2017 Investor Roundtable:General Discussion

[Bet TSLA]

What is everyone's cost estimate of building a 50 GWh Gigafactory?

I don't have an estimate, and I'm going to reason by analogy, so you may want to avert your eyes.

Tesla has been quite explicit that they would be treating the Gigafactory as a product. The first one is therefore quite obviously a prototype, and costs ridiculous amounts of money as they figure out what they're doing. It is, in fact, more expensive because they are treating it as a product and therefore "one off" solutions are unacceptable.

So clearly Gigafactories two through twenty will be much, much cheaper. It's anybody's guess how cheap. The minimum would be cost of materials. Perhaps they could 3D print most of them?

 

[brian45011]

GF update: New drone flyby shows the new section complete;
Tesla Gigafactory: take a close look with a new drone video
"<Prior to this small addition, which is the only new structure so far this year, the overall structure had a 1.9 million square-foot footprint. Including several levels, the factory currently is about 4.9 million square feet of operational space. This represents only ~30 percent of the total finished Gigafactory.>"

If only 30% of the exterior footprint has been completed, does that imply the need for additional capital soon since "We had cumulatively incurred and capitalized costs of $1.98 billion and $825.3 million, respectively, for Gigafactory 1 as of June 30, 2017 and December 31, 2016."?

"As of July 31, 2017, there were 166,887,023 shares of the registrant’s common stock outstanding...As of April 30, 2017, there were 164,259,736 shares of the registrant’s common stock outstanding" That's about a 1.6% increase in 3 months without any public offering.

 

[ValueAnalyst]

I don't have an estimate, and I'm going to reason by analogy, so you may want to avert your eyes.

Tesla has been quite explicit that they would be treating the Gigafactory as a product. The first one is therefore quite obviously a prototype, and costs ridiculous amounts of money as they figure out what they're doing. It is, in fact, more expensive because they are treating it as a product and therefore "one off" solutions are unacceptable.

So clearly Gigafactories two through twenty will be much, much cheaper. It's anybody's guess how cheap. The minimum would be cost of materials. Perhaps they could 3D print most of them?

I agree with your post. I think Elon & Team may prove me too conservative in my ~$40m-45m per GWh cost estimate for subsequent Gigafactories.

I'll allow reasoning by analogy ( for this purpose only) since none of us may have the necessary detail to reason from first principles.

 

[ValueAnalyst]

They could tier the referral awards. 1k for S&X and 500 for Model 3 (or less or none).

Just remembered Model 3 is not eligible for the referral program anyway.

 

[Alketi]

By the end of 2017, Model 3 reservations will exceed one million.

ONE MILLION!

VA, come back to us!

Tesla won't promote the Model 3 at all in 2017, as there would be no point.

If one assumes the anti-sell/production-hell post-event (impossible) rate of 1,800 net orders were to continue for the rest of the year (despite Tesla's silence), one would only have:

455K + (1,800*152) = 728K orders

550-600K would be a more conservative bet, IMO.

They're not going to build more Gigafactories until they can at least prove defect-free manufacturing on their first product line. They will announce more Gigafactories, but not start building.

The Model 3 is, in many ways, superior to a Model S. I'd argue that size and interior are two ways. I don't want something as large as a Model S and I'm in love with the minimalist interior. I'd spend $70K on a fully loaded Model 3 before I ever spent $75K on a base Model S. But, for those looking to buy a $35K car, you can't deliberately price them into an S. They'll look elsewhere.

Here's where I agree with you:

1. Gigafactory cost is vastly overstated. The land/taxes will be less than expected. And, a Gigafactory can become functional with a minimal amount of square footage and can be scaled section-by-section over time. I expect they'll break ground early next year on one or more sites.

2. Tesla will have a world-wide monopoly for a desirable $35K EV until ~2020 with (performance / looks / features / availability / charging).

3. The Model 3, once large enough numbers are in the field, will be sold word-of-mouth and be back-ordered for years. It has no equal.

So, yes, more Gigafactories need to be ramped. But not until next year. One step at a time.

