2017 Investor Roundtable:General Discussion

[Words of HABIT]

[QUOTE="We just don't know until we do."[/QUOTE]

Thanks Waiting4M3. What I am implying is that we should have a rough idea, by checking the M3 thread we have a good approximation of the numbers. Over the next several weeks and months we will be inundated with M3 weekly rates from various sources. Unless it comes from Elon himself, fact check for highest VIN spotted before blindly accepting the information. Best, Daniel

 

[Waiting4M3]

"We just don't know until we do."

Thanks Waiting4M3. What I am implying is that we should have a rough idea, by checking the M3 thread we have a good approximation of the numbers. Over the next several weeks and months we will be inundated with M3 weekly rates from various sources. Unless it comes from Elon himself, fact check for highest VIN spotted before blindly accepting the information. Best, Daniel

I've been tracking VINs, on average we see ~ 30 VINs per week (in Sep/Oct it was ~7/wk). The basic assumption is that the chance of seeing a M3 VIN scales with the # of M3 out there, but I think it's highly unscientific. On one hand as the employees get more and more cars, it's no longer very new and exciting for them, so fewer of them post about it on social media, so we're tilting lower. On the other hand, now we get batch sighting like in Playa Vista parking garage, and also at Fremont parking lot earlier this week, which also breaks the scaling assumption, tilting on the high side. We already know that Tesla doesn't build/deliver sequentially based on VIN, and VIN registration seems to have no regular interval, and invitation to configure only happened last week, with no new data to indicate any interval or frequency. There is nothing more we can read from these pieces of info until we get more. I suspect when we do it will be a lot more, and will make all these tea-leaf reading totally irrelevant.

 

[neroden]

WRT 45k Roadster reservations, I think that number is entirely feasible. I agree with Starno in that I don't trust employee rumors, but I put my deposit in last Tuesday. Haven't heard a peep out of Tesla since. It's possible that they're overwhelmed with reservations and that's the reason they haven't confirmed receipt of my deposit.

Or they're just lazy.

...they have a record of sloppy communications.

 

[mongo]

Based on VIN sighting I think we're at <500/wk for sure, maybe 100-200/wk. But that's delivery rate, there is a 1-2 week delay from production.

Also if we don't see many trucks, remember that they could be on trains, train tracks run right up to the factory based on what I've seen:

The cars would be in covered cargo containers. There is no way to tell from just the trucks. Also they would not use trains for low volume transportation, no point loading just 1 car in a train. But once the volume goes up , it would be a lot more economical to use trains.

We just don't know until we do.

Just a note, that pic is from before Tesla built the towers.
I think the train line runs inside now (easier loading?)

 

[Dennisj00]

I agree.
I'm tempted to put reservation in; even-though I don't know how would I pay for it, would wife kill me mercilessly quickly or would she take her own sweet time, and if my credit line has enough space to draw from for it...

If she kills you slowly, you ca still enjoy it for a while!

 

[Snapdragon III]

From my limited understanding of Australian peak power charges, they should make a killing on the spot market.

They might not make that much, but instead ruin it for everyone else?

 

[DragonWatch]

GM links electric vehicle effort to autonomous driving in attempt to compete with Tesla

One interesting piece of info from GM here is that trucks account for 1/3 of the miles traveled. Tesla will attack that 1000B miles market with one semi model, while GM is working on attacking the other 2000B miles, but with 18 models based on their EV platform. I'm not an expert but the economy of scale seems backward for GM.

So, are they using Google, their own or Mobileye self driving software/hardware? Since they are slow on the uptic, I hope they have someone good. Maybe they (Lutz) could ask Tesla

GM is the dimmest light bulb on the national tree:-(

Just keep doing what you have been doing for years until their are no needs for fossil fuel. Oh, I guess that was a clifffffff

 

[Bobfitz1]

But you're looking at profit from the battery POV, not from the vehicle. The model 3 production line is huge (almost half of the fremont factory is devoted to building the model 3, plus two new buildings to stage model 3 parts), because of the rate of production that's needed. Would a semi production line be smaller (due to lower production rate goals)? Would it fit within the gigafactory? If so, then wouldn't it be more profitable to sell the semi versus building out a new factory to build the model 3?

