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

[Fred42]

They said tera kilowatt hours. I.e. petawatts.

Is the correct term terawatt hours? Or was tera kilowatt hours correct?

 

[KarenRei]

Is the correct term terawatt hours? Or was tera kilowatt hours correct?

Should have been terawatt hours. That's the right scale for the task at hand.

 

[wnorris]

There wasn't anything technologically complex about what Amazon has achieved in the online retailer business. It has been relatively easy to understand but yes, costly and time consuming to achieve.

What Amazon (via Jeff Wilke) did during the early 2000s was indeed 'technologically complex'. I was just starting out at the time and was fortunate enough to cross paths during Jeff's initial tour through Amazon's locations. A lot of the technology that went into Amazon's logistics systems helped shape Amazon's ability to continually lower costs to consumers and squeeze out competitors. What he did for Amazon's distribution centers was transformational for the company.

I only wish I had not sold stock during that time period. I see very similar parallels with where Tesla is at now.

 

[mongo]

What Amazon (via Jeff Wilke) did during the early 2000s was indeed 'technologically complex'. I was just starting out at the time and was fortunate enough to cross paths during Jeff's initial tour through Amazon's locations. A lot of the technology that went into Amazon's logistics systems at the time helped shape Amazon's ability to continually lower costs to consumers and squeeze out competitors. What he did for Amazon's distribution centers was transformational for the company.

I only wish I had not sold stock during that time period. I see very similar parallels with where Tesla is at now.

Agreed, freespace based product storage was revolutionary and required a non-typical SW back end to support. Then they pared that with a robot army to bring the shelves to the people.

 

[Cosmacelf]

Just did some math. Sure looks like Raven is going to need a new shift in Q3, or Q4 at the latest, barring unforseen circumstances.

From the ER, we learn that:

* They produced "over 14,500" S/X in Q2
* They delivered 17,722 S/X in Q2
* Inventory has fallen to 18 days of sales

From this we can calculate that they ran a production deficit of "less than" (but probably within 100 of) 3222 S/X. We can also calculate that they have about 3500 S/X in global inventory. Barring:An inexplicable drop in demand (which is growing instead, including the fact that Raven appears to have quite a waitlist on deliveries)

• A price rise to deliberately reduce demand, earning more margins in exchange (Tesla is usually loathe to do this)
• Tesla finding a way to produce significantly more S/X with a single shift (would be awesome, but over 20% QoQ growth with existing lines?)

.... they're going to need to add a second shift. If not in Q3, then in early Q4 at the very latest.

A second shift and a higher Raven fraction will hurt FCF but improve net income / profits vs. Q2.

A small nit ... Raven doesn’t appear to have much of a waitlist here in the US. A friend of mine ordered in late June and got his Model S within 2 weeks, and I ordered my LR Model X 2 weeks ago and will get delivery today. Both In southern CA, so short delivery logistics.

This might just mean that Tesla is ramping production very tightly to meet supply, and we know they have production headroom.

 

[SPadival]

I cant be bothered to wade through the accounting posts to check if this has been linked - Ashlee Vance has an article out on JB and he managed to talk to JB

“It’s been an amazing time, and I really, really love the mission and this personal connection and ownership with the whole company,” Straubel said in a phone interview. “Tesla has evolved. What we need now is a focus on sales, delivery and manufacturing. I have been helping with that in recent years, but it’s not what I am best at. There are people in the world who are better at this stuff and enjoy it more.”

Bravo JB. As a shareholder, I am really appreciative of the fact that Tesla has such great leaders.

Oh.. and this is bullish AF. As babycharts would have said, LFG

Tesla Loses a Founder, and a Piece of Its Soul

 

[KarenRei]

A small nit ... Raven doesn’t appear to have much of a waitlist here in the US. A friend of mine ordered in late June and got his Model S within 2 weeks, and I ordered my LR Model X 2 weeks ago and will get delivery today. Both In southern CA, so short delivery logistics.

This might just mean that Tesla is ramping production very tightly to meet supply, and we know they have production headroom.

Well, in Europe they're only just recently starting to appear. Will be nice when that changes

 

[wnorris]

The emergence of Zack Kirkhorn is in my view the single most positive feature of the past year.

