Demand

[AlMc]

Let's agree to check back here during the rest of 2016.

Agreed.

First, Thanks to all for moving the discussion over here.

Second, we do not have enough facts to make any position the 'right' one. I tend to agree with 'VG' because of all the research he has done in compiling numbers/wait times and having a wealth of EM quotes at his fingertips to support that demand is not an issue.

We will have more data points soon about Q2 deliveries and an ER/CC where if guidance is to be changed it most likely will be.

 

[JRP3]

While the price of the car has been rising Tesla has also improved the car significantly making it more expensive to produce. Clearly a bigger battery option will be more expensive, likewise in production.

How so? The larger battery is only possible through improved specific energy, i.e. each cells contains more energy. That likely means the cost per kWh goes down, not up, since the same volume/weight of material holds more energy. The end result is that Tesla can charge more for the same volume/weight of materials, which will increase profit margins.

 

[Perfectlogic]

How so? The larger battery is only possible through improved specific energy, i.e. each cells contains more energy. That likely means the cost per kWh goes down, not up, since the same volume/weight of material holds more energy. The end result is that Tesla can charge more for the same volume/weight of materials, which will increase profit margins.

Larger batteries isn't the only improvement Tesla has been making. What is the most relevant is to just look at the gross margin which hasn't gone up much in spite of scale, production efficiencies and battery costs improving. When accounting for cost not part of COGS but still relevant for the S sales I would argue that they are still only breaking even, which would suggest that Tesla might not have much more up their sleeve.

 

[electracity]

Outside of the NYC area the rest of NY State doesn't have a single store. There are large numbers of high net worth households untouched by Tesla in NY State alone.

Of course Tesla can build more showrooms and get more sales. But the cost of customer acquisition rises as the density of of high net worth households decreases. The economics of showrooms in second tier markets improves greatly with the model 3.

Mercedes and BMW sell mostly less expensive cars. There would be far fewer Merc. and BMW dealerships if the didn't make made most of their unit sales over $55K.

 

[Yggdrasill]

Larger batteries isn't the only improvement Tesla has been making. What is the most relevant is to just look at the gross margin which hasn't gone up much in spite of scale, production efficiencies and battery costs improving. When accounting for cost not part of COGS but still relevant for the S sales I would argue that they are still only breaking even, which would suggest that Tesla might not have much more up their sleeve.

Percentage-wise, gross margin hasn't increased much. But the cars are getting more expensive, and 25% of a more expensive car means more money in. For the past few months you've also had the X pulling down the gross margin. The gross margin for the X won't improve until they can produce them in large numbers with few quality issues. That should happen by the end of this quarter. This is too late for the results in Q2, but Q3 should be quite good.

 

[Perfectlogic]

Percentage-wise, gross margin hasn't increased much. But the cars are getting more expensive, and 25% of a more expensive car means more money in. For the past few months you've also had the X pulling down the gross margin. The gross margin for the X won't improve until they can produce them in large numbers with few quality issues. That should happen by the end of this quarter. This is too late for the results in Q2, but Q3 should be quite good.

I was talking about the S gross margin which still seems to be around 25%, I think 30% have been the target for a while that keeps getting pushed out further in the future.

 

[vgrinshpun]

Indeed you have to account for the weight of each region to the the most precise, but the difference in wait fluctuations isn't nearly big enough for this factor to skew the whole picture. Compared to previous quarters China has gone up maybe 2 weeks on average in wait while the US has gone down about the same, 2 weeks on average.

So unless China for Q2 was a larger market than the US, the (small) influence from this principle is actually the opposite of what you are trying to prove. And the last time I checked, China sales are very small compared to the US.

As I mentioned before, I do not believe that Schonelucht's graphs are accurate. I actually reviewed all the data that I accumulated for NA wait times and result, as I suspected, is actually the opposite of the one's in Schonelucht's graphs.

The trend line for the NA wait time is UP, not down. You can see details here.

Given that deliveries were increasing over the same period, your conclusion about the demand are just plain wrong.