Here's what I'm listening to this fine Saturday morning:

 

[ValueAnalyst]

ONE MILLION!

VA, come back to us!

Tesla won't promote the Model 3 at all in 2017, as there would be no point.

If one assumes the anti-sell/production-hell post-event (impossible) rate of 1,800 net orders were to continue for the rest of the year (despite Tesla's silence), one would only have:

455K + (1,800*152) = 728K orders

550-600K would be a more conservative bet, IMO.

They're not going to build more Gigafactories until they can at least prove defect-free manufacturing on their first product line. They will announce more Gigafactories, but not start building.

The Model 3 is, in many ways, superior to a Model S. I'd argue that size and interior are two ways. I don't want something as large as a Model S and I'm in love with the minimalist interior. I'd spend $70K on a fully loaded Model 3 before I ever spent $75K on a base Model S. But, for those looking to buy a $35K car, you can't deliberately price them into an S. They'll look elsewhere.

Here's where I agree with you:

1. Gigafactory cost is vastly overstated. The land/taxes will be less than expected. And, a Gigafactory can become functional with a minimal amount of square footage and can be scaled section-by-section over time. I expect they'll break ground early next year on one or more sites.

2. Tesla will have a world-wide monopoly for a desirable $35K EV until ~2020 with (performance / looks / features / availability / charging).

3. The Model 3, once large enough numbers are in the field, will be sold word-of-mouth and be back-ordered for years. It has no equal.

So, yes, more Gigafactories need to be ramped. But not until next year. One step at a time.

Here's what I'm listening to this fine Saturday morning:

Solid music choice.

I recognize the argument that the new reservation rate of 1,800 per day may not continue, but I'm not sure if the argument will prove correct. Press around Model 3 is unlikely to slowdown in the coming months and thousands of Model 3's soon hitting the streets may even further accelerate this rate. It's hard to predict this rate, which I think is why Tesla is struggling to put a cap on unmet demand growth even with a higher ASP. The following is a fact: the demand for Model 3 has surprised the bears, the bulls, and Tesla management, so it would not be advisable to continue to underestimate this growth rate.

Regarding your other point, I think we'll be surprised how quickly the additional Gigafactories get under way once locations are announced. Tesla is highly incentivized to exponentially ramp up its production, and whatever you incentivize will happen.

 

[brian45011]

...



Why do you say that it is happening now? Tesla realized this efficiency gain already way back in 16Q3 when automotive revenue nearly doubled but SG&A increased by only 5%. Your comparison with 16Q2 only reflects that quarter's gains. Since then SG&A (when accounting for SCTY like you do) has been relatively stable versus revenue.

I am not expecting another step up (or down, really) for the SG&A over revenue metric before the M3 deliveries scale up, ie 18Q1. But then it will once again be as drastic as 16Q3 (hopefully).

From the 2Q16 CC about SG&A:
"This quarter, SG&A would have been flat quarter-over-quarter in spite of our continued expansion in service and sales, were it not for the payroll taxes we paid on our CEO's options exercises. SG&A was up $19 million quarter-over-quarter on a non-GAAP basis. $17 million of that was the payroll expense associated with those option exercises."

 

[Fallenone]

...



Why do you say that it is happening now? Tesla realized this efficiency gain already way back in 16Q3 when automotive revenue nearly doubled but SG&A increased by only 5%. Your comparison with 16Q2 only reflects that quarter's gains. Since then SG&A (when accounting for SCTY like you do) has been relatively stable versus revenue.

I am not expecting another step up (or down, really) for the SG&A over revenue metric before the M3 deliveries scale up, ie 18Q1. But then it will once again be as drastic as 16Q3 (hopefully).

As I wrote earlier in response to another post that also mentioned 16Q3, I do think that quarter was an outlier because Tesla had easy and near 0 cost of converting Model 3 reservation holders to the Model S. It's a tactic that only worked once and cannot be repeated to that extent. As a results, we saw the ratio of SGAA/Atuo Rev rose back quite a bit in the following 16Q4.

