2017 Investor Roundtable:General Discussion

[schonelucht]

The fact is that it cost you only the cost of power and ink to produce that neighborhood leaflet, plus wear and tear on the printer.

And the costs associated with the huge building around it. And the costs associated with calibrating the damn thing for one run. And the fact that you can just ask 100 copies but that it immediately spits out a few thousands at a minimum. And that you need a crew of 10 to run it instead of the single guy pressing a button on the copier. etc. etc.

Maybe you don't have any experience with manufacturing through processes that are set up for scale? The engineering reality is that you can't scale them down and expect running costs per unit to stay the same. It just doesn't work that way.

Anyway. I don't think we are getting any further. Let's wait and see what happens next quarter. I think I said the same thing to @vgrinshpun when I had about the same discussion with him last quarter. We've been waiting for how many quarters now for the TE ramp up now? One more will hurt no-one, right.

 

[neroden]

I have no references to add to the subject, but the quote specifically says he led 'battery cell material sourcing' so someone has made an error.

edit: I also incidentally am not proposing cobalt has anything to do with current output.

Well, I can't find the reference which I remember in the past.... to Tesla hiring a "director of global supply chain" to handle such things.

But Tesla's management is frankly a bit chaotic. Who knows, several people may have been working on battery cell material sourcing.

 

[schonelucht]

I think it's not a great gross margin, but they've been pretty disciplined about not selling below marginal cost, since the early Roadster problems.

You mean like service & maintenance, the zero-profit center that is now second quarter in a row generating huge gross margin losses?

 

[neroden]

The engineering reality is that you can't scale them down and expect running costs per unit to stay the same. It just doesn't work that way.

See, now you're making a different argument. This is the argument that they are paying for workers to operate at high volume while not actually doing so. (Recurring fixed costs.) This is *not* the same as the depreciation. The equipment doesn't deteriorate faster per battery if you only produce one battery. The people do. (Another argument for automation...)

This is quite definitively happening on the Model 3. Maybe it's happening with the stationary batteries too. Actually, correct that, with only 40% uptime, it's almost certainly happening.

Now, based on accounting principles, *none of this should be in the gross margin*. But unfortunately it probably is.

I suspect this isn't enough to bring gross margins actually negative, but it's hard to tell because we really don't have number broken out. In addition to a definite muddle regarding how much of the battery revenue and cost is allocated to "SolarCity" in Q2, the recurring fixed costs for the stationary batteries are impossible to truly distinguish from the Model 3 battery costs: they're all Gigafactory operations costs; I seriously doubt they're allocated between them in any rational fashion; the moment you get into cost allocation, accounting usually gets sloppy.

I'm not sure it matters, since your contention is that it's a negative gross margin and mine is that it's an insignificant positive gross margin.

Anyway. I don't think we are getting any further. Let's wait and see what happens next quarter. I think I said the same thing to @vgrinshpun when I had about the same discussion with him last quarter. We've been waiting for how many quarters now for the TE ramp up now? One more will hurt no-one, right.

Sure, let's wait for the next quarter. I think the number to actually ask Musk about is that uptime percentage. As long as it's 40% that's bad.

 

[schonelucht]

Schonelucht apparently thinks this means the Gigafactory batteries are more expensive than buying off the market long-term, but Musk and Straubel have said absolutely nothing to indicate that.

I think this all points to a problem with the production ramp.

Except that JB Straubel is on record saying that cell production will not be the problem and that Elon is on record that cell production at the gigafactory is already right now surpassing production at any other factory in the world.

 

[neroden]

You mean like service & maintenance, the zero-profit center that is now second quarter in a row generating huge gross margin losses?

Two points there:
(1) They underpriced the extended warranty and service plans early on, as I stated at the time. (Lucky me, I got them.) However, the underpriced plans won't roll off the books until 2022.
(2) There's something completely fishy about the accounting for this. Properly speaking, only the variable costs should be included. So none of the service center rent or utilities cost, none of the salaries (IIRC they have a lot of salaried workers), none of that -- all of that should be in SG&A.

From what I've heard, that's not how the accounting is being done. How do you get $271 million in *parts and hourly labor* costs?

When substantial amounts of fixed recurring costs end up in gross margin, it breaks it as a measurement. Economically speaking it's supposed to exclude those but that's not always accounting practice.

 

[Yggdrasill]

Except that JB Straubel is on record saying that cell production will not be the problem and that Elon is on record that cell production at the gigafactory is already right now surpassing production at any other factory in the world.

