2017 Investor Roundtable:General Discussion

[dhrivnak]

Anyone here borrowing money to buy shares? I currently have $30k of my own money in tsla. I have a fee free offset mortgage on my paid off house. It would cost me $1200 in interest to borrow 30k and buy tsla. Worst case scenario, tsla drops by half, I can still sell everything and pay my loan back.

So far I'm up 35% on my initial investment. I'm very very certain that I'll make enough to pay for the interest, as I see tsla only going upwards from the model 3 launch.

Worst worst case, I'm adding a year to pay my mortgage back. Seems pretty low risk.

This is just me but I would not borrow money to buy stock. I am anti risk and want to keep my house paid off. Now with som cash I would buy Tesla I just would not borrow to do so.

 

[Snapdragon III]

I updated the unit projection for 2H17 and kept margin projection the same.

I was too optimistic for 3Q17 and slightly optimistic for 4Q17. I underestimated the time it takes to take delivery of parts ordered, assemble them into cars, and deliver to customers.

Having said that, however, the +100% month-over-month exponential growth rate throughout 4Q17 gives me more confidence for my 2018 projections, which is really what matters to the future of Model 3, and hence, the stock price.

The following table presents my Model 3 expectation for the next six quarters. Note that I have my EAD/FSD option uptake projection in a separate section, hence the Model 3 ASP assumption of $40k.

View attachment 233867

I think you are off base predicting linear, or exponential increases in production past 5000 cars per week. My memory of the last conference call, (I could be misremembering, or have misunderstood) is that they discussed this. I think they said the initial line they were building is sized for 5000 cars a week, and it will be an S curve getting it up to capacity. After that, to get to 10,000 cars a week they will have to make a major capital investment. I assume they will have to duplicate at least a portion of the line. They said it would not be nearly as expensive as the initial line, but still significant. JB came into the discussion and mentioned that the Gigafactory part of things was already built. Elon also talked about maybe moving the time frame up a few months, as the months cash flow from 5000 cars was amazing, and just a couple months difference made a big difference. I took this last to be implying another capital raise might be necessary to do this, but it may be worth it?

My impression was that they would get to 5000/week at the end of 2017, and the original plan was to build the new line similar in timing to this year, maybe start it in July, and s curve it up to full speed to hit 10,000 at the end of 2018. With a possible option to bump it all forward a few months if everything is going well.

Again, This is from my fuzzy memory, and anyone reading it should go back and listen to the call themselves before making an investment decision.

 

[neroden]

I don't understand why you need Level 5 to have a robust network of autonomous cars....Eventually the Level 4 network could be active the vast majority of the time in most metropolitan areas without needing to be Level 5.

What am I missing?

Level 4 can't reliably arrive at your house on its own. It needs a sober, awake person in the car... or it may simply abort and park at the side of the road and never show up at your house. (I suppose it could be programmed to phone you and tell you that it bailed out.)

I think that's what you're missing. Level 4 is still great, but it doesn't readily enable all these "transportation as a service" scenarios people are throwing around.

Or maybe what you're missing is that we're still a long way from Level 4 working well on arbitrary city streets.

 

[neroden]

I am personally turn between the camp that say FSD is harder then anyone thinks and won't be here in 10 years and those who say it will be here by year end. I could see level 3 by December. My opinion is that the only difference between 4 and 5 is a human in the car. If you can do 4, 5 is only about validation and proof that it is 10x safer then a human.

Regarding going from level 4 to level 5: In trains, it was already proven that it was much safer, and public opinion still wouldn't let the trains operate without a human. Musk said it had to be 100x safer than a typical human before people would allow cars with no drivers... and I think the evidence shows that this is about right to get public acceptance.

Regarding going from level 3 to level 4.... well, I think it's much harder than any of the people working on it believe. Ice storm, tree across the expressway. Kangaroos. Bizarre, poorly marked detours. Bridges out. Flooding. On and on and on, it's just an endless list of corner cases. They'll have to take one corner case at a time (well, technically, the work can be parallelized, but that requires way more employees than they have), and they haven't really started in on making a *list* of corner cases yet. They can't handle unstriped roads yet.