 

[EinSV]

I think it is worth making an observation about long-term demand for Tesla vehicles that is pretty obvious but I don't see expressed very often.

One of the biggest long-term bear arguments against Tesla is that EVs are inherently a niche market, and that the total EV market size is limited. I believe this bear thesis was conclusively disproven by the success of the Model S, but Tesla skeptics still argued (contrary to the data) that the Model S is a plaything for the rich, and that "regular" people who need their car to go to work and for vacations would never buy EVs in large numbers due to range anxiety, etc.

The incredible response to the Model 3 demolishes this argument forever. 373,000 people have now placed a reservation, without ever test driving the car and enduring a 1.5-2.5 year wait, despite Tesla holding back many details about the Model 3 that will bring in many more customers.

Based on the level of reservations, there is no question in my mind that there is sufficient demand for the Model 3 to quickly outsell the current leader in its class -- the BMW 3 series. BMW 3 Series - Wikipedia, the free encyclopedia -- in the same way the Model S now outsells all other large cars in the premium market.

But the larger point is that the incredible reception to the Model 3 conclusively disproves the central bear thesis that the overall market for EVs is limited. Therefore, the main hurdles to Tesla blowing through the 1,000,000 target in 2020 and reaching 6,000,000-10,000,000 vehicle sales by 2025 are execution, macro risks, the ability to raise modest amounts of capital, continuing to reduce costs so lower priced cars can earn a reasonable margin and competition if and when it wakes up.

 

[Oil4AsphaultOnly]

I haven't said Tesla isn't at production capacity right now. It is all about the ramp from 1300 to 2000/week end of Q2. The wording Elon has used about the second half of the year is that the plan is to sell 50k, and of course it is.

Hmmm, how is this not to be interpreted as condescending?

 

[Perfectlogic]

As I mentioned before, I do not believe that Schonelucht's graphs are accurate. I actually reviewed all the data that I accumulated for NA wait times and result, as I suspected, is actually the opposite of the one's in Schonelucht's graphs.

The trend line for the NA wait time is UP, not down. You can see details here.

Given that deliveries were increasing over the same period, your conclusion about the demand are just plain wrong.

Might as well set the record straight in this thread too. My theory isn't based on waiting times alone, but rather a function of production and fluctuations in waiting times. Even if your graph is more precise than Schonelucht's the scale of the difference is tiny which means it changes nothing.

 

[electracity]

Might as well set the record straight in this thread too. My theory isn't based on waiting times alone, but rather a function of production and fluctuations in waiting times. Even if your graph is more precise than Schonelucht's the scale of the difference is tiny which means it changes nothing.

Why do historical trends matter when Tesla has doubled its products for sale? No one knows what happens after the initial model X orders are filled. I'm interested in numbers after Europe MX orders ship.

Not to mention that the effects on near term demand due the interplay between S/X and the announced model 3 is unknown. This is a new Tesla, with three models in play affecting consumer demand.

 

[vgrinshpun]

Might as well set the record straight in this thread too. My theory isn't based on waiting times alone, but rather a function of production and fluctuations in waiting times. Even if your graph is more precise than Schonelucht's the scale of the difference is tiny which means it changes nothing.

The difference is that you seemingly based your conclusions on assumption that demand (rate of incoming orders) was reducing with the increase in production and deliveries.

The problem, however, is that the rate of incoming orders actually increased faster than the increase in production and deliveries.

So now by claiming that this does not affect your conclusions you are effectively saying that your conclusions are independent of the data, which is exactly what I feel is happening. Your conclusions do not have any link with reality.

 

[Perfectlogic]

The difference is that you seemingly based your conclusions on assumption that demand (rate of incoming orders) was reducing with the increase in production and deliveries.

The problem, however, is that the rate of incoming orders actually increased faster than the increase in production and deliveries.

So now by claiming that this does not affect your conclusions you are effectively saying that your conclusions are independent of the data, which is exactly what I feel is happening. Your conclusions do not have any link with reality.