If they're currently producing 2 GWh/quarter, that's insufficient for Q4 Model 3 production. We don't know the mix of NMC/NCA cells currently being produced, but it is quite possible the production ramp has been going badly, and NMC cell production is being sacrificed for NCA cell production, to avoid impacting the Model 3 production ramp.

In that case it makes sense to cover the shortfall with Samsung cells, until the production ramp can catch up with cell demand. I would have liked to see the cell production ramp faster, so that this wouldn't be an issue, but if this is what's happening, Tesla seems to be making the right decisions.

 

[jeewee3000]

Production ramp regarding battery cells or not:

1) The use of Samsung 2170 cells - which are manufactured as per Tesla's designs and instructions - in the powerpacks installed in Australia (put together in the GF), shows that this deal has been in the works for a long time. At least 6 months ago, but most likely more. So it's not like Tesla panicked and quickly ordered some cells from Samsung. This partnership has existed for some time now.

2) The benefits of the above partnership are:
- multiple suppliers of the same cells (Panasonic, Samsung and maybe others) keeps suppliers competing/honest and gives you backup supply;
- a larger part of the global battery market gets fed to Tesla. Other car companies looking for battery suppliers to produce BEV's have less options, and those options will be more expensive -> competitive advantage for Tesla.
-...

3) The downsides of the above partnership are:
- purchasing cells from Samsung is more expensive than producing those cells in-house.

But as Tesla's battery demand is expanding more rapidly than their battery production, it is a good thing they get in business with more suppliers. It means faster growth for TSLA, period.

So I'm not worried too much about the whole "Is the GF having production problems?".

 

[schonelucht]

See, now you're making a different argument. This is the argument that they are paying for workers to operate at high volume while not actually doing so. (Recurring fixed costs.) This is *not* the same as the depreciation. The equipment doesn't deteriorate faster per battery if you only produce one battery. The people do. (Another argument for automation...)

Ugh. That's one major misunderstanding then. When I was saying "that's just how costs works", I obviously meant all costs, not just depreciation but also running costs.

I suspect this isn't enough to bring gross margins actually negative, but it's hard to tell because we really don't have number broken out.

So now the crucial question is : why do you think that?

Why is it so impossible to think that early in a ramp up of TE, gross margins were negative? Q4 conference call, Elon was very clear that initial Model 3 was going to have horrible negative margins. (Tesla Model 3: 4 New Things to Know -- The Motley Fool) Why would that hold for the car but not for early powerpacks? Why is it so impossible to assume that maybe automation isn't fully calibrated and working at full speed? That there are batches of cells that were out of spec because internal procedures were not yet optimized for how to correct for certain conditions early in the manufacturing process? That they need to do more quality control testing than usual on the first products of the line? That more product needs to be set aside for long term verification of the current batches? That some of the labour that needs to happen hasn't yet become routing for the personal and therefore takes longer than it will in a few months. Etc. etc. But no, instead it must be the official SEC reporting that somehow isn't representative.

It's some kind of mantra on here : nop, Tesla's gross margin must be positive (and I use that term exactly within the context of GAAP as reported in official filings because despite all it's failings it is still at least somewhat defined) I don't even understand why it would be controversial : I told repeatadly that I think negative gross margins today do not indicate mean they will continue to lose money on them tomorrow. On the contrary. It's just a metric. But seems to be one that has taken some holy significance on the board "Thou shalt not be negative". I don't get it. Elon doesn't have that taboo, see above. Neither should you.

 

[schonelucht]

So two things are going to happen:

(1) Tesla is going to focus on its two competitive advantages:
(a) the "integrated / turnkey system"
(b) the cooled/warmed/fireproofed/stabilized pack technology, where they are *way ahead* of competitors

Citation needed. How is Tesla, for example, way ahead in those technologies than AES? If Tesla provides an unparalleled turnkey integrated system, why is it relying on AMS to offer things like virtual grids? Why is it relying on AMS at all? Do you think Altagas didn't integrate their 80MWh battery plant on site at their gas peaker plant as well as Tesla's powerpack would do it out of the box? In fact, why do you think Tesla understands better how to integrate with utilities than companies like Siemens, GE or ABB that have been traditional suppliers in these markets for decades?

(2) Tesla will buy parts from the outside as much as possible and only in-house when they have to

Maybe, maybe not. Tesla seems to like vertical integration. But that just makes it more easier for other parties to copy their product.