Even if FSD us more then 2 years away, I expect we will start to see FSD related safety features sooner then later. Red light and stop sign breaking, even if AP is not enabled or engaged for example.

THAT stuff can be implemented RIGHT NOW and I really hope it is implemented immediately. I would like all cars to *refuse to tailgate*, *refuse to crash into things they see*, *refuse to run stop signs*, etc. (I've been referring to cars with these features as "I'm sorry, Dave, I can't do that" cars. I want everyone to be forced to have such cars, so they will stop running into me.)

 

[Reciprocity]

Level 4 can't reliably arrive at your house on its own. It needs a sober, awake person in the car... or it may simply abort and park at the side of the road and never show up at your house. (I suppose it could be programmed to phone you and tell you that it bailed out.)

I think that's what you're missing. Level 4 is still great, but it doesn't readily enable all these "transportation as a service" scenarios people are throwing around.

Or maybe what you're missing is that we're still a long way from Level 4 working well on arbitrary city streets.

The main difference between 4 and 5 is the person is required for 4 and not for 5. The reason that a person is required is because there will be geofenced areas that require a person. In theory, 4 wouldn't require a person if it never went to those places, but you would not be able to get sign off from regulators to do that, so you must have 5 for a service as it has been described by Elon. But, that does not mean that you could have the cars park at a destination and wait for another passenger that sits in the drivers seat, but does not need to drive if they dont want to. This is not what Elon has described so I doubt Tesla would provide this service, but it wouldn't stop someone else from doing it. Rentcentric is a software that allows for such a scheme today, even control the locks and allow the car to start and it does work with Tesla's as some are already using it. This would also only work in more densely populated areas as people would take your car to the suburbs and no one would there to bring it back.

 

[ValueAnalyst]

I think you are off base predicting linear, or exponential increases in production past 5000 cars per week. My memory of the last conference call, (I could be misremembering, or have misunderstood) is that they discussed this. I think they said the initial line they were building is sized for 5000 cars a week, and it will be an S curve getting it up to capacity. After that, to get to 10,000 cars a week they will have to make a major capital investment. I assume they will have to duplicate at least a portion of the line. They said it would not be nearly as expensive as the initial line, but still significant. JB came into the discussion and mentioned that the Gigafactory part of things was already built. Elon also talked about maybe moving the time frame up a few months, as the months cash flow from 5000 cars was amazing, and just a couple months difference made a big difference. I took this last to be implying another capital raise might be necessary to do this, but it may be worth it?

My impression was that they would get to 5000/week at the end of 2017, and the original plan was to build the new line similar in timing to this year, maybe start it in July, and s curve it up to full speed to hit 10,000 at the end of 2018. With a possible option to bump it all forward a few months if everything is going well.

Again, This is from my fuzzy memory, and anyone reading it should go back and listen to the call themselves before making an investment decision.

I am assuming a linear (not exponential) increase beyond 4Q17. Also, no new line or massive incremental investment is needed to get to 10,000 per week (see below from 4Q16 call). I agree that, if they can pull forward Model 3 production to say "10,000 per week some time in 1Q18," a capital raise would be "worth it," but I don't think it will be necessary given what I'm assuming: 10,000 per week some time in 4Q18.

----------

Elon Reeve Musk - Tesla, Inc.

Yes, I mean, there's obviously going to be a fair bit of incremental investment to go from 5,000 cars a week to 10,000 cars a week, but it's going to be a lot less than getting to 5,000 cars a week in the first place. We don't know exactly what that's going to be except I'm confident it'll be less. Because the first thing we'll try to increase output is going back to rocket equation is to increase exit velocity of the line. And we don't know exactly where the trouble points (64:53) are going to be. We tried to model it out as carefully as possible, but there'll be things that aren't captured in the model.