I went back and examined all ERs since 2014 Q1. Since 2016 Q4 part of Tesla's earnings reflects SCTY. I'm using the following basis:

1. Before acquisition (2016 Q1-Q3), SCTY was running about $250M/Q on OpEx, with SGAA being the overwhelming part of it (RND was below $15M).
2. In 2016 Q4, SCTY reported $135M OpEx with $12M of RND and TSLA reported $85 OpEx related to SCTY. So I took $80M out of TSLA SGAA in 2016 Q4 to get auto related SGAA.
3. I assumed SCTY costing around $150M of SGAA for 2017 Q1 and Q2 due to letting people go.
4. 2017 Q1 had another $65M one time cost related to acquisition that was taken off to calculate this ratio.
5. I assume SGAA related to storage is negligible. This assumption would overestimate this ratio.

End results look like this

2014 Q1: 20%
2014 Q2: 18%
2014 Q3: 19%
2014 Q4: 22%
2015 Q1: 22%
2015 Q2: 23%
2015 Q3: 28%
2015 Q4: 26%
2016 Q1: 31%
2016 Q2: 27%
2016 Q3: 16%
2016 Q4: 19%
2017 Q1: 17%
2017 Q2: 17%

Yesterday I didn't notice take the SCTY related SGAA cost in 2016 Q4 so got a much higher ratio that quarter. I also forgot the $65 acquisition cost in 2017 Q1. So what I thought was 16%, 23%, 20%, 17% for Q3, Q4, Q1, Q2, and thought we have a trend here. Now the improvement since Q4 was only modest and not really worth to be excited.

 

[Fallenone]

"As of July 31, 2017, there were 166,887,023 shares of the registrant’s common stock outstanding...As of April 30, 2017, there were 164,259,736 shares of the registrant’s common stock outstanding" That's about a 1.6% increase in 3 months without any public offering.

Could this be achieving milestones for Model 3 for employees? Each individual doesn't get that much to be mentioned like the CEO and management incentive grants, but pooled together could be significant if they have this kinda thing.

 

[ggr]

Since you specifically asked for comments, I will give you some.

Tesla is making an important strategic mistake!

As I laid out in my article in August 2016 titled, Tesla Profitability: A Game Theory Perspective, (which forum rules preclude my linking), I presented the next logical steps Tesla would take with its Model 3 program:

Bottom Line: With demand already at impossible levels, game theory would suggest that the next logical moves are (1) reach lower on the Model S demand curve to tap into Model 3 reservations, (2) price Model 3 options high in order to capture more economic profits during the initial ramp-up through 2018 and (3) focus solely on bringing Gigafactory to full capacity immediately before credible competition arrives, which will not happen until 2019 at the earliest.

Well, as a professional mathematician, I don't believe Game Theory says any such thing.

Tesla followed these steps to a tee. It even set Model 3 ASP higher than most estimated. This level, however, turned out to be still very low compared to the level at which demand for Model 3 would closely align with how quickly Tesla can possibly increase production in the next twelve months. By the end of 2017, Model 3 reservations will exceed one million.

Now that the prices of options are set, however, Tesla can't increase the Model 3 ASP. The steps it can take, however, are:

• Move planned capex from Supercharger buildout to accelerating additional Gigafactory buildout. Tesla should be building five or more Gigafactories simultaneously, not three or four, and it should break ground yesterday.

You don't know for sure that they haven't already broken ground on at least one other Gigafactory, or are not at least well into the planning phase. It makes sense for them to hoard capital at the moment to ramp up M3 production, but there's plenty they can do to prepare Gigafactory production without (yet) spending lots of money. In the meantime, supercharger buildout is already very important, and with M3s hitting the roads, and an existing promise to build them, and the relatively small outlay compared to a gigafactory, it doesn't make any sense at all to slow down superchargers.