It might just be me, but Elon seems to be letting his subordinates take more questions than in the past. This past Q's call it was evident that Zack's answers actually address the question being asked. Elon's answers tend to be repetitive and focused more on some point in the future. I always enjoy the more succinct answer from Zach, JB, or Jerome.

 

[S3XY]

Because sometimes you do.

Because not everybody has 300 plus mile range on their Tesla.

Because not everybody drives in ideal conditions all the time. Because sometimes having a AC full blast or Sentry mode eats into your range.

Because sometimes you forget to plug in or because you don't have access to a plug.

Because sonetimes it's the middle of Winter.

 

[EVDRVN]

I cant be bothered to wade through the accounting posts to check if this has been linked - Ashlee Vance has an article out on JB and he managed to talk to JB



Bravo JB. As a shareholder, I am really appreciative of the fact that Tesla has such great leaders.

Oh.. and this is bullish AF. As babycharts would have said, LFG

Tesla Loses a Founder, and a Piece of Its Soul

Good thing he thinks CS needs no attention.

 

[KarenRei]

It might just be me, but Elon seems to be letting his subordinates take more questions than in the past. This past Q's call it was evident that Zack's answers actually address the question being asked. Elon's answers tend to be repetitive and focused more on some point in the future. I always enjoy the more succinct answer from Zach, JB, or Jerome.

Elon is great on stage, speaking to people who get and share his vision. He's not so great dealing with doubting analysts; it feels as though it's because the premises he operates on are so established to him on so many levels that it feels unnatural to him to have to explain them. So people hear things like "2 TWh per year", and if you're onboard with Tesla's vision, that's 'duh', of course they need to be thinking of that now, because they're going to be needing that in, say, five years time, and you have to start laying the roadmap this early. But for people who question whether Tesla is even going to survive a year, or think EVs are going to be a niche or slow-growing market and can't imagine an S-curve, you need someone who's willing to act like a grade school teacher, go back to basics, and walk them up to speed.

 

[SPadival]

Good thing he thinks CS needs no attention.

What does CS stand for? Concern-trolling Service?

It is obvious to anyone that he mentioned only areas where he helped out.

 

[CorneliusXX]

Is the correct term terawatt hours? Or was tera kilowatt hours correct?

the base unit is watthours. Kilo, Mega, Giga, Tera are all just orders of magnitude of the watthour. So a tera kilowatt hour is multiplying orders of magnitude.

 

[AlMc]

I cant be bothered to wade through the accounting posts to check if this has been linked - Ashlee Vance has an article out on JB and he managed to talk to JB



Bravo JB. As a shareholder, I am really appreciative of the fact that Tesla has such great leaders.

Oh.. and this is bullish AF. As babycharts would have said, LFG

Tesla Loses a Founder, and a Piece of Its Soul

IF JB had left two years ago I would have vacated all TSLA positions. As it is now I am OK with it and wish him well. (OK, he is not gone but a 'consultant'...right).

Now, if Jerome leaves before the semi is out in full production....I am out of TSLA.

Anyone following EM for the past few years, (he warned us about volatility in the SP), realizes he does not care about the ST SP go TSLA. He only cares, IMO, when it impacts the mission. Completing the mission means access to $.

I have learned that most people (including me) should not be 'all in' on TSLA unless you have the ability to replenish all your risked assets through earnings/inheirtance/bank robbery..j/k) AND you are willing to hold your position for a minimum of five years.

 

[dc_h]

Great observations.

Consider too that Drew Baglino has been an essential (but not in limelight) part of the team for ages. He's now well positioned to take over many of J.B.'s duties. Tesla has a unique culture and I think promoting from within will be the company's best strategy going forward. By the time someone outside of Tesla has acquired the talent to be a top decision maker, they're most often no longer of a disposition to transition into the pressure cooker, working with Elon, and their time at the company is typically short. Additionally, changing strategies on a dime to something unconventional when needed is likely foreign to most non-Tesla types.

