 3 standard range would go the same way as the Model S 40.

 

[neroden]

They have basically been throwing darts at the wall and seeing what sticks.

That does seem to be their style

When they should easily already know what is going to stick.

Fair criticism.

To deny they are desperately against the ropes is to ignore reality.

Ludicrous. They're doing fine --- certainly much better than 2 years ago.

Their biggest problem is that they're stuck at financial-breakeven production rates of 5000/week and they need to get to financial-profit production rates of 10000/week. (And be able to ship them to Europe.) This is a real and serious problem, but for some reason people keep pretending there are other serious problems. This is the only financially serious problem in the short term.

 

[Fact Checking]

And where's your evidence that Tesla has more orders than they can fulfill?

Here is the evidence that Tesla has more Model 3 orders than they can fulfill:

• Until today half of Europe and China couldn't order anything but $60k+ Model 3 configs.
• Still no sales to the other half of Europe. At all.
• Still no right hand drive sales. At all. 35% of the world is using right hand drive cars.
• Still no sales to the rest of the LHD world. Tesla still doesn't sell any Model 3's to @KarenRei, at any price.
• Even the U.S. market is production constrained: in the Q1 delivery report Tesla reported: "Despite pull forward of demand from Q1 2019 into Q4 2018 due to the step down in the federal tax credit, US orders for Model 3 vehicles significantly outpaced what we were able to deliver in Q1."

These are all readily available, easily verified facts.

 

[neroden]

Tesla is being rather quiet about the fact that they're still production constrained, though it occasionally gets mentioned. Maybe they don't want to admit that they're still in production hell, for whatever reason (I suppose it would look bad). I think the evidence is overwhelming, though. If they could have produced 7000 cars/week and shipped them to Europe in January or Feburary, they would have. Clearly they couldn't.

 

[sundaymorning]

I wouldn’t underestimate Tesla’s ability to reach FSD within 3 years. They could very well work on geofencing urban areas such as SF, LA, etc, combine the technology with AP/FSD and it’ll be more advance than anything on the road. If you go to Silicon Valley today, you’ll see a lot of self driving cars being tested by engineers in the passengers seat where their cars would be geofenced in a particular region. Tesla’s fleet doesn’t need to be FSD for the entire geography of the US within 3 years; they just need it working in a few cities in 2022/23 and Uber/Lift will be in a bad position.

 

[T@oo]

I think if it had been the only trim with huge demand, they could have eventually made it work. But when the percentage of people buying it was similar to the S40? No way -- it wouldn't justify setting up the supply chain for the cloth seats. Let alone the downtime from production switching...


I am now absolutely certain of this. I believe that switching betweeen variants was killing production rates. They haven't admitted it but I think every piece of evidence indicates this.

Unfortunately, they haven't solved their problems with production downtime due to variant switching; so they're using the brute-force method of reducing options. That's definitely unfortunate; it means they really don't know how to make a lot of variants on a single production line without major downtime. That's simply something where they are worse at it than the old-line auto companies.

I would really like to you to stop stating facts about things you speculate about and could only have indications about at most. But no evidence or knowlegde. If things you boldly state as facts on such a basis are proven to be not true in near future, it won't make you look too good.

 

[electronblue]

Actually, that's a great deal: $37k happens to be $35k inflation adjusted 2016 dollars.

Not really. The $35k was always a future price, not one you could buy in the past.

 

[neroden]

All facts are derived from evidence. I had a strong hypothesis, and now it's been confirmed with substantial additional evidence.

I would love to get some questions in on the conference call to find out what's causing the production constraints.

 

[Leo9]

I wouldn’t underestimate Tesla’s ability to reach FSD within 3 years. They could very well work on geofencing urban areas such as SF, LA, etc, combine the technology with AP/FSD and it’ll be more advance than anything on the road. If you go to Silicon Valley today, you’ll see a lot of self driving cars being tested by engineers in the passengers seat where their cars would be geofenced in a particular region. Tesla’s fleet doesn’t need to be FSD for the entire geography of the US within 3 years; they just need it working in a few cities in 2022/23 and Uber/Lift will be in a bad position.

Tesla may be able to achieve FSD in less than 3 years, but that doens't mean that regulators will approve the current sensor suite for that. I actually think it's likely that only 8 cameras with no redundancy will not be able to get approval, no matter the neural network advancements.

 

[Artful Dodger]

In the USA, you may be right, but think “Robotistan”. Some place with good roads and signage willing to take a chance to be a world leader. Likely with oil money. Likely in control of the media so they don’t fear backlash from the first casualty. Casualties will happen, though less frequently than with humans in control whom we tend to forgive if they were not being reckless.

The cars are easily shipped. Autonomous cars will be worth more than their owners paid for them, back when they lacked the capability.

Next market session will be a measure of investor intelligence. This *is* a catalyst. Market may or may not work it out immediately.

Yup. Already planned for Dubai:

Dubai wants to make 25% of car trips driverless by 2030, and it’s starting with Tesla taxis

 

[Fact Checking]

Tesla is being rather quiet about the fact that they're still production constrained, though it occasionally gets mentioned. Maybe they don't want to admit that they're still in production hell, for whatever reason (I suppose it would look bad). I think the evidence is overwhelming, though. If they could have produced 7000 cars/week and shipped them to Europe in January or Feburary, they would have. Clearly they couldn't.

