Elon Musk: "50% annual growth" + "500,000 in 2020". Which is most likely to occur?

[chickensevil]

The "few million cars by 2025" was only after the interviewer pressed hardly on where Elon thinks the company can get to. He had just gotten through talking about how it is technically possible that they cap out at 100k and can't go any higher than that. Then on bequest of the interviewer gave the other end of the spectrum (from the lowest possible to the highest). So he said off the cuff that he thinks he could see them doing a few million cars by 2025.

Everything in context... seriously guys... There is a difference between asking about the CEO's dream for the company and actually holding them to it. If you don't dream big then where does that leave you? Small, tiny... safe... iterations each year that does nothing to cause the paradigm shift in transportation that he is hoping to accomplish. GM, Nissan, BMW, VW, Mercedes... that is what you get when you come at this from the "small, slow, safe" iterations... I don't want that dream... And If you didn't buy into that dream then by all means short the stock... bet against the dream. And of course if valuation jumps tomorrow to assume that the dream is going to happen exactly as stated, then don't get upset when a bump in that road causes the price to drop.

To that end, if they strive VERY HARD to reach 3M by 2025 and fall short to *just* 2M... wouldn't you think that 2M was still a fantastic achievement?

 

[Benz]

Elon Musk has made two projections about future growth: "50% annual growth" + "500,000 in 2020". Which is most likely to occur?

Is it possible that both projections can be correct at the same time? Or can both projections not be correct at the same time?
How do the numbers work out?

If we assume that there were (at least) 32,000 Tesla Model S deliveries in 2014, then "50% annual growth" would imply:
2013 - 22,477 Tesla Model S deliveries
2014 - 32,000 Tesla Model S deliveries
2015 - 48,000 Tesla Model S deliveries
2016 - 72,000 Tesla Model S deliveries
2017 - 108,000 Tesla Model S deliveries
2018 - 162,000 Tesla Model S deliveries
2019 - 241,000 Tesla Model S deliveries
2020 - 361,500 Tesla Model S deliveries

Then that would imply that a very large portion of the 500,000 cars in 2020 would be already occupied with the Tesla Model S only.

Then we haven't even included the delivery numbers of the Tesla Model X and the Tesla Model 3 yet.

What does this mean:
1 - The Tesla Model S delivery numbers just cannot keep on growing at 50% year after year.
2 - Or the total number of cars in 2020 is much higher.

Or maybe a bit of both?
And that is most likely, I think.

So, what is a more likely scenario then?

Tesla Model S deliveries:
2015 - 50,000
2016 - 60,000
2017 - 70,000
2018 - 80,000
2019 - 90,000
2020 - 100,000

Or would that be a too conservative projection of growth?

What projection of growth would be more realistic according to you?

I think that the key is production capacity of battery packs and cars.

I actually think that Tesla Motors will announce in 2016 that they will open two new factories to produce their cars. One in Europe and one in China. In these two factories Tesla Motors will initially produce only the Tesla Model 3. That would mean that there will be additional production capacity available.

That will change the game. Then this 500,000 number for 2020 will not be considered to be a ceiling (maximum), but instead it will be considered as a floor (minimum).

 

[Johan]

Elon Musk: "50% annual growth" + "500,000 in 2020". Which is most likely to o...

First: I'm sure that when Elon is talking production estimates for 2015 and 2016 he means S+X combined (regardless of what the analyst said on the call). When he talks numbers longer term he's talking about all models combined. Remember that unit-for-unit the Model 3 will be cheaper and easier to build, but also have lower ASP and not unlikely lower margins. That would kind of make it easier to maintain 50% yearly growth even as the absolute numbers become larger.

The only real factors I see stopping Tesla's growth up until 2020 or even 2025 are:
Themselves messing something up badly (for example committing to a battery chemistry or technology in the GF that somehow becomes outdated over night) or something safety related

or

Beginning to touch on total global demand constraints for cars in any given class.

 

[tftf]

The "few million cars by 2025" was only after the interviewer pressed hardly on where Elon thinks the company can get to. He had just gotten through talking about how it is technically possible that they cap out at 100k and can't go any higher than that. Then on bequest of the interviewer gave the other end of the spectrum (from the lowest possible to the highest). So he said off the cuff that he thinks he could see them doing a few million cars by 2025.

Pressed hardly? You can view the entire video interview with Musk here:

http://bloom.bg/1zbulPP




Minute 09:45 and following for the exact quote about 2025 sales.

