2017 Investor Roundtable:General Discussion

[everman]

Not at all. I want Tesla to learn from their past efforts how long it actually takes to implement hardware changes on the production line and forecast deliveries accordingly. That simple.

Ok, I understand. I just took your original point as arguing against the introduction of a HUD, as it will lead to delays. I'm sure they are working hard to minimize them each time, but I hope they don't delay features due to some delivery forecast months prior.

 

[techmaven]

Question for any resident accounting gurus:

Now that Solarcity is part of Tesla, if stationary storage projects are bought on a PPA basis like in Kauai, how does that affect Tesla's bottom line? If it is financed with non-recourse debt, what is the effect on GAAP profit and cash from operations metrics?

Thx!

 

[tander]

Ok, here is another interpretation of this misalignment in announcements. It is a little "out there", but bear with me.

While Elon has publicly set the target date for M3 production at July 1st, 2017, I always translated this to myself as September(ish), with deliveries starting end of October (long validation for the first few batches, working out quality and production issues... we know the deal, Model X ramp process, except, hopefully, with much less trouble).

Yesterday Tesla said volume cell production started at GF1 for TE and will start for M3 in Q2. Now, once again, I translated that to Elon-time, meaning, second half of June, with enough stockpile of cells for that September production start of M3. But that too was a little odd to me, I mean how much in advance do you need to start spitting out cells. 2-3 months, really?

So what if Panasonic's comment was a little bit of a slip-up? At this point TE cell production is probably a fraction of what M3 cell production will be, so what if by April, Pana was referring to M3 cells?

The implication of that would be huge. If they actually started volumes for cars in April, that could mean they anticipate vehicle production much sooner than September. Dare I say, around the actual target date (for a change)? This would also coincide with recent reports by Fred, that suppliers say Tesla is Hell-bent of getting all ducks in a row by July 1st.

So if they started "pilot" M3 production in July and actually ramping in August, that would give them almost 2 quarters of production. Meaning, we could be looking at the 100k+ M3 numbers this year that Elon was predicting. Hmm...

One thing to consider regarding timelines; a big part of why the x was late or delayed was because it was deliberate. A big part of doing the x was just to bridge the gap and keep peoples' interest between the s and m3, and they knew they needed time to get battery costs lower and the volume up. So in part the x was deliberately designed to be hard to design and make, while the model 3 is the opposite. So applying production timelines of s or x to the 3 is not really apples to oranges. The x is like a custom wedding cake, the 3 is designed to be twinkies pumped out at volume. I think you're right about the timeline, there could be hiccups and speed bumps, but if the supply chain is ready there is no reason not to pump 3 out as soon and fast as possible.

 

[Irishjugg]

it's quite a long list... but I have a personal opinion based on my software engineering experience that companies are grossly over simplifying and have their time lines way too aggressive for when we might start seeing anything like Tesla Mobility.

Fair enough. I assume that's focused on the self driving ai part rather than the Uber type problems?

 

[Yggdrasill]

Do you have an estimate on how many 2170 are in each pod in Powerpack 2? We can calculate Wh per 2170 TE cell this way as well.

Well, if we assume the Powerpack/Powerwall is using the heat pipe-based pack design, a powerpack 2 could contain around 16,000 21-70s.

Each pod has around 700x700 mm of batteries, or 490,000 mm^2, which is effectively around 443,500 mm^2 when packed with circles. Each 21-70-cell will effectively occupy around 363 mm^2 with 0.5 mm between each cell, 380 with 1 mm between each cell and 415 mm^2 with 2 mm between each cell.

That works out to 1221, 1167 and 1068 cells. This doesn't include the center cooling pipe, so lets assume around 1000 cells. With 16 pods in a Powerpack, that's 16,000 cells, and the energy per cell is around 13 Wh/cell. And the machine would output up to 0.81 GWh/year with no downtime. 13 Wh/cell is roughly a 20% improvement in chemistry.