Not at all. Perhaps you did not see the original post which pointed out that 1,000 KWh of batteries can power 1 semi or 13 M3s at 75 KWh. This is simple math. Divide 1000 by 75. Tesla seems to strive for 25 or 30% Gross Margin on its vehicles. Say they are confident they will reach that on the M3 once production ramps to planned level. If the Semi reaches the GM Tesla aims for, ideally its GM would be same percentage as M3. Say that was 30% GM. $180,000 x .30 = $54,000 gross profit. The current LR M3 runs about $50K.
$50,000 x .30 = $15,000 gross profit. If you only had 1,000 Kwh of batteries available and had to choose between using them to make 1 semi or 13 LR M3, which would be the more profitable option just from the vehicle sales standpoint?
13 x $15K = $195,000. So if you went with using your batteries to build/sell 13 M3 instead of the 1 Semi, that would yield 3 1/2 times the profit compared to selling 1 Semi. So no, this is not looking at profit from the batteries, but from the vehicles, using a given KWh of batteries. Where they build the Semis is irrelevant. Tesla has calculated ALL costs for building Semi when they set the price for the 500 mile range version at $180K

I keep reminding commenters that this is a thought experiment. I'm not saying Tesla won't have all the batteries it needs to build tons of Semis in addition to all the M3 and MYs the public may want. Nor am I saying there may not be another way to make profits off Semi, like selling electricity it produces and stores to run fleets of Semis. Only if the supply of batteries was not sufficient to make all the cars Tesla can produce and sell when Semi ramps up, would they need to choose whether to limit production of Semis to make higher profits by maximizing car production. Human nature being what it is, it seems that many can't read and react to what's actually in a posting. Instead they read some conclusion into the observation they disagree with, and then explain why one conclusion or another is wrong. The post was only an observation about how, strictly from sales of cars vs Semis, a given quantity of batteries can generate higher profit used in cars than in Semis. No replies I've seen have refuted the straightforward math that shows this is the case.

 

[Reciprocity]

Not at all. Perhaps you did not see the original post which pointed out that 1,000 KWh of batteries can power 1 semi or 13 M3s at 75 KWh. This is simple math. Divide 1000 by 75. Tesla seems to strive for 25 or 30% Gross Margin on its vehicles. Say they are confident they will reach that on the M3 once production ramps to planned level. If the Semi reaches the GM Tesla aims for, ideally its GM would be same percentage as M3. Say that was 30% GM. $180,000 x .30 = $54,000 gross profit. The current LR M3 runs about $50K.
$50,000 x .30 = $15,000 gross profit. If you only had 1,000 Kwh of batteries available and had to choose between using them to make 1 semi or 13 LR M3, which would be the more profitable option just from the vehicle sales standpoint?
13 x $15K = $195,000. So if you went with using your batteries to build/sell 13 M3 instead of the 1 Semi, that would yield 3 1/2 times the profit compared to selling 1 Semi. So no, this is not looking at profit from the batteries, but from the vehicles, using a given KWh of batteries. Where they build the Semis is irrelevant. Tesla has calculated ALL costs for building Semi when they set the price for the 500 mile range version at $180K

I keep reminding commenters that this is a thought experiment. I'm not saying Tesla won't have all the batteries it needs to build tons of Semis in addition to all the M3 and MYs the public may want. Nor am I saying there may not be another way to make profits off Semi, like selling electricity it produces and stores to run fleets of Semis. Only if the supply of batteries was not sufficient to make all the cars Tesla can produce and sell when Semi ramps up, would they need to choose whether to limit production of Semis to make higher profits by maximizing car production. Human nature being what it is, it seems that many can't read and react to what's actually in a posting. Instead they read some conclusion into the observation they disagree with, and then explain why one conclusion or another is wrong. The post was only an observation about how, strictly from sales of cars vs Semis, a given quantity of batteries can generate higher profit used in cars than in Semis. No replies I've seen have refuted the straightforward math that shows this is the case.

You keep discounting the fact that the semi battery demand allows for the economies of scale required to make that 30% margin on the model 3. Otherwise, you don't build 150Gwh battery factory, you only build a 50GWh factory and the batteries cost more because you don't have the same economies of scale. Raw materials, vertical integration with partners and venders, custom built systems and large scale hardware all enhanced by offering 3x as much committed volume. Tesla's goal is not to maximize the profit per battery cell, but to sell as many batteries as humanly possible because that serves the mission and helps profitability long term.

That is why I questioned the timing, why a semi now and not a pickup truck or the model Y. The reason is 150GWh battery factory that gives Tesla the only edge they really have and the semi leverages that capacity with far less vehicles produced (100,000=3.4 million 3s). If Tesla could make 3.4 million 3s they shouldn't make the semi yet, but they can't. Tesla's main advantage over the next 5 years is that they own their own battery supply and no one else is even close to making that kind of commitment. But gigafactories only work at very large scales. Baby gigafactories do not allow for enough scale.