Musk repeats the goal twice. Even the interviewer is stunned and is asking Musk if he didn't mean 2030 (counting 2020 and 500k cars as a starting point instead of 2015), yet Musk reiterates once again 2025.

The interview is online and so is the old 2009 interview, a podcast where Musk announced the Model 3 by late 2014 / early2015 for around $20-30k, he even hinted that the Model 3 may arrive sooner than that thanks to a "secret project". Earlier (2008, GigaOm interview) he even said the Model 3 might sell for as low as $20k.

Over five years passed since that late 2009 interview and Musk is still making the same kind of sky-high promises into the future.

Anyway, everybody can form their own opinion based on a number of sales projections and estimates.

 

[chickensevil]

Pressed hardly? You can view the entire video interview with Musk here:

http://bloom.bg/1zbulPP




Minute 09:45 and following for the exact quote about 2025 sales.

Musk repeats the goal twice. Even the interviewer is stunned and is asking Musk if he didn't mean 2030 (counting 2020 and 500k cars as a starting point instead of 2015), yet Musk reiterates once again 2025.

The interview is online and so is the old 2009 interview, a podcast where Musk announced the Model 3 by late 2014 / early2015 for around $20-30k, he even hinted that the Model 3 may arrive sooner than that thanks to a "secret project". Earlier (2008, GigaOm interview) he even said the Model 3 might sell for as low as $20k.

Over five years passed since that late 2009 interview and Musk is still making the same kind of sky-high promises into the future.

Anyway, everybody can form their own opinion based on a number of sales projections and estimates.

Isn't that pretty much exactly what I just said and was getting at? I don't know why you make it like you are fighting me on us being on the same page here.

I just don't view the sky high promise as a negative since I would rather the guy dream big. Of course if people start to assume that number is going to happen and the price goes way up to match that and then nothing becomes of the number within that timeframe the stock will bubble up and burst of dissappointment.

I view it as an attainable number, even if it is challenging to reach... But will not be surprised if they fall short. How goes the saying? "Shoot for the moon and if you miss you will end up among the stars" The reason I got into the company was for the vision and lofty goals... At least someone is trying to revolutionize the car industry.

- - - Updated - - -

Bottom line, look at the 2025 number as a lofty goal for the company... But don't put millions of dollars of faith in that goal until we seem the steps that would be required to get there start to take shape. (That is, announcements and plans to build a bunch more factories).

 

[Benz]

Pressed hardly? You can view the entire video interview with Musk here:

http://bloom.bg/1zbulPP




Minute 09:45 and following for the exact quote about 2025 sales.

Musk repeats the goal twice. Even the interviewer is stunned and is asking Musk if he didn't mean 2030 (counting 2020 and 500k cars as a starting point instead of 2015), yet Musk reiterates once again 2025.

The interview is online and so is the old 2009 interview, a podcast where Musk announced the Model 3 by late 2014 / early2015 for around $20-30k, he even hinted that the Model 3 may arrive sooner than that thanks to a "secret project". Earlier (2008, GigaOm interview) he even said the Model 3 might sell for as low as $20k.

Over five years passed since that late 2009 interview and Musk is still making the same kind of sky-high promises into the future.

Anyway, everybody can form their own opinion based on a number of sales projections and estimates.

Elon Musk has a vision and he also has plans to accomplish that vision. The point is that his plans evolve/improve over time. They get better and more realistic and more accurate. The circumstances and the means to execute his plans have improved very much as well.

 

[Benz]

I think that the key is production capacity of battery packs and cars.

I actually think that Tesla Motors will announce in 2016 that they will open two new factories to produce their cars. One in Europe and one in China. In these two factories Tesla Motors will initially produce only the Tesla Model 3. That would mean that there will be additional production capacity available.

That will change the game. Then this 500,000 number for 2020 will not be considered to be a ceiling (maximum), but instead it will be considered as a floor (minimum).

The economic forcing function that will lead to the expansion of the production capacity (of both the cars and the battery packs) of Tesla Motors (simply meaning more factories), is the rising rate of global reservations (including the deposits) that then are changed into orders of both the Tesla Model S and the Tesla Model X. Simply because all cars that have been ordered for, simply have to be built and delivered in a certain and reasonable timeframe. When they will have to produce more than their maximum capacity allows them, then they will simply have to expand their production capacity. Of course they will be able to control when this will happen (up to a certain level) by the levers they can pull (when and if they would want to do that).