• Lower referral awards and limit number of referral awards one owner can at most receive to three. Now that the brand value of Tesla has a life of its own, the purpose of this program should move slightly from increasing demand to motivating more people to talk up their Tesla's, not motivating Tesla enthusiasts to talk more. There is a slight but important difference in the benefit-to-cost ratio to Tesla.
• Further increase efforts to get Model 3 reservation holders to buy Model S. This is very important in 2018.
  
• Lower warranty coverage further for base-model Model 3 (i.e. decrease warranty cost/accrual as a percentage of revenue) until subsequent Gigafactories start coming online in 2019 and plow savings to Gigafactory buildout. This would also increase Tesla's borrowing capacity as it would improve its income statement and balance sheet metrics.
• Borrow non-dilutive debt to accelerate Gigafactory buildout. Tesla's cost of incremental debt = 7-8% but its Return on Capital = 100%+. By managing its balance sheet so conservatively, Tesla is foregoing substantial economic profits.

I agree with some of these. Mostly because I hardly ever see anyone, working from home, to refer. Don't agree about warranty coverage, though. It's a cheap value add for marketing.

 

[ValueAnalyst]

Since you specifically asked for comments, I will give you some.


Well, as a professional mathematician, I don't believe Game Theory says any such thing.

You don't know for sure that they haven't already broken ground on at least one other Gigafactory, or are not at least well into the planning phase. It makes sense for them to hoard capital at the moment to ramp up M3 production, but there's plenty they can do to prepare Gigafactory production without (yet) spending lots of money. In the meantime, supercharger buildout is already very important, and with M3s hitting the roads, and an existing promise to build them, and the relatively small outlay compared to a gigafactory, it doesn't make any sense at all to slow down superchargers.

I agree with some of these. Mostly because I hardly ever see anyone, working from home, to refer. Don't agree about warranty coverage, though. It's a cheap value add for marketing.

Thank you for your feedback; I appreciate it very much.

Based on your and others' comments, combined with my estimate that the planned Supercharger and destination charger buildout will cost less than $1 billion, I'll drop my idea of slowing down Supercharger buildout.

 

[TradingInvest]

Tesla is making an important strategic mistake!

As I laid out in my article in August 2016 titled, Tesla Profitability: A Game Theory Perspective, (which unfortunately moderators do not allow me to link here, but you can find it with a Google search), I presented the next logical steps Tesla would take with its Model 3 program:

Bottom Line: With demand already at impossible levels, game theory would suggest that the next logical moves are (1) reach lower on the Model S demand curve to tap into Model 3 reservations, (2) price Model 3 options high in order to capture more economic profits during the initial ramp-up through 2018 and (3) focus solely on bringing Gigafactory to full capacity immediately before credible competition arrives, which will not happen until 2019 at the earliest.

Tesla followed these steps to a tee. It even set Model 3 ASP higher than most estimated. This level, however, turned out to be still very low compared to the level at which demand for Model 3 would closely align with how quickly Tesla can possibly increase production in the next twelve months. By the end of 2017, Model 3 reservations will exceed one million.

Now that the prices of options are set, however, Tesla can't increase the Model 3 ASP. The steps it can take, however, are:

• Move planned capex from Supercharger buildout to accelerating additional Gigafactory buildout. Tesla should be building five or more Gigafactories simultaneously, not three or four, and it should break ground yesterday.
• Lower referral awards and limit number of referral awards one owner can at most receive to three. Now that the brand value of Tesla has a life of its own, the purpose of this program should move slightly from increasing demand to motivating more people to talk up their Tesla's, not motivating Tesla enthusiasts to talk more. There is a slight but important difference in the benefit-to-cost ratio to Tesla.
• Further increase efforts to get Model 3 reservation holders to buy Model S. This is very important in 2018.
  
• Lower warranty coverage further for base-model Model 3 (i.e. decrease warranty cost/accrual as a percentage of revenue) until subsequent Gigafactories start coming online in 2019 and plow savings to Gigafactory buildout. This would also increase Tesla's borrowing capacity as it would improve its income statement and balance sheet metrics.
• Borrow non-dilutive debt to accelerate Gigafactory buildout. Tesla's cost of incremental debt = 7-8% but its Return on Capital = 100%+. By managing its balance sheet so conservatively, Tesla is foregoing substantial economic profits.