(The calculations line up pretty well with my previous calculations: Short-Term TSLA Price Movements - 2016 Previously I assumed immersion cooling, but heat pipe cooling is fairly equivalent.)

 

[tander]

Question for any resident accounting gurus:

Now that Solarcity is part of Tesla, if stationary storage projects are bought on a PPA basis like in Kauai, how does that affect Tesla's bottom line? If it is financed with non-recourse debt, what is the effect on GAAP profit and cash from operations metrics?

Thx!

Not a guru, but have been looking into that. The big question with gaap right now regards a change in their policy on revenue recognition. They voted to change the rule and now companies can elect to start applying this new rule in 2017 or they are obliged to do it by 2018. Tesla has not confirmed whether they plan to start doing this in 2017 or not but if they do it could have material and packs on the "bottom line", my understanding is it applies retroactively too, so its possible that when they make the change you could make many previous quarters or years look much better than before, since revenue recognition has previously been a big point of contention and confusion particularly with solar city .
I don't know if they have to publicly disclose when they plan to start using this new method, I've emailed them and received no replly, the last 10q said they hadn't decided yet. However it makes sense to me that if you're in the middle of a merger and you're already combining the books of two companies and starting a new year, seems like it makes sense to start using these new rules now as well? Tesla and other Silicon Valley companies tend to adopt a new accounting practices faster than other companies to I don't know I guess we'll see. Hope that helps.

 

[MitchJi]

One thing to consider regarding timelines; a big part of why the x was late or delayed was because it was deliberate. A big part of doing the x was just to bridge the gap and keep peoples' interest between the s and m3, and they knew they needed time to get battery costs lower and the volume up. So in part the x was deliberately designed to be hard to design and make, while the model 3 is the opposite.

That's foolishness!

They wanted to delay the income from the MX production?

Elon was planning on going through production hell and sleeping on the production line?

How did those things help them transition to the M3?

 

[bonaire]

They can package and resell any of their PPA contracts via bond-market type of financing. So, they build the equipment and install it and then can claim that the equipment will make a certain amount of money and then sell that "future money" in a type of bond. I bet it's complex to read the contracts. For instance, they can re-sell PPAs on a homeowner's solar PV array based on nameplate PV size and not future production. They just do not indicate if the roof is facing the sun properly, if there is shading or not or the expected PV degradation over time. So, the sale guys at home depot who are selling "would you be interested in a free solar upgrade?" to consumers probably have incentive to just put solar on any roof these days. The PPA agreements will be sold off. I have a neighbor who recently had SCTY install - E and W facing roofs - plenty of huge trees blocking production. Sort of a heartbreaking install knowing that their array will produce so little.

 

[Turing]

New research note this morning from Goldman Sachs. Likes the sizzle, wants more steak.

So Goldman was disappointed Elon didn't disclose concrete 'material information' at a private investor event? Where's the fainting couch when you need it?

 

[EnzoXYZ]

One thing to consider regarding timelines; a big part of why the x was late or delayed was because it was deliberate. A big part of doing the x was just to bridge the gap and keep peoples' interest between the s and m3, and they knew they needed time to get battery costs lower and the volume up. So in part the x was deliberately designed to be hard to design and make, while the model 3 is the opposite. So applying production timelines of s or x to the 3 is not really apples to oranges. The x is like a custom wedding cake, the 3 is designed to be twinkies pumped out at volume. I think you're right about the timeline, there could be hiccups and speed bumps, but if the supply chain is ready there is no reason not to pump 3 out as soon and fast as possible.

My read is all the analysts and bears isn't doubting M3 is a simplification of S/X but that actual mass production is not a cake wake. I don't think there is anyone that doesn't believe the 3 will be easier car to make. There are plenty of evidence that hiccups and slowdowns already plagued the current production line. Continual slowdown and execution problems burns cash and cash drained will be start of the end for Tesla... so the bears say.

I agree we shouldn't compare S/X to M3 in the manufacturing process and also we should not assume M3 will have highly technical design difficulties. If so we also should not think Tesla will perfect M3 manufacturing process in the first crack.