Hopefully moving Drew up also means some upward mobility for his direct reports. Turnover isn’t all bad. You need a path up for new stars, or they’ll go to Apple and Waymo, etc. I like the idea of an Engineering Emeritus for distinguished Tesla staff. Get them out of the grind and let them focus on a single issue or set of issues that require singular focus. Give them a chance to have only technical meetings, more normal hours and recharge doing what they love.

 

[Doggydogworld]

Just did some math. Sure looks like Raven is going to need a new shift in Q3, or Q4 at the latest, barring unforseen circumstances.

From the ER, we learn that:

* They produced "over 14,500" S/X in Q2
* They delivered 17,722 S/X in Q2
* Inventory has fallen to 18 days of sales

From this we can calculate that they ran a production deficit of "less than" (but probably within 100 of) 3222 S/X. We can also calculate that they have about 3500 S/X in global inventory. Barring:An inexplicable drop in demand (which is growing instead, including the fact that Raven appears to have quite a waitlist on deliveries)

• A price rise to deliberately reduce demand, earning more margins in exchange (Tesla is usually loathe to do this)
• Tesla finding a way to produce significantly more S/X with a single shift (would be awesome, but over 20% QoQ growth with existing lines?)

.... they're going to need to add a second shift. If not in Q3, then in early Q4 at the very latest.

A second shift and a higher Raven fraction will hurt FCF but improve net income / profits vs. Q2.

18 days inventory is an overall measure, dominated by Model 3 which is now >80% of unit sales.

Through 6/30/19 Tesla has built 424.9k S/X and delivered 414.3k. Of the 10.6k undelivered cars, they've moved about 1.6k older loaners from Finished Goods Inventory to PP&E. I doubt they're actively trying to sell those, but would if they got an offer. That leaves 9k later model S/X in inventory. This includes S/X in transit to buyers plus demos and loaners with mileage adjustments on the asking price.

Total undelivered S/X/3 is 25.5k, of which ~24k are in FGI and ~1.6k in PP&E. That's more than 18 days at Q2 average sales rate, but my guess is they subtracted the 7.4k in transit before doing the calculation. Alternatively, they may have used a higher sales rate based on the back half of the quarter.

I don't foresee a 2nd S/X shift. They would have mentioned it in the letter and/or call. Instead they talked about managing inventory, working capital, etc. @neroden thinks they improved the S/X lines while reconfiguring/combining them and can now build 18-20k S/X with a single shift. 18k over 12 production weeks is 1500/week. That's a 96 second takt time, still very modest. Seasonal OT could boost output to 20k in high demand quarters.

 

[Doggydogworld]

Tesla's $500m Phase 1 capex guidance is just too low to be feasible for full capex of Phase 1 and 3K per week production (c.$7bn annual revenue, $1.5-2bn gross profit). I don't think Tesla is lying though, so the only conclusion is that Tesla themselves are not paying the full capex of Phase 1

They are building:

• A huge high tech factory at rapid speed with high construction staff overtime
• Full surrounding infrastructure including roads and electricity stations.
• Stamping, Body shop, Welding robots, General Assembly Line and Paint shop equipment.

Admittedly they are saving capex on Battery module/pack lines by sending the old LR pack lines from GF1.

In addition to the $500m Phase 1 capex guidance, Tesla did already spend $141m in 2018 on land purchase in Shanghai (accounted as a prepaid 50 year lease rather than capex).

We know Nio previously made a deal with Shanghai for its factory (though China later banned Nio from building the factory until Tesla's is finished). The deal was that Shanghai will build Nio's factory & rent it to Nio for free for 5 years & discounted for another 5 years. Shanghai also offered to guarantee local debt to finance half of Nio's $650m equipment capex.

One bit of info we got from Tesla on GF3 earlier this year was:
"As part of this project, we have agreed with the local government to spend approximately $2 billion in capital expenditures over the next five years (which is already included in our capital expenditure plans), and to generate approximately $270 million of annual tax revenues starting at the end of 2023. We believe the tax revenue target will be easily attainable even if our production were far lower than the volumes we are planning, although if we are unwilling or unable to meet such target or obtain periodic project approvals, we would be compensated for the remaining value of the land lease, buildings and fixtures and revert the site to the government."