If you're thinking these steps would anger potential customers, you know what would anger them more? Waiting for what you want for three years.

ValueAnalyst,

I usually agree with your views concerning Tesla's long term future, but I have different views regarding the bullet list above.

• Move planned capex from Supercharger buildout to accelerating additional Gigafactory buildout. Tesla should be building five or more Gigafactories simultaneously, not three or four, and it should break ground yesterday.

Elon sincerely cares about Tesla's customers/supporters. It is impossible to think that he will pause the SuperCharger expansion in favor of more Gigafactories. From a business point of view, we will urgently need more SuperChargers in certain areas considering more than 400k Tesla cars will be delivered in 2018.

• Lower referral awards and limit number of referral awards one owner can at most receive to three. Now that the brand value of Tesla has a life of its own, the purpose of this program should move slightly from increasing demand to motivating more people to talk up their Tesla's, not motivating Tesla enthusiasts to talk more. There is a slight but important difference in the benefit-to-cost ratio to Tesla.

The referral program is mainly to help selling cars, it's also a way to give back to the supporters. In Elon's view, these supporters (and long term investors) are part of the Tesla family. Elon will not try to squeeze a few extra dollars from them.

• Lower warranty coverage further for base-model Model 3 (i.e. decrease warranty cost/accrual as a percentage of revenue) until subsequent Gigafactories start coming online in 2019 and plow savings to Gigafactory buildout. This would also increase Tesla's borrowing capacity as it would improve its income statement and balance sheet metrics.

Tesla wants to provide the best product package, which includes warranty and service. Soon, all other brands will have their EVs with decent warranty terms (they can afford to lose money on small number of EVs). If Tesla reduces coverage, we would see dozens of bashers saying Tesla has the worst warranty.

• Borrow non-dilutive debt to accelerate Gigafactory buildout. Tesla's cost of incremental debt = 7-8% but its Return on Capital = 100%+. By managing its balance sheet so conservatively, Tesla is foregoing substantial economic profits.

Currently Tesla is moving as fast as humanly possible, anything faster will likely lead to trouble. The best analogy I can think of is our stock accounts. We know TSLA will go much higher, do you think it's wise to borrow full margin and invest in TSLA? We can argue that if we don't use full margin, we would forgo substantial economic profits.

Tesla's ultimate goal is to accelerate the world’s transition to sustainable energy. Keeping this in mind can help us understand Elon's decisions better.

 

[TradingInvest]

At this time, I don't think it's a good idea to persuade shorts to cover and get out of trouble. If they escape, they will come back to hurt Tesla again in the future (by shorting and spreading lies). In the past, when shorts asked me why I was so bullish about Tesla's long term future, I didn't give them detailed explanations. From Elon's point view, do you think he prefers to see shorts escape without a scratch? I do argue with shorts occasionally, but I will never write convincing details to help them.

 

[brian45011]

Could this be achieving milestones for Model 3 for employees? Each individual doesn't get that much to be mentioned like the CEO and management incentive grants, but pooled together could be significant if they have this kinda thing.

About half was: Tesla - Current Report
Add up the form 4s for most of the rest.

Those who think Tesla can issue non-dilutive debt at will to raise new capital should check the restrictions in the Asset Based Credit Agreement (and other existing borrowing agreements) about Additional Indebtedness and other financial covenants.

 

[Lhan]

Poor Prius
A Funny Story — Used Tesla Model S vs Toyota Prius Prime

 

[ValueAnalyst]

ValueAnalyst,

I usually agree with your views concerning Tesla's long term future, but I have different views regarding the bullet list above.

I'm okay with disagreement, because if no one is disagreeing with me, I'm not doing enough to push the limits of our TMC bubble. It may be helpful to keep in mind that I feel more free to express some of my more aggressive/optimistic views here, and then scale back based on feedback, rather than include these aggressive/optimistic views in my free-to-public articles that are read by thousands of people who may not be knowledgeable enough in finance/economics/accounting/manufacturing etc. as some TMC members. Rest assured that I'm mindful of that.