 

[EinSV]

Not at all. Perhaps you did not see the original post which pointed out that 1,000 KWh of batteries can power 1 semi or 13 M3s at 75 KWh. This is simple math. Divide 1000 by 75. Tesla seems to strive for 25 or 30% Gross Margin on its vehicles. Say they are confident they will reach that on the M3 once production ramps to planned level. If the Semi reaches the GM Tesla aims for, ideally its GM would be same percentage as M3. Say that was 30% GM. $180,000 x .30 = $54,000 gross profit. The current LR M3 runs about $50K.
$50,000 x .30 = $15,000 gross profit. If you only had 1,000 Kwh of batteries available and had to choose between using them to make 1 semi or 13 LR M3, which would be the more profitable option just from the vehicle sales standpoint?
13 x $15K = $195,000. So if you went with using your batteries to build/sell 13 M3 instead of the 1 Semi, that would yield 3 1/2 times the profit compared to selling 1 Semi. So no, this is not looking at profit from the batteries, but from the vehicles, using a given KWh of batteries. Where they build the Semis is irrelevant. Tesla has calculated ALL costs for building Semi when they set the price for the 500 mile range version at $180K

I keep reminding commenters that this is a thought experiment. I'm not saying Tesla won't have all the batteries it needs to build tons of Semis in addition to all the M3 and MYs the public may want. Nor am I saying there may not be another way to make profits off Semi, like selling electricity it produces and stores to run fleets of Semis. Only if the supply of batteries was not sufficient to make all the cars Tesla can produce and sell when Semi ramps up, would they need to choose whether to limit production of Semis to make higher profits by maximizing car production. Human nature being what it is, it seems that many can't read and react to what's actually in a posting. Instead they read some conclusion into the observation they disagree with, and then explain why one conclusion or another is wrong. The post was only an observation about how, strictly from sales of cars vs Semis, a given quantity of batteries can generate higher profit used in cars than in Semis. No replies I've seen have refuted the straightforward math that shows this is the case.

I think your math is hard to argue with and the scenarios you lay out do make for an interesting thought experiment. My guess is that your assumptions about projected margins are as good as can be made w/o further guidance from the company.

So assuming the math is basically correct why accelerate the Semi over other vehicles? $ for charging is one possibility. Another is that battery supply is not seen to be as big a constraint as we might expect by 2020 or 2021.

Another possibility is that this is a long-term play where Tesla is choosing to leverage its advantages with battery cost and volume to quickly snatch up market share and then benefit from the network effect of having the largest and best charging network.

Also, convoying has the potential to provide an enormous cost advantage for EVs that may be a relatively easy and early use case for demonstrating a relatively simple form of AP technology.

Another possible factor: Since trucking is a business the transition to electric should be driven mostly by economics, which if Tesla can deliver on pricing should overwhelmingly favor EVs. This may be quicker and easier than the consumer car market where old habits die hard and fears around new technology can create resistance to change. If trucking quickly switches to electric that may convince reluctant consumers that the technology is mainstream and can be trusted.

On the other hand, it is possible that batteries are expected to be a major production choke point through the early 2020s, or that Tesla needs time to refine Semi production, in which case we may see a slow initial Semi ramp up in favor of less battery hungry vehicles. I tend to think this is not going to happen and that Tesla will aim for a relatively quick ramp up but it’s really guesswork at this point.

 

[Reciprocity]

GM links electric vehicle effort to autonomous driving in attempt to compete with Tesla

One interesting piece of info from GM here is that trucks account for 1/3 of the miles traveled. Tesla will attack that 1000B miles market with one semi model, while GM is working on attacking the other 2000B miles, but with 18 models based on their EV platform. I'm not an expert but the economy of scale seems backward for GM.

This is the answer to why a semi now. The semi seems so out of place with rest of Tesla's products, but not when you consider that 100,000 semis = 3.4 million model 3s when you compare consumption only. If you think about Tesla's goal is not to sell cars but to displace fossil fuels, then what they really want to do is to sell as much solar + batteries as possible regardless of product or service. Also, you need a product like the semi to consume the 150GWh/Y of batteries that they will make in the gigafactory1 when it's completed. Remember, 500,000 model 3s is only 30-35GWh of batteries. Storage will require some, but not another 100GWh, which is.... 100,000 semis worth?