 

[Yggdrasill]

Production ramp regarding battery cells or not:

1) The use of Samsung 2170 cells - which are manufactured as per Tesla's designs and instructions - in the powerpacks installed in Australia (put together in the GF), shows that this deal has been in the works for a long time. At least 6 months ago, but most likely more. So it's not like Tesla panicked and quickly ordered some cells from Samsung. This partnership has existed for some time now.

2) The benefits of the above partnership are:
- multiple suppliers of the same cells (Panasonic, Samsung and maybe others) keeps suppliers competing/honest and gives you backup supply;
- a larger part of the global battery market gets fed to Tesla. Other car companies looking for battery suppliers to produce BEV's have less options, and those options will be more expensive -> competitive advantage for Tesla.
-...

3) The downsides of the above partnership are:
- purchasing cells from Samsung is more expensive than producing those cells in-house.

But as Tesla's battery demand is expanding more rapidly than their battery production, it is a good thing they get in business with more suppliers. It means faster growth for TSLA, period.

So I'm not worried too much about the whole "Is the GF having production problems?".

It is certainly a possibility that Tesla wants to include Samsung in the plans for the future.

The Gigafactory One is planned to produce 105 GWh of cells and 150 GWh of packs. That leaves 45 GWh worth of cells up for grabs. Most of this will likely come frem Panasonic's Japanese factories, but some could come from Samsung and other suppliers.

Getting them into the supply chain can also help with getting them as a partner on a Gigafactory in Asia. I assume Tesla would need Samsung to contribute significant capital, so they might need some persuasion. Having several large partners, Tesla also increases the odds of continuing to have the best technology.

 

[schonelucht]

If they're currently producing 2 GWh/quarter, that's insufficient for Q4 Model 3 production. We don't know the mix of NMC/NCA cells currently being produced, but it is quite possible the production ramp has been going badly, and NMC cell production is being sacrificed for NCA cell production, to avoid impacting the Model 3 production ramp.

Elon was specifically asked how the ramp up in the gigafactory on cell production was going on the Q1 conference call (end of april). Here is the answer :

Elon Reeve Musk - Tesla Motors, Inc.

Yeah, I think it's ramping very rapidly (47:24) but we're not really seeing anything that's standing in the way of that.

Jonathan McNeill - Tesla Motors, Inc.

Yeah, and I completely agree, and we're basically tracking slightly ahead of where we need to be on vehicles, but that's sort of as was planned.

Elon Reeve Musk - Tesla Motors, Inc.

Yeah.

Jonathan McNeill - Tesla Motors, Inc.

We don't want to be too far ahead or else we'd have a pretty massive inventory issue showing up. So we run it in batches, and we run at high rates and then pause and validate the throughput, but, yeah, it's where we expect it to be.

At just about every opportunity they were asked about potential Model 3 bottleneck. Cell battery production was never mentioned, always the unluckiest supplier etc.

It makes no sense that they were ahead for battery cell production for cars in April but at the same time today have to sacrifice energy cell production for car cell production given the enormous volumes they are already running today while continuing to reaffirm production targets.

Feels a bit like the twilight zone : I am on the biggest Tesla bull forum and people are trying to convince me there is a production problem at the gigafactory.

In that case it makes sense to cover the shortfall with Samsung cells, until the production ramp can catch up with cell demand. I would have liked to see the cell production ramp faster, so that this wouldn't be an issue, but if this is what's happening, Tesla seems to be making the right decisions.

You can't just call up a new supplier and drop the cells in there on a moments notice. This has been in the works for a longer time.

 

[Yggdrasill]

Elon was specifically asked how the ramp up in the gigafactory on cell production was going on the Q1 conference call (end of april). Here is the answer :



At just about every opportunity they were asked about potential Model 3 bottleneck. Cell battery production was never mentioned, always the unluckiest supplier etc.

It makes no sense that they were ahead for battery cell production for cars in April but at the same time today have to sacrifice energy cell production for car cell production given the enormous volumes they are already running today while continuing to reaffirm production targets.

Feels a bit like the twilight zone : I am on the biggest Tesla bull forum and people are trying to convince me there is a production problem at the gigafactory.



You can't just call up a new supplier and drop the cells in there on a moments notice. This has been in the works for a longer time.

The ramp is certainly going slower than I expected. Whether it is going slower than planned is an open question.

Edit: This is basically what I was expecting back in January:

I added estimated Model 3 revenue to my spreadsheet based on my assumptions for ASP. 42k USD for base Model S (based on Musk tweet), 48k USD for 70 kWh pack and 63k USD for 80 kWh pack. I've based the ASPs on my thoughts for option pricing, but the numbers are probably low for 2017: Attempting to price out options for Model 3 (Part 2!)