• Move planned capex from Supercharger buildout to accelerating additional Gigafactory buildout. Tesla should be building five or more Gigafactories simultaneously, not three or four, and it should break ground yesterday.

Elon sincerely cares about Tesla's customers/supporters. It is impossible to think that he will pause the SuperCharger expansion in favor of more Gigafactories. From a business point of view, we will urgently need more SuperChargers in certain areas considering more than 400k Tesla cars will be delivered in 2018.

Based on your feedback and others' comments, I dropped this one from my article.

• Lower referral awards and limit number of referral awards one owner can at most receive to three. Now that the brand value of Tesla has a life of its own, the purpose of this program should move slightly from increasing demand to motivating more people to talk up their Tesla's, not motivating Tesla enthusiasts to talk more. There is a slight but important difference in the benefit-to-cost ratio to Tesla.

The referral program is mainly to help selling cars, it's also a way to give back to the supporters. In Elon's view, these supporters (and long term investors) are part of the Tesla family. Elon will not try to squeeze a few extra dollars from them.

• Lower warranty coverage further for base-model Model 3 (i.e. decrease warranty cost/accrual as a percentage of revenue) until subsequent Gigafactories start coming online in 2019 and plow savings to Gigafactory buildout. This would also increase Tesla's borrowing capacity as it would improve its income statement and balance sheet metrics.

Tesla wants to provide the best product package, which includes warranty and service. Soon, all other brands will have their EVs with decent warranty terms (they can afford to lose money on small number of EVs). If Tesla reduces coverage, we would see dozens of bashers saying Tesla has the worst warranty.

I'll keep the warranty point but word it differently and state the risks around this.

Tesla's ultimate goal is to accelerate the world’s transition to sustainable energy. Keeping this in mind can help us understand Elon's decisions better.

Note that an accelerated gigafactory buildout would in fact accelerate the world's ultimate transition to sustainable energy, not slow it down.

 

[MitchJi]

I think a successful Model 3 ramp could change everything. It should give even more confidence to investors and lenders that Tesla is capable of high volume production, which will make it easier to raise cash if necessary to accelerate the completion of GF1, and buildout of GFs 3-6. I personally don't think we are there quite yet though. But soon, with a little luck.

I believe that there's a good chance that they will decide (or have already decided) to use FWD's on the MY. If that's correct that will test the market's confidence in Tesla's manufacturing competence.

And note that analysts confidence is still very low.

 

[22522]

I'm okay with disagreement, .

A few things. I don't like the game theory suggests.

If you look at the cars that sell over $50K, there are not enough people there.

The end of the rebates did not seem to play a big enough role.

Tesla's mission is affordable transportation.

1. They have been holding off a bunch of idealistic students building cars for rich people.
2. The team wants ordinary people driving these cars. As soon as they are able they need to build cars for ordinary people.
3. There is not that much capital in-between where they are now and building a lot of cars for ordinary people - in time for the tax credit.

This winter is when Tesla will show if they are really a different kind of company or not. Game theory means not.

 

[MitchJi]

At this time, I don't think it's a good idea to persuade shorts to cover and get out of trouble. If they escape, they will come back to hurt Tesla again in the future (by shorting and spreading lies). In the past, when shorts asked me why I was so bullish about Tesla's long term future, I didn't give them detailed explanations. From Elon's point view, do you think he prefers to see shorts escape without a scratch? I do argue with shorts occasionally, but I will never write convincing details to help them.

In the big picture the effects of shorts on the SP are less than mouse nuts.

 

[kenliles]

weekend interest post
Interesting historical perspective re: GF
A legendary Ford plant could be inspiring Elon Musk's ideas about how to re-imagine factories — Business Insider UK

My kids (with our grandkids) just took this pic visiting the Smithsonian today
We've come far-- but not yet far enough