But I think in a lot of cases, we'll simply be able to run the lines faster as opposed to duplicate the line. That's by far the best CapEx maneuver is just to make it go faster. But I would say it's going from 5,000 to 10,000 is probably – this is a total wild-ass guess, so (65:35) right way to think about, but it's like somewhere between 50% to 70% of the cost of the 5,000 line. Something like that.

If you're lucky and smart, 50% is only half the game, which obviously is pretty awesome from a CapEx standpoint. I can't imagine it being more than about 70% as much as there is. So, JB, what do you think?

Jeffrey B. Straubel - Tesla, Inc.

Yes, I think that's right. And it's maybe helpful to realize, but a lot of the infrastructure investments to get all the way to 10,000 are already completed, Gigafactory in particular.

 

[EnzoXYZ]

Anyone here borrowing money to buy shares? I currently have $30k of my own money in tsla. I have a fee free offset mortgage on my paid off house. It would cost me $1200 in interest to borrow 30k and buy tsla. Worst case scenario, tsla drops by half, I can still sell everything and pay my loan back.

So far I'm up 35% on my initial investment. I'm very very certain that I'll make enough to pay for the interest, as I see tsla only going upwards from the model 3 launch.

Worst worst case, I'm adding a year to pay my mortgage back. Seems pretty low risk.

I do it as it is such low risk for a nice potential pay off. If you can wait 2-3 years out with out financial difficulties all of the better. I'm very risk tolerant as I'm pretty much 100% TSLA.

 

[EinSV]

Level 4 can't reliably arrive at your house on its own. It needs a sober, awake person in the car... or it may simply abort and park at the side of the road and never show up at your house. (I suppose it could be programmed to phone you and tell you that it bailed out.)

I think that's what you're missing. Level 4 is still great, but it doesn't readily enable all these "transportation as a service" scenarios people are throwing around.

Or maybe what you're missing is that we're still a long way from Level 4 working well on arbitrary city streets.

To me, it's helpful to separate two distinct issues: (1) what capabilities are required for a commercial self-driving network to be viable; and (2) how long will it take to develop those capabilities. I focused on item 1, which I think is useful because I don't see much discussion of it and because there are widely varying opinions on item 2.

Back to item 1. ...

I don't think your post proves the case for needing Level 5. Driving a car currently is far from perfectly reliable. Cars break down, overheat and get flat tires. Owners become incapacitated for various reasons (fatigue, alcohol/drugs, health issues, etc.).

Just as normal driving does not need to be perfectly reliable to be useful, neither does an autonomous network.

If the self-driving car malfunctions, the network can send another self-driving car to get you. Or just like calling a tow truck if you get a flat tire or have a breakdown on the freeway, the network could maintain a roster of drivers to pick you up if you are stuck in a situation that for whatever reason the network of self-driving cars can't handle. It seems to me that given the dollars at stake, there are any number of solutions that could be gap fillers until the networked cars have fewer hiccups.

 

[neroden]

I think you are off base predicting linear, or exponential increases in production past 5000 cars per week. My memory of the last conference call, (I could be misremembering, or have misunderstood) is that they discussed this. I think they said the initial line they were building is sized for 5000 cars a week, and it will be an S curve getting it up to capacity. After that, to get to 10,000 cars a week they will have to make a major capital investment. I assume they will have to duplicate at least a portion of the line. They said it would not be nearly as expensive as the initial line, but still significant. JB came into the discussion and mentioned that the Gigafactory part of things was already built. Elon also talked about maybe moving the time frame up a few months, as the months cash flow from 5000 cars was amazing, and just a couple months difference made a big difference. I took this last to be implying another capital raise might be necessary to do this, but it may be worth it?

My impression was that they would get to 5000/week at the end of 2017, and the original plan was to build the new line similar in timing to this year, maybe start it in July, and s curve it up to full speed to hit 10,000 at the end of 2018. With a possible option to bump it all forward a few months if everything is going well.