 

[mmd]

Yesterday's headlines were amazingly incorrect. Tesla doesn't make any cells, Panasonic does. Both technology and equipment are owned by Panasonic. Yet I didn't see any headline saying "Panasonic starts cell production at Gigafactory". The qualification and packaging of cells is identical, no matter the cells come from across a line in the factory or across the Pacific.
From the Gigafactory opening party in July, Tesla employees are not even allowed in the Panasonic area where cells are produced. I wonder, has Tesla started paying them now for advertizing that they are putting out favorable headlines?
Tesla starts battery cell production at gigafactory
Tesla begins churning out battery cells at Nevada Gigafactory
Tesla starts cell production at Gigafactory, shows it off to investors
Tesla starts battery cell production at gigafactory
Tesla starts battery cell production at gigafactory


BTW, "we sell every car we make" is a tautology. Everyone does. If it doesn't sell, lower prices till it sells and then lower production till you have nothing left to sell. That's exactly what Tesla is doing. Every quarter, quite a bit of downtime, except may be Q3 of last year.
Deutsche Questions Whether Tesla Has Overcome Production Challenges

 

[larmor]

So Goldman was disappointed Elon didn't disclose concrete 'material information' at a private investor event? Where's the fainting couch when you need it?

In many CCs and presumably in this investors meeting, people have tried to get an exact cost per kwh for the batter, to which EM has responded (in the past) that comes under the company business umbrella, and not public knowledge to share the entire business plan.

 

[jhm]

Even the Wikipedia entry for that book is an interesting read and applicable to Tesla. Thanks.

It's curious that Clayton Christensen has maintained that his theory does not apply to Tesla or electric vehicles generally. He views electrification as a sustaining technology for the automotive industry, rather than a disruptive. This is largely because his theory focuses on cheap technologies finding niches that then advance to take on more established technologies. He thinks if cheap golf carts take over cars that would be disruption, but not if a high end electric progressively becomes more affordable.

I do believe he is missing the point of what exactly is being disrupted. It is not cars per se, but the internal combustion engine and gasoline that is getting disrupted by little battery cells. Who would have thought that laptop batteries would come to imperil ICE and fossil fuels? Clearly, batteries are not sustaining technologies for ICE makers and the oil industry.

It is surprising that Christensen makes this mistake. Much of his book traces the evolution of hard drives. Hard drives go progressively smaller. This did not make mainframe computers go away. It only meant that with time even mainframes would incorporate smaller disk drives, even as the suppliers of larger disk drives went out of business. The analogy here is that a batter drive train is more like smaller disk drive than a mainframe computer. The outer shell of cars or computers may not be disrupted even as critical internal components are disrupted and fall into the dust bin of history. So I suspect that Christensen himself has particular blinders regarding EVs.

This I believe reflects poorly on his theory, in that it is not crisp enough to identify disruption in the face of other intellectual biases that a researcher may have. Indeed the point of such a theory should be to help practitioners get beyond their organizational blinders and see the disruption potential before it happens. It should not be used to discount a potential disruptor because it fails to meet a certain set of expectations of a given theory. Clayton made the same mistake with iPhones. He said it was a high end sustain technology for cell phone. Only in hindsight did Christensen recognize that the disruption was to PCs than to cell phones. People were willing to pay more for smart phone, not because it was a replacement for cell phones, but because it was a replacement for other computing devices that added enormous mobility. And yet PCs were much more powerful computers at the time. So PCs were clearly disrupted, not so much phones per se. Camera and film were also disrupted. So it a poor theory of disruption that can only recognize disruption after the fact. Christensen should be asking himself, if EVs are a potential disruptor, what exactly would it be disrupting? Put this way one can easily see many technologies at risk, and the incumbents which profit from them are also at risk.

Now the Gigafactory is mass producing cells. Every cell is displacing older technologies. The list of technologies and companies within range of this dreadnought is quite long, but this alien ship will keep firing with increasing speed.