Are you not listening to what I'm saying? I already explained everything very detailed in the other thread. Production has increased around 20% compared to 1 year ago, waiting times have increased 1 week. If the waiting time had been completely steady like I have been saying all along the demand increase would be the 20% from increased production, if the wait increased 1 week as you claim (still not convinced as you haven't included the latest data point with a short wait) then that adds 2%. Increasing the total from 20% to 22%, that is clearly not a large difference and doesn't change the fact that the demand growth is slowing for the S as it grew 50% the year before.

 

[lango]

Are you not listening to what I'm saying? I already explained everything very detailed in the other thread. Production has increased around 20% compared to 1 year ago, waiting times have increased 1 week. If the waiting time had been completely steady like I have been saying all along the demand increase would be the 20% from increased production, if the wait increased 1 week as you claim (still not convinced as you haven't included the latest data point with a short wait) then that adds 2%. Increasing the total from 20% to 22%, that is clearly not a large difference and doesn't change the fact that the demand growth is slowing for the S as it grew 50% the year before.

You keep failing to be clear what your message is. Is it now that growth is slowing from 50% for S? I don't think anyone would have an issue with that statement, it is also guided for by Tesla. The yoy growth for S from 2015 to 2016 is expected to be quite low, partly on purpose to free up production for X.

Then you keep saying X demand will be low and the backlog will not keep up with production so production would have days with no input, but you have no reliable data for that whatsoever. With no good data the logical thing would be to go by guidance and maybe 95% of it because historically they have fallen short, but you somehow keep trusting your faulty methods more despite all flaws people point out with those methods. The X is not a normal car purchase until very recently. People are savvy and don't want to buy a beta car. You are also wrong that X is failure, even the ramp is not. What has happened is just an unrealistic time schedule, the actual time it took them to get here is pretty fast. To give an analogy, if Usain Bolt kept promising under 9 s but run at 9.6 s, that is not a failure. Same thing with the X ramp... it seems they have solved most issues with it now.

If queue is increasing for the average customer or steady while production is increasing then demand keeps up with production? Very simple.

What do you think the growth will be the coming years for S and X?

 

[Oil4AsphaultOnly]

Are you not listening to what I'm saying? I already explained everything very detailed in the other thread. Production has increased around 20% compared to 1 year ago, waiting times have increased 1 week. If the waiting time had been completely steady like I have been saying all along the demand increase would be the 20% from increased production, if the wait increased 1 week as you claim (still not convinced as you haven't included the latest data point with a short wait) then that adds 2%. Increasing the total from 20% to 22%, that is clearly not a large difference and doesn't change the fact that the demand growth is slowing for the S as it grew 50% the year before.

That's bad logic (and math). Let me illustrate:

1 year ago: production was at 1000/week, and 4 weeks of wait time = 4000 unit backlog
now: production is at 1200/week (20% higher), and 5 weeks of wait time = 6000 unit backlog (5 weeks x new production rate)
now (assuming 0 increase in wait time): 4800 unit backlog (4 weeks x 1200/week)

I'm impressed that vgrinshpun has been so patient with you for so long.

Edit: In case you need it spelled out, even without an increase in wait times, there's still a 20% increase of orders. In the 20% increase in wait time scenario, there's actually a 50% increase in orders.

 

[Perfectlogic]

You keep failing to be clear what your message is. Is it now that growth is slowing from 50% for S? I don't think anyone would have an issue with that statement, it is also guided for by Tesla. The yoy growth for S from 2015 to 2016 is expected to be quite low, partly on purpose to free up production for X.

My message has been perfectly clear the entire time, if you don't understand it you should read my posts again. Clearly the reason for the discussion about if demand = production + fluctuations in waiting times have gone for so long is because people don't agree with it.

The problem for most people probably is that the demand curve predicted by that equation makes it unlikely that Tesla will be able to fill the 2k production capacity. Unless ofcourse you believe the X demand to suddenly jump to 800/week from what is suggested by the simple logic of the equation to have been much, much lower so far.