View attachment 209767

Gigafactory output spreadsheet.zip

 

[FinFan]

My theory is that the Samsung cells were ordered long time before Gigafactory went online. Puropose of the order was to use those incase Gigafactory ramp up didn't go to plan. So now they just have some extra cells to put somewhere, so why not Australia. Maybe that was the original reason why Elon was so eager to hammer the 100 day deal. He just wanted to get rid of those cells as fast as possible. Who knows how much Samsung cells they orderer originally.
Is this reasonable theory or just crazy talk?

 

[jeewee3000]

My theory is that the Samsung cells were ordered long time before Gigafactory went online. Puropose of the order was to use those incase Gigafactory ramp up didn't go to plan. So now they just have some extra cells to put somewhere, so why not Australia. Maybe that was the original reason why Elon was so eager to hammer the 100 day deal. He just wanted to get rid of those cells as fast as possible. Who knows how much Samsung cells they orderer originally.
Is this reasonable theory or just crazy talk?

This is rather crazy talk:

- the cells delivered by Samsung are of the new 2170 type. This type didn't exist before the GF production came online.
- Musk is always eager to hammer quick deals. He will never "take it slow". That's what makes him stand out from the competition. This does not connect to hidden agenda of battery cells lying around.

 

[22522]

Tesla Meets With California, Nevada on Autonomous Truck Testing

My question is around axle weight and if they will try to cycle through different lead trucks while platooning - Take turns pulling.

Or have subordinate driverless trucks that can only follow, but can carry more cargo and less battery because on the reduced aerodynamic load. The picture is a long train of trucks with enhanced braking capability and turning capability vs mechanically connected trucks. Would LIDAR make sense (I don't want to think so) from a public acceptance point of view?

There is a lot of weight stacked up behind that lead truck. I would use multiple radars, with some in the A-Pillars.

The redundancy wants to be triple.

 

[ValueAnalyst]

lols, I know, a few months ago you bet me it would be under 10% by the end of this year never to rise above 10% again. well, with "end-18," you're moving more in agreement to my position now (fwiw, sounded like you are moving closer to my view on our other bet as well from what you wrote the other day about your surprise at how slowly the rest of the automakers are moving to EVs ; ).

I hope you are right on all this, but, for over 5 years that I've been watching Tesla extremely closely, the short position has been there the whole time, never dipping to market norms, and the FUD has only gotten far more aggressive with time.

The point is short interest will decline below 10% soon, much sooner than you expect.

I don't remember our second bet. Doesn't sound like something I would say, but I may be misremembering.

There's just wayyy to many bet rumors, share/dividend donations swirling around. So I created this thread:

Bet ValueAnalyst at your own risk

Please submit your proposed bets in that thread. First come, first serve for each bet.

 

[ValueAnalyst]

This is about the 4th time in the past 5 years that Tesla has effectively selectively released material information in a private setting without a press release for the rest of us at the same time.

Tesla is very transparent to us all about long term goals and strategy. That is very helpful to us as investors. The level playing field I understood Techmaven to be bringing up was about releasing material information to everyone at the same time. I've only followed one other company this closely before, but, my sense is that Tesla slips up on this far more than the average company. Again, in the past 5+ years, I've seen this happen at least 4 times. Perhaps this is "an order of magnitude better" for those inside the meetings, the opposite of a level playing field. To be fair, Tesla is under far more of a microscope than most companies, so, maybe this is all part of Wall Street's unwritten rules and typically goes unreported. I realize this may have been the first time you experienced this, so, I can see why you might of thought tech was overstating things. Over several years, we've seen a strong pattern.

I don't know exactly what is wink wink, nod nod, re complying with the rules on the book, but, I have some concerns about whether this repeated practice might bite us all in the rear one day. I take Elon as he is re speaking style (actually like it) and talking about the stock price, but I don't like this practice and have written to IR about it in the past.

We clearly disagree 180 degrees on this. I understand you have only followed one other company this closely, but having interacted with dozens of management teams of both private and public companies, I can tell you from experience, that Tesla is significantly better than most other companies with respect to sharing material long-term information with its all of its investors.

 

[ItsNotAboutTheMoney]

Elon reiterated this at the roadshow for new Tesla Bonds in Manhattan Monday.

Many thought he backtracked with the announcement Model Y would be based on the Model 3. That is incorrect.

Citation needed.

 

[RobStark]

Citation needed.

See originally post linking Automotive News.

Or simply Google