 

[Benz]

In this video JB Straubel is talking about energy storage and at 12:28 minutes in the video he says: "Our target is to build half a million Gen 3 vehicles per year."

Does he mean in 2020 already?

Link: https://www.youtube.com/watch?v=zWSox7mLbyE

 

[SteveG3]

...

The interview is online and so is the old 2009 interview, a podcast where Musk announced the Model 3 by late 2014 / early2015 for around $20-30k, he even hinted that the Model 3 may arrive sooner than that thanks to a "secret project". Earlier (2008, GigaOm interview) he even said the Model 3 might sell for as low as $20k.

Over five years passed since that late 2009 interview and Musk is still making the same kind of sky-high promises into the future.

Anyway, everybody can form their own opinion based on a number of sales projections and estimates.

We know in 2012, early 2013, Musk's aspiration was reaching volumes of 20K Model S and 15K Model X. By the spring of 2013, he saw that demand was MUCH higher. Musk now has fairly consistently expressed volume targets of over 100K for the Model S and Model X with increased Fremont capacity in 2016. As things looked in 2009 (until spring of 2013 for that matter) 35K Model S/X sales in 2015, it seemed it would help to have a Model 3 developed and on sale in 2015. At 60K S/X sales in 2015 heading to 100K+ in 2016 it's pointless to introduce a Model 3 which you wont even have batteries available to sell volumes of it higher than your S/X volumes. This is especially so as Model S/X ASP is more than double what Model 3 is likely to be AND have better margins than Model 3.

Pretty sure he's said Panasonic non-GF battery supply could get up to volumes capable of producing 150K vehicles. At some point in 2013 Tesla could see S/X approaching those levels on their own, no battery manufacturers independently making a big move to increase supply and correctly realized the company was better off first improving supply issue (GF) and then releasing Model 3.

TFTF there was another concern you'd expressed about Tesla's prospects I saw earlier this week that I also found deeper examination revealed was nothing to be concerned about in my view. I wrote a response to your concerns there a few days ago, and have not seen a reply from you as to whether you agree with what I've written or are looking at it differently. Link to thread below if you care to reply to what I discussed there.

No Model III until 2019? - Page 5

 

[Benz]

Total annual Tesla Model S deliveries in 2013: 22,477
Total annual Tesla Model S deliveries in 2014: 31,655

That's a 40.83% increase.

Maybe 50% is not realistic.

What would 40% annual growth lead to?

Tesla Model S deliveries based on 40% annual growth:
2015 - 44,317
2016 - 62,044
2017 - 86,862
2018 - 121,607
2019 - 170,250
2020 - 238,350

Probably it's more likely that the % annual growth will decline as the years go by and the annual delivery numbers keep increasing.

 

[adiggs]

Total annual Tesla Model S deliveries in 2013: 22,477
Total annual Tesla Model S deliveries in 2014: 31,655

That's a 40.83% increase.

Maybe 50% is not realistic.

What would 40% annual growth lead to?

Tesla Model S deliveries based on 40% annual growth:
2015 - 44,317
2016 - 62,044
2017 - 86,862
2018 - 121,607
2019 - 170,250
2020 - 238,350

Probably it's more likely that the % annual growth will decline as the years go by and the annual delivery numbers keep increasing.

I prefer to look at production rather than deliveries. The reason is that deliveries are subject to variation that production is not subject to. For instance, needing to pay for expedited car shipping to get a car across country (or ocean) due to low availability of transport. That delivery gets delayed, but the parts streaming into the factory are (I presume) less subject to weather and oddball transportation issues.

In the direct example you're using above, we know that there were a number of incremental cars produced and not delivered in Q4. By using Q4 delivery numbers, you're baking into your model that shortfall in this Q4, and expanding it by 40% every year from here on out. Then again, I haven't plotted or analyzed the data to determine whether production or deliveries represent the more stable and projectable process. I just presume that production is the better process.

 

[Benz]

I prefer to look at production rather than deliveries. The reason is that deliveries are subject to variation that production is not subject to. For instance, needing to pay for expedited car shipping to get a car across country (or ocean) due to low availability of transport. That delivery gets delayed, but the parts streaming into the factory are (I presume) less subject to weather and oddball transportation issues.

In the direct example you're using above, we know that there were a number of incremental cars produced and not delivered in Q4. By using Q4 delivery numbers, you're baking into your model that shortfall in this Q4, and expanding it by 40% every year from here on out. Then again, I haven't plotted or analyzed the data to determine whether production or deliveries represent the more stable and projectable process. I just presume that production is the better process.