Again, This is from my fuzzy memory, and anyone reading it should go back and listen to the call themselves before making an investment decision.

Hmm. You may be right.

Given Tesla's current capital structure (recent increase in ABL etc) I strongly suspect they can build those duplicated portions of the line without issuing more stock. Musk stated that the Model 3 ramp-up was "self-financing" due to the terms they're getting from suppliers: delivery now, pay 60 or 90 days later. This doesn't really kick in immediately; they're getting parts for 1000 cars a week in July and aren't producing that many cars in September; but come December it really does kick in, and for any new / duplicated section of line, this will be operational immediately.

So financing needs are *just* the capex for setting up the duplicated portion of the line. Given that Tesla is actually pretty cash-heavy at the moment, I suspect they've already placed their robot orders and are seeing up that duplicated portion ASAP. If they're getting 60-day terms on the robots too, it really should be easy to finance this... the only thing they have to pay up front is labor!

 

[neroden]

That is not how 90% plus of potential and actual customers will look at it.

The bottom line is the monthly payment. In most markets it will compete with those entry level luxury vehicles and particularly the PHEV versions.

Maybe, but I think people will pay extra for an electric car. (The PHEVs may have the same bonus.) At any rate, people who have test-driven them will pay extra for an electric car. I've been spitballing a $5000 premium on direct new-car purchase for the sheer niceness of electric traction, though I'm not sure how this would translate to a "monthly payment" (maybe $100 - $150?)

 

[22522]

I am assuming a linear (not exponential) increase beyond 4Q17. Also, no new line or massive incremental investment is needed to get to 10,000 per week. See below from 4Q16 call:

Elon Reeve Musk - Tesla, Inc.
But I would say it's going from 5,000 to 10,000 is probably – this is a total wild-ass guess, so (65:35) right way to think about, but it's like somewhere between 50% to 70% of the cost of the 5,000 line. Something like that.

If you're lucky and smart, 50% is only half the game, which obviously is pretty awesome from a CapEx standpoint. I can't imagine it being more than about 70% as much as there is. So, JB, what do you think?

Jeffrey B. Straubel - Tesla, Inc.

Yes, I think that's right. And it's maybe helpful to realize, but a lot of the infrastructure investments to get all the way to 10,000 are already completed, Gigafactory in particular.

Thanks, I missed this.

I would reuse the capital up to the paint facility and split into two lines after that to provide surface area to reduce in factory material handling by building right off the suppliers delivery truck. The stuff prior to paint should all be in house. The stuff after paint is from suppliers. Two lines double the access area for delivery.

They also make the process more robust to failure and easier to evolve. Two lines can leap frog each other on speed.

Seems like the cost of down time would bury the capital cost of a parallel line after paint...

 

[neroden]

Have you accounted for the lack of a complete assembly line robotic installation to date? It's very possible that we'll see the rapid ramp a few quarters later than presently imagined.

On the topic of robotic assembly lines, the interesting thing is that each individual step has to be automated individually: it takes a specific invention to automate each step. Some are harder than others. I was very impressed when they automated the process of bolting the battery pack to the car, since bolting was considered hard-to-automate. The reason the Model 3 is "alien dreadnaught 0.5" is that they haven't managed to automate all the steps yet... but they've automated more than they had automated in the S or X.

It would be fascinating to know which steps they've automated and which steps they're still working. But most of that is a trade secret. I am quite sure they haven't finished automating wiring -- they've talked about several different techniques to try to make the wiring more efficient / more automation-ready, which tells me it isn't ready yet. I don't know about the other final assembly steps.

 

[neroden]

They will sell cars outside the US as well.. haha. My point is that model 3s swarming everywhere will drive demand and the tax credit phase out will set the sense of urgency around the purchase. The question is if they can produce 2x as many, I dont think they will have any problem selling them, US, China the moon, wherever.