 

[Bobfitz1]

We'll find out, Feb 7th (or so) at the latest. They've said that they are behind in notifications due to high volumes. The market has way underestimated demand for the S as well as the 3 in the past, and I believe that this time is no different.

So you believe that in 1 - 2 weeks 45,000 people put deposits in on a 250K car, when only 6 times that many, in a week, put in a 1K deposit for M3 which was advertised with a base price of $35K?

 

[Mkorpal]

Another reason I see the Semi being moved up is that Tesla can save quite a bit of money moving hundreds of thousands of cars and battery packs with their own fleet of trucks. How much do you think it costs them to ship each battery pack from Sparks to Fremont?

 

[EinSV]

You keep discounting the fact that the semi battery demand allows for the economies of scale required to make that 30% margin on the model 3. Otherwise, you don't build 150Gwh battery factory, you only build a 50GWh factory and the batteries cost more because you don't have the same economies of scale. Raw materials, vertical integration with partners and venders, custom built systems and large scale hardware all enhanced by offering 3x as much committed volume. Tesla's goal is not to maximize the profit per battery cell, but to sell as many batteries as humanly possible because that serves the mission and helps profitability long term.

That is why I questioned the timing, why a semi now and not a pickup truck or the model Y. The reason is 150GWh battery factory that gives Tesla the only edge they really have and the semi leverages that capacity with far less vehicles produced (100,000=3.4 million 3s). If Tesla could make 3.4 million 3s they shouldn't make the semi yet, but they can't. Tesla's main advantage over the next 5 years is that they own their own battery supply and no one else is even close to making that kind of commitment. But gigafactories only work at very large scales. Baby gigafactories do not allow for enough scale.

This could be a factor when Semi production is low volume but aren’t they going to need an additional GF to produce 100K+ Semis? My assumption is that they are also going to be ramping up Model Y production and TE using GF1 batteries in roughly the same time period.

 

[Oil4AsphaultOnly]

Not at all. Perhaps you did not see the original post which pointed out that 1,000 KWh of batteries can power 1 semi or 13 M3s at 75 KWh. This is simple math. Divide 1000 by 75. Tesla seems to strive for 25 or 30% Gross Margin on its vehicles. Say they are confident they will reach that on the M3 once production ramps to planned level. If the Semi reaches the GM Tesla aims for, ideally its GM would be same percentage as M3. Say that was 30% GM. $180,000 x .30 = $54,000 gross profit. The current LR M3 runs about $50K.
$50,000 x .30 = $15,000 gross profit. If you only had 1,000 Kwh of batteries available and had to choose between using them to make 1 semi or 13 LR M3, which would be the more profitable option just from the vehicle sales standpoint?
13 x $15K = $195,000. So if you went with using your batteries to build/sell 13 M3 instead of the 1 Semi, that would yield 3 1/2 times the profit compared to selling 1 Semi. So no, this is not looking at profit from the batteries, but from the vehicles, using a given KWh of batteries. Where they build the Semis is irrelevant. Tesla has calculated ALL costs for building Semi when they set the price for the 500 mile range version at $180K

I keep reminding commenters that this is a thought experiment. I'm not saying Tesla won't have all the batteries it needs to build tons of Semis in addition to all the M3 and MYs the public may want. Nor am I saying there may not be another way to make profits off Semi, like selling electricity it produces and stores to run fleets of Semis. Only if the supply of batteries was not sufficient to make all the cars Tesla can produce and sell when Semi ramps up, would they need to choose whether to limit production of Semis to make higher profits by maximizing car production. Human nature being what it is, it seems that many can't read and react to what's actually in a posting. Instead they read some conclusion into the observation they disagree with, and then explain why one conclusion or another is wrong. The post was only an observation about how, strictly from sales of cars vs Semis, a given quantity of batteries can generate higher profit used in cars than in Semis. No replies I've seen have refuted the straightforward math that shows this is the case.

Fair enough, I did miss your initial reasoning. But gross margin is simply the profit margin from building the vehicle. The operating margin for the semi should be higher shouldn't it? Sales and service (warranty support) for 13 model 3's would be significantly more than the sales and service of a single semi. Perhaps enough to overcome the gross profit difference?

Anyway, leveraging off my original point, this thought exercise is flawed, as it's scoped under the assumption that Tesla is battery-production-limited. The situation is that in 2019, once the gigafactory is nearing completion, Tesla will be vehicle-production-limited.