The price of the car is paid when the car is delivered. That's when the transaction takes place. That's when revenue is realised. And that's what it's all about, I think.

Manufacturing cars is not enough, the delivery of the cars is important as well.

 

[Robert.Boston]

The price of the car is paid when the car is delivered. That's when the transaction takes place. That's when revenue is realised. And that's what it's all about, I think.

Manufacturing cars is not enough, the delivery of the cars is important as well.

If and when Tesla becomes "delivery constrained," I'd agree. But Elon says they are "production constrained." Thus tracking production tells us how the rate-limiting system is performing. Everything else is just about timing.

 

[Mario Kadastik]

Well another angle to take in is that Elon on the Q4 CC mentioned revenue growth YoY of ca 50% for next 10 years and that this would effectively bring 10y from now TSLA to be worth about what Apple is now ($700B). So maybe that clears up a bit the ambiguity of S vs X vs 3 mixture. If the revenues go up then addition of Mod 3 will have to be ~2x that of Mod S/X growth to be about the same equivalent ASP. So for every two-three Mod 3 it's equivalent to one more Mod S instead. And Mod 3 is meant to be volume production so instead of 50k Mod S it'd be 100-150k Mod 3 that gives equivalent revenue.

 

[Ocelot]

hopefully by 2025 they will have 10 or more models to choose from. once the model 3 hits, i can see a model a year to year and a half being revealed, all based on the three platform.

 

[Benz]

If the total number of delivered Tesla vehicles in 2020 indeed will be 500,000, then about how many of that total number will be Tesla Models Ses (in 2020)?

 

[kenliles]

If and when Tesla becomes "delivery constrained," I'd agree. But Elon says they are "production constrained." Thus tracking production tells us how the rate-limiting system is performing. Everything else is just about timing.

that's a good point; but I believe Tesla is organizing a real-time order and delivery system that always ensures production is driven by customer orders directly rather than an inventoried dealership model. From a realization of income then, it would always be on delivery and modeled as production contained (albeit constrained via it's own forcing function of active customer orders). In other words, it will always take 30 days (time of production and delivery) to get a Tesla car (excluding the lease and used rotation inventories of course)

 

[Robert.Boston]

that's a good point; but I believe Tesla is organizing a real-time order and delivery system that always ensures production is driven by customer orders directly rather than an inventoried dealership model. From a realization of income then, it would always be on delivery and modeled as production contained (albeit constrained via it's own forcing function of active customer orders). In other words, it will always take 30 days (time of production and delivery) to get a Tesla car (excluding the lease and used rotation inventories of course)

That's not what I would call production constrained. Production constrained to me is when the factory is cranking out as many cars as it can per shift. I've never read or heard anything to make me believe that Tesla's lines were ever slowed because lack of backlog.

 

[kenliles]

That's not what I would call production constrained. Production constrained to me is when the factory is cranking out as many cars as it can per shift. I've never read or heard anything to make me believe that Tesla's lines were ever slowed because lack of backlog.

agreed- I was addressing more the discussion around delivery vs production tracking of realized revenue etc. I think deliveries to customers will always be Tesla's production - 'built-delivered-to-order'; I think they see delivery as part of production in terms of revenue etc. So going forward production-deliveries will be forced to match orders; so I guess what I'm projecting is they will never be delivery constrained. probably just nuance I guess- I thought that's what the form of the discussion was though- maybe I misread it

 

[30seconds]

agreed- I was addressing more the discussion around delivery vs production tracking of realized revenue etc. I think deliveries to customers will always be Tesla's production - 'built-delivered-to-order'; I think they see delivery as part of production in terms of revenue etc. So going forward production-deliveries will be forced to match orders; so I guess what I'm projecting is they will never be delivery constrained. probably just nuance I guess- I thought that's what the form of the discussion was though- maybe I misread it

I think Elon's messages are a bit mixed when it comes to the future of build to order. To me it seems that this will always be an option for Tesla customers, but to drive significant Gen3 volume Tesla is leaving open the option for dealerships. One way to drive volume is to produce batches of Gen3 with common configuration options and then sell them through national or dealer level inventory. Kind of like "inventory" cars today, but larger scale. If they push this inventory to the dealer level, then this sale could be at the time of the transfer from Tesla to the dealer (like other manufacturers).