I think S/X are subject to a pretty hard production limit at the moment, as all the work in increasing production line capacity is going to Model 3.

That said, Body Line 2 is clearly still not operating at full intended capacity. So Tesla might be able to crank up S/X production some, sometime in 2018 (after Model 3 is running smoothly).

I expect a bias towards supplying all US production first until the US tax credit runs out completely; Tesla has done its best to give its customers the full advantage of expiring tax credits in other jurisdictions.

 

[neroden]

What happens if sales divert to international to push the tax credit crossover date to early 2Q18?

It's not possible.

Tesla's going to be at 180K - 200K cumulative by the end of 2017. We know Model 3 is going to California first. Any reasonable Model 3 production rate will blow through 200K by the end of February if not January. So the crossover date is in Q1 2018, pretty much for sure.

(P.S. I guess if they didn't deliver a single S or X in the US for the next six months, they might be able to push it out to Q2 2018, but that's not happening.)

 

[Snapdragon III]

I am assuming a linear (not exponential) increase beyond 4Q17. Also, no new line or massive incremental investment is needed to get to 10,000 per week (see below from 4Q16 call). I agree that, if they can pull forward Model 3 production to say "10,000 per week some time in 1Q18," a capital raise would be "worth it," but I don't think it will be necessary given what I'm assuming: 10,000 per week some time in 4Q18.

----------

Elon Reeve Musk - Tesla, Inc.

Yes, I mean, there's obviously going to be a fair bit of incremental investment to go from 5,000 cars a week to 10,000 cars a week, but it's going to be a lot less than getting to 5,000 cars a week in the first place. We don't know exactly what that's going to be except I'm confident it'll be less. Because the first thing we'll try to increase output is going back to rocket equation is to increase exit velocity of the line. And we don't know exactly where the trouble points (64:53) are going to be. We tried to model it out as carefully as possible, but there'll be things that aren't captured in the model.

But I think in a lot of cases, we'll simply be able to run the lines faster as opposed to duplicate the line. That's by far the best CapEx maneuver is just to make it go faster. But I would say it's going from 5,000 to 10,000 is probably – this is a total wild-ass guess, so (65:35) right way to think about, but it's like somewhere between 50% to 70% of the cost of the 5,000 line. Something like that.

If you're lucky and smart, 50% is only half the game, which obviously is pretty awesome from a CapEx standpoint. I can't imagine it being more than about 70% as much as there is. So, JB, what do you think?

Jeffrey B. Straubel - Tesla, Inc.

Yes, I think that's right. And it's maybe helpful to realize, but a lot of the infrastructure investments to get all the way to 10,000 are already completed, Gigafactory in particular.

That was the discussion that I was thinking of, but you and I obviously interpret it differently. I hope you are right. My take on that is that they know they need to duplicate a portion of the line to get from 5000-10,000. Maybe only the final assembly area? The part they are not sure about is how much of the line they need do duplicate. For instance, if they can crank the BIW part of the line up to 10,000 then they only need to branch to 2 lines afterwards, which would be more like the 50% number. If they can't get they BIW line going fast enough they may need to duplicate it too, but there would still be only one paint shop. That would be the 70% cost.

I think speeding things up will obviously help, but I believe there are portions of the current line that they know they can not run over 5000/week. If they were just going to speed up the whole thing, how could it possibly cost 70% of the initial cost? My main point is that I don't thing the increase in production will be linear in 2018. I think it will be an s curve in 2017 flattening out at around 5000/week, then another S curve later in 2018 flattening out around 10,000/week.

 

[Willuknight]

This is just me but I would not borrow money to buy stock. I am anti risk and want to keep my house paid off. Now with som cash I would buy Tesla I just would not borrow to do so.

I've been in since $180. Felt the same as you, until now.