That's not an apples to oranges comparison though:

• They shipped almost full spec units to Europe and China: LR packs, AWD drive train that doubles the powertrain installation overhead, the full complex interior - plus performance units that go through a manual line. The drive unit production consumption on those would be 10k drive units per week alone - which they were able to peak-produce in Q3 already, but which nevertheless puts a limit on assembly time and on overall production speed.
• Their latest disclosures shows that they are still "fundamentally" Panasonic cell supply limited at GF1. With the Standard Range Grohmann Machine not coming online until ~March they were consuming ~78 kWh of cells for every unit made, plus they didn't want to completely starve their storage business either.

With more SR units made, and by turning SR+ into a high margin configuration they can now make more units even if cell output grows slower than expected: if Q1 2170 cell output was 62k * 78 kWh (4.84 GWh), then the 55 kWh pack of the SR+ allows up to 87k units in Q2.

But Panasonic also said they upgraded cell output to 35 GWh by the end of March. If that is close to effective capacity, then that's another ~20% increase over the cell output estimated by Carsonight.

I'm sure Fremont has some other bottlenecks down the road, paint shop looming large, but I'm increasingly convinced that the primary bottleneck in Q1 was 2170 cell supply plus maybe hitting a cash low watermark in early March (as @schonelucht hypothesized) when most of the units weren't paid for yet, and the (understandable) unwillingness to build a too large inventory at the end of the quarter.

This explains the 'fundamental constraint' language of Tesla's latest PR release, I believe.

 

[neroden]

With the Standard Range Grohmann Machine not coming online until ~March they were consuming ~78 kWh of cells for every unit made,

Hmmmm.... that is interesting timing. So you're suggesting they were cell-limited during the crucial months of January and February when they were shipping to Europe and Asia, and stopped being cell-limited just in time to switch back to US production (which was obviously suffering a hangover from the tax credit expiration)?

Possssible. Doesn't account for (a) the January downtime, which is known from leaks, (b) the removal of variants from the production line; those still tell me that they're losing uptime due to variant changes. (LR RWD was clearly popular enough, from the "waiting rooms" over on the other side of this forum, so its removal must be for production reasons, not demand reasons.) Doesn't account for that fact that they seem to still be making US production right now, rather than switching back to Euro production, which tells me either they are trying to do variant changes less often or there is a shortage of Euro-spec parts.

We might never know -- if they did have production problems and they solve them during Q2 they probably won't tell us, they'll just announce the new production numbers. If they didn't have production problems at Fremont and the production numbers go up, it'll look exactly the same to us outsiders.

If production remains low in Q2, of course -- with the European backlog still present -- then they *definitely* have production problems.

 

[RobStark]

In the overall market Tesla Model 3 was #1 in March and #4 in Q1 where the next direct competitor Mercedes C Class was #9.

EV Sales: Switzerland March 2019

 

[Tslynk67]

From what I’ve seen, they started off saying none of SpaceX’s stuff was all that great. That, of course, didn’t stick, so they moved on to claiming that he’s just a puppet figurehead and it’s other people doing all the cool stuff.

Well seems they may have conceded that it's not CGI, so now they move onto the profitability question.

This example idiot is part of the $TSLAQ brigade too - and claims he's a "progressive liberal".

 

[T@oo]

Good thing you always have the choice in RL and on the Internet to stay or leave.

That came along exactly different than it was meant. Sorry.

 

[UnknownSoldier]

So how f*cked is the stock price tomorrow? I heard that $265 was a very important resistance level, if it goes below that tomorrow who knows where the floor will be. I was thinking of clearing out a bunch of shares too but now it's looking bad.

The magic number for Chanos, if anyone cares, is around ~$260 a share in order for him to finally turn a profit. He may end up winning the war after all at the rate Tesla is going these days.

 

[BioSehnsucht]

(B) Expect them to attempt to remove differences between Euro, China, and US spec cars. The charge port differences may be forever, but don't be surprised if we start getting Euro-spec taillights in the US, just to minimize downtime on production line switches.

I'd happily pay a reasonable price to have my tail lights upgraded to EU spec amber turn signals. On older vehicles this is easily remedied by importing said parts and swapping them on, but on newer cars even getting the parts can be hard (especially Teslas) and there's no guarantee it will even work without some factory reprogramming. The fact that they continue to ship shitty red signals in NA is a dumb move for the sake of "style" since safety is one of Tesla's selling points.

 

[Prunesquallor]

I think what is unclear is whether Autopilot will also be standard on the 35k model (meaning it will sell for 37k minimum). I would assume it would be if Tesla is to be consistent.

Agree. If they want to maintain the $35k car, offer it with AP disabled.

Unless they simply cannot make a profit and they expect a fair number of people to actually buy them. Then if avoiding those losses is more important than being able to claim the ”35k car", flip the bit and upsell it.

 

[neroden]

So, there is currently a serious threat to my investment thesis for Tesla. (And that for anyone else with a similar model.)

It doesn't depend on any of this autonomy stuff. It doesn't depend on high-trim sales. It doesn't depend on good customer service. It doesn't depend on competent communications. And I am definitely not worried about demand.

It depends on one thing: they have to get to 10k/week production with minimal capex. Having to build the Chinese factory to get to 10k/week was already a setback. Having Fremont stuck at 5k/week is a *serious* setback.

We need to get questions asked at the conference call about production rates and bottlenecks.

 

[Carl Raymond]

Yup. Already planned for Dubai:

Dubai wants to make 25% of car trips driverless by 2030, and it’s starting with Tesla taxis

A question for every voting age person to consider is this:

Which matters more, fewer road deaths, or zero road deaths by robot?

I would like to see it put to referendum. Force everybody to consider it, make the right choice, then collectively accept responsibility for the result. If this is not done, finger pointing and law suits will be the inevitable result of every (less frequent) mishap.