I think Tesla's deal is different to Nio's, most likely Tesla do actually own the building, but they are just getting China to pay for it, or at least to offer a far below cost bid for the construction contract. It looks to me like Tesla promised a certain level of tax revenues in return for heavily subsidised construction costs, and if Tesla does not meet the tax targets they have to return the site to China as penalty.

What are your thoughts on the structure of Tesla's deal for GF3? @neroden @Doggydogworld ?

You pretty much covered the high points. I think the government is paying for the building and equipment. Tesla's 500m is for tooling and incidentals. I think it's the same deal Nio described. Tesla is technically a tenant paying nominal rent as long as they meet milestones. Similar to GF2 in Buffalo, also.

As with Buffalo Tesla will put the GF3 buildaing and equipment on the balance sheet, perhaps as an Operating Lease Right of Use asset, with an offsetting liability. But also like Buffalo they won't run it through the capex account. If Tesla were the legal owner of the building and equipment they'd have to run it through capex, even if financed 100% with loans. Construction and equipment costs would be in "Cash Flows from Investing Activities" and borrowing would be "Cash Flows from Financing Activities".

Perhaps Tesla will explain this in the 10-Q. But don't hold your breath.

 

[KarenRei]

18 days inventory is an overall measure, dominated by Model 3 which is now >80% of unit sales.

"18 days of sales" is listed under the Model S/X section. It would be highly misleading for S/X to be, say, 40 days of sales and Model 3 be 12 days of sales, and then write, under Model S/X, "Our deliveries increased sequentially to 17,722 as we continue to prioritize inventory reduction (working capital management). As a result, our total new car inventory levels have fallen to just 18 days of sales...". I'd say almost criminally misleading. I guess they could get away with it for including the word "total"?

Is there perhaps some confusion between new and used inventory here?

Through 6/30/19 Tesla has built 424.9k S/X and delivered 414.3k. Of the 10.6k undelivered cars, they've...

From what estimate are you getting all this? I write "estimate" because I can't imagine that you have any more detailed hard data about Q2 than what's been put out in the ER - e.g. no 10-Q that might have provided more colour. Or is this just some accumulation of past reports, incl. accumulating any errors?

@neroden thinks they improved the S/X lines while reconfiguring/combining them and can now build 18-20k S/X with a single shift.

So you're thinking that you're going to take two lines, merge them into one (e.g. one line gone), and get more throughput than when you had two? And that you're going to do said merger in a small fraction of a single quarter? Or do you think that they've already done said merger? I've talked with people who've done tours since the Q2 shutdown - both lines are still there, there's been no merger. The changes were exactly what was said on the earnings call: they were freeing up parts warehousing space for the Y line.

In all reasonableness, 18-20k in Q3 in a single shift is not happening. They might (might) see some improvement in S/X line productivity, but Q3 isn't going to see anything close to 18-20k in a single shift, unless Tesla pulls off a minor miracle. Q2 S/X production rose by only 354 vehicles relative to Q1.

 

[Webeevdrivers]

Curious if Tesla has published how many model 3’s have been built. Are they over 300k yet?

 

[ReflexFunds]

You pretty much covered the high points. I think the government is paying for the building and equipment. Tesla's 500m is for tooling and incidentals. I think it's the same deal Nio described. Tesla is technically a tenant paying nominal rent as long as they meet milestones. Similar to GF2 in Buffalo, also.

As with Buffalo Tesla will put the GF3 buildaing and equipment on the balance sheet, perhaps as an Operating Lease Right of Use asset, with an offsetting liability. But also like Buffalo they won't run it through the capex account. If Tesla were the legal owner of the building and equipment they'd have to run it through capex, even if financed 100% with loans. Construction and equipment costs would be in "Cash Flows from Investing Activities" and borrowing would be "Cash Flows from Financing Activities".

Perhaps Tesla will explain this in the 10-Q. But don't hold your breath.

I agree this looks plausible, the only odd things if Tesla did infact get the same deal as Nio is: 1) Why did they have to buy the 50 year land lease? and 2) Why would Tesla be compensated for the remaining value of the land, buildings and fixtures when transferred to the government if they had never owned these in the first place?