 

[ValueAnalyst]

That was the discussion that I was thinking of, but you and I obviously interpret it differently. I hope you are right. My take on that is that they know they need to duplicate a portion of the line to get from 5000-10,000. Maybe only the final assembly area? The part they are not sure about is how much of the line they need do duplicate. For instance, if they can crank the BIW part of the line up to 10,000 then they only need to branch to 2 lines afterwards, which would be more like the 50% number. If they can't get they BIW line going fast enough they may need to duplicate it too, but there would still be only one paint shop. That would be the 70% cost.

I think speeding things up will obviously help, but I believe there are portions of the current line that they know they can not run over 5000/week. If they were just going to speed up the whole thing, how could it possibly cost 70% of the initial cost? My main point is that I don't thing the increase in production will be linear in 2018. I think it will be an s curve in 2017 flattening out at around 5000/week, then another S curve later in 2018 flattening out around 10,000/week.

Sure I agree with you on the two S-curves, but why assume the second S-curve will start only after the first S-curve ends?

If the exponential portion of the second S-curve overlaps with the logarithmic portion of the first S-curve, doesn't that add up to a linear increase in production throughout 2018?

 

[neroden]

That was the discussion that I was thinking of, but you and I obviously interpret it differently. I hope you are right. My take on that is that they know they need to duplicate a portion of the line to get from 5000-10,000. Maybe only the final assembly area? The part they are not sure about is how much of the line they need do duplicate. For instance, if they can crank the BIW part of the line up to 10,000 then they only need to branch to 2 lines afterwards, which would be more like the 50% number. If they can't get they BIW line going fast enough they may need to duplicate it too, but there would still be only one paint shop. That would be the 70% cost.

I think speeding things up will obviously help, but I believe there are portions of the current line that they know they can not run over 5000/week. If they were just going to speed up the whole thing, how could it possibly cost 70% of the initial cost? My main point is that I don't thing the increase in production will be linear in 2018. I think it will be an s curve in 2017 flattening out at around 5000/week, then another S curve later in 2018 flattening out around 10,000/week.

I think you're right. Which raises the question: how soon can they get this "second final assembly" (or maybe even only portions of the final assembly) up and running? I think we don't really have enough information to answer that question. But they've already got permits to expand the factory, and they have a lot of cash. They could be starting to set it up right now, or they might not start for three months. Who knows?

 

[Snapdragon III]

Hmm. You may be right.

Given Tesla's current capital structure (recent increase in ABL etc) I strongly suspect they can build those duplicated portions of the line without issuing more stock. Musk stated that the Model 3 ramp-up was "self-financing" due to the terms they're getting from suppliers: delivery now, pay 60 or 90 days later. This doesn't really kick in immediately; they're getting parts for 1000 cars a week in July and aren't producing that many cars in September; but come December it really does kick in, and for any new / duplicated section of line, this will be operational immediately.

So financing needs are *just* the capex for setting up the duplicated portion of the line. Given that Tesla is actually pretty cash-heavy at the moment, I suspect they've already placed their robot orders and are seeing up that duplicated portion ASAP. If they're getting 60-day terms on the robots too, it really should be easy to finance this... the only thing they have to pay up front is labor!

I like the way you are thinking, I suspect the problem is that there are some portions of the line they know they will have to duplicate to get to 10,000 per week. There are other portions that are borderline and they are hoping that they can just crank up the speed on the line they already have, and not duplicate it. The problem is if you order the robots now what do you order? If you just order the ones for the part you know will need to be duplicated, and you are unable to crank up the speed on the borderline section, you end up with a bunch of expensive stranded assets you can't use till you duplicate the borderline section. If you order the whole thing, then it turns out you can speed up the borderline section, you have wasted a bunch of money on robots you don't need. It is a hard problem.

It may turn out that they look at the extra cash flow that would come from getting to 10,000 sooner, and say f* it, lets order the whole thing now and who cares if we have a little wasted capacity, we will still come out ahead.

 

[erthquake]

If production goes as Musk tweeted last night, the Model 3 should easily outsell the Bolt this year with less than six months of production. It will likely outsell the Bolt's entire 2018 production before mid-February.