 

[myusername]

Fair enough. I assume that's focused on the self driving ai part rather than the Uber type problems?

yes.

 

[Turing]

My read is all the analysts and bears isn't doubting M3 is a simplification of S/X but that actual mass production is not a cake wake. I don't think there is anyone that doesn't believe the 3 will be easier car to make. There are plenty of evidence that hiccups and slowdowns already plagued the current production line. Continual slowdown and execution problems burns cash and cash drained will be start of the end for Tesla... so the bears say.

I agree we shouldn't compare S/X to M3 in the manufacturing process and also we should not assume M3 will have highly technical design difficulties. If so we also should not think Tesla will perfect M3 manufacturing process in the first crack.

Has the model S really had noticeable issues with production in the last two years? S was basically on time in terms of guidance, and production has been pretty smooth for a while now, no?

 

[The Blue Owl]

I agree with the rest of your post, but I don't think this is true:

Clayton made the same mistake with iPhones. He said it was a high end sustain technology for cell phone. Only in hindsight did Christensen recognize that the disruption was to PCs than to cell phones. People were willing to pay more for smart phone, not because it was a replacement for cell phones, but because it was a replacement for other computing devices that added enormous mobility. And yet PCs were much more powerful computers at the time. So PCs were clearly disrupted, not so much phones per se.

If by "disrupted" we mean "replaced" (which is the only interpretation of any consequence), then the existing cell phones were very much disrupted out of existence. PCs still sell today, albeit in smaller numbers. Dumb cell phones, or even smarter ones like the BlackBerry and the Nokia Communicator disappeared completely. Huge, dominant companies ceased to matter almost overnight (RIM, Nokia, Motorola).

PCs also felt the impact, but it was comparatively much smaller, and they are still relevant today (i.e., selling in significant numbers).

 

[DaveT]

Not sure if this has been posted or not. Q&A from yesterday's Gigafactory investor event.

EDIT: it appears that this video is from last summer's GF event. Apologies.

 

[Mike Smith]

So Goldman was disappointed Elon didn't disclose concrete 'material information' at a private investor event? Where's the fainting couch when you need it?

Bloomberg is describing the lack of privileged inside information at the event as a 'caution flag on the Tesla Plant'.

Wall Street Waves Caution Flag on Tesla Plant as Investors Snap Up Shares

 

[techmaven]

Yesterday's headlines were amazingly incorrect. Tesla doesn't make any cells, Panasonic does. Both technology and equipment are owned by Panasonic. Yet I didn't see any headline saying "Panasonic started producing cells at Gigafactory". The qualification and packaging of cells is identical, no matter the cells come from across a line in the factory or across the Pacific.
From the Gigafactory opening party in July, Tesla employees are not even allowed in the Panasonic area where cells are produced. I wonder, has Tesla started paying them now for advertizing that they are putting out favorable headlines?
Tesla starts battery cell production at gigafactory

Quibbling over semantics is all you have? The Gigafactory equipment is lease-to-own. And overwhelmingly, the headlines for the start of cell production in Smyrna, TN was Nissan, not AESC or NEC. The plant itself is owned by Tesla. And there are certainly big cell production parts that are not closed off from Tesla. Did you visit this summer? The parts that are restricted areas do not consist of the entire Panasonic section. I believe anode and final cell assembly is restricted. It was made very clear that Tesla had a hand in overall plant design, especially in energy management. I don't think you've been listening to Panasonic people talk about the Gigafactory.

BTW, "we sell every car we make" is a tautology. Everyone does. If it doesn't sell, lower prices till it sells and then lower production till you have nothing left to sell. That's exactly what Tesla is doing. Every quarter, quite a bit of downtime, except may be Q3 of last year.
Deutsche Questions Whether Tesla Has Overcome Production Challenges

Except that production is increasing, not lowering.

 

[Fallenone]

Not sure if this has been posted or not. Q&A from yesterday's Gigafactory investor event.

This is last summer's one. The press only before the opening party.