 

[Bobfitz1]

I think your math is hard to argue with and the scenarios you lay out do make for an interesting thought experiment. My guess is that your assumptions about projected margins are as good as can be made w/o further guidance from the company. So assuming the math is basically correct why accelerate the Semi over other vehicles? $ for charging is one possibility. Another is that battery supply is not seen to be as big a constraint as we might expect by 2020 or 2021.

Another possibility is that this is a long-term play where Tesla is choosing to leverage its advantages with battery cost and volume to quickly snatch up market share and then benefit from the network effect of having the largest and best charging network.

On the other hand, it is possible that batteries are expected to be a major production choke point through the early 2020s, or that Tesla needs time to refine Semi production, in which case we may see a slow initial Semi ramp up in favor of less battery hungry vehicles. I tend to think this is not going to happen and that Tesla will aim for a relatively quick ramp up but it’s really guesswork at this point.

Thank you @EinSV for reading the post as written. One or all of the other possibilities you list are very plausible and one or more are likely to play out. I posted the simple math to stimulate ideas about how or why the vehicle profit differential would not lead to going slow on Semi ramp up to high volumes in first few years. A subsequent post replying to @MitchJi ends by emphasizing the obvious. Many, many GFs will need to be built in only a few years to meet the demand created by the popularity of Tesla's exceptional products. The scale is enormous and no other company in the world even contemplates doing anything like it. Today Electrik reported on SK Innovation announcing plans to build a 7.5 GWh 'GigaFactory' in Hungry, opening in 2020, to provide batteries to the big Euro ICE companies.
The lack of significance to this is clear when a little easy math shows that this amount of batteries will only cover building 100,000 75Kwh vehicles annually. Starting in 2020!

 

[mongo]

Overly simplistic late evening thoughts:

Cells will be at the lowest cost when GF is at maximum production rate (first for the equipment they have, then for the max equipment that fits). Semi, and to a smaller extent Roadster, boost cell usage which drops cell cost which makes everything more profitable.

Additionally, perhaps there is a concern that without these extra vehicles (along with huge storage products) GF will be underutilized in a few years.

Maybe the emergency fix on 3's module line revealed an efficency improvement such that production can be higher than expected. (Though everything can be paralleled with space, so likely not a big factor.)

 

[Reciprocity]

This could be a factor when Semi production is low volume but aren’t they going to need an additional GF to produce 100K+ Semis? My assumption is that they are also going to be ramping up Model Y production and TE using GF1 batteries in roughly the same time period.

If they could make 3.4 million S3XY vehicles, they should do that instead. But they can't do that as quickly as they can make half a million S3X cars and 50,000 semis. If a fully operational gigafactory truly can make 150GWh of batteries, then that is like 2.5 million model 3s. If the gigafactory maxes out at 100GWh, that's 500K cars and 50k semis and some storage. The semi timing seems two fold, gigafactory supply and offsetting large amounts of fossil fuel consumption with one product instead of a half dozen.. S3XY+roadster+pickup. Those other car programs will come, but not as quickly as Tesla can build and sell 50,000-100,000 semis. They bed now car factories as well, at 500,000 they are maxed out at Fremont. Specially after they start making semis there.

 

[Reciprocity]

Overly simplistic late evening thoughts:

Cells will be at the lowest cost when GF is at maximum production rate (first for the equipment they have, then for the max equipment that fits). Semi, and to a smaller extent Roadster, boost cell usage which drops cell cost which makes everything more profitable.

Additionally, perhaps there is a concern that without these extra vehicles (along with huge storage products) GF will be underutilized in a few years.

Maybe the emergency fix on 3's module line revealed an efficency improvement such that production can be higher than expected. (Though everything can be paralleled with space, so likely not a big factor.)

Great minds think alike and simplistically?

 

[generalenthu]

I believe the original driver for marginal demand was supposed to be Tesla energy. The reality is TE can only scale at a glacial pace, because the customers are utilities (monopolies) and they are driven by politics more than economics. The SA battery is the exception rather than the rule because the situation got really dire.

My hypothesis is Tesla figured the razor + blade model of semi and mega charging let's them diversify their demand side a bit more efficiently.

Plus there is the argument around scaling quickly to get market share in as many big markets as possible.

Edit: Puerto Rico would / could be a major demand driver for TE, but there has been no news on it recently. Not sure they got cash to do anything given their financial troubles, FEMA funds notwithstanding.