What do you think the growth will be the coming years for S and X?

I would guess around 55k Model S this year and <10% growth for the S over the coming years, but still positive. For the X I expect 15-20k tops for this year and perhaps 10% growth over the next few years, it is much harder to say than for the S as we have less data.

 

[techmaven]

I would like to point out something about U.S. and European delivery times. Back in the spring of 2015, Tesla started to use rail for transporting vehicles from the west coast to the east coast. That includes European deliveries. Early within a quarter, the use of rail is high. As the quarter reaches near the end, the use of trucking is high. Therefore, there is a ~2-3 week drop in delivery times just by switching from rail to trucking. The switch between rail and trucking completely messes up any delivery time estimates as useful data, as they switch between them with a lot of internal data, but it is difficult for us on the outside to see enough of the variables to hazard decent guesses as to when, how much and ultimately why. Their switching also depends on their costs, the availability of rail and trucking, and their desire to hit a particular number of deliveries.

 

[Perfectlogic]

That's bad logic (and math). Let me illustrate:

1 year ago: production was at 1000/week, and 4 weeks of wait time = 4000 unit backlog
now: production is at 1200/week (20% higher), and 5 weeks of wait time = 6000 unit backlog (5 weeks x new production rate)
now (assuming 0 increase in wait time): 4800 unit backlog (4 weeks x 1200/week)

I'm impressed that vgrinshpun has been so patient with you for so long.

Edit: In case you need it spelled out, even without an increase in wait times, there's still a 20% increase of orders. In the 20% increase in wait time scenario, there's actually a 50% increase in orders.

You are correct, that is the first post you have made that have actually made sense, congratulations. The problem is that we are still only talking about a few percent, we are not arguing about details, but rather if the principles I'm using to calculate demand makes sense or not. If we are getting into details we should also be dealing with the waiting times seperately for each market as they are different in size (1 week wait is a different number of cars in the backlog).

If we do that we will find that the biggest market, the US, has actually seen shortening wait times while the China wait is pulling the average up, and China is the smallest market. So accounting for all the details and using the latest data points the total backlog probably hasn't move up at all.

 

[lango]

I would guess around 55k Model S this year and <10% growth for the S over the coming years, but still positive. For the X I expect 15-20k tops for this year and perhaps 10% growth over the next few years, it is much harder to say than for the S as we have less data.

55K S for this year is in line with my estimate for the year. The production might be higher if X production lags and they feel the need to clear the channels again in Q4 like they did last year. As for X, they still have a sizable backlog and production is increasing. It is absolutely possible though that the rate of new orders will be too low because they don't have demo cars and the launch has been tainted.

Personally, I think they will be able to grow S demand with 15% yoy until it reaches 100k a year end of 2020. X would also be able to reach 100k a year at that time.

This prediction is based on the 11 points I wrote about in the short term thread and also the fact that growth for a product often continues rather than stall, so the default is to assume it continues unless you got evidence it won't (which you don't have). X will follow S because of the SUV market being bigger.

Your method and short term wait time data is NOT proving this to be false at all for reasons stated by many over and over.

Now S and X might stall for one other reason, that they focus and invest on Model 3 only, and not build out capacity for them or drive demand in any meaningful way.

No models for Tesla assumes 50% yoy continued growth for X and S, many actually do assume a steady state of 100k combined for them which seems to be way too low.

Btw, there is a big, big difference between 9% and 1% for a long term growth investment.

 

[vgrinshpun]

Are you not listening to what I'm saying? I already explained everything very detailed in the other thread. Production has increased around 20% compared to 1 year ago, waiting times have increased 1 week. If the waiting time had been completely steady like I have been saying all along the demand increase would be the 20% from increased production, if the wait increased 1 week as you claim (still not convinced as you haven't included the latest data point with a short wait) then that adds 2%. Increasing the total from 20% to 22%, that is clearly not a large difference and doesn't change the fact that the demand growth is slowing for the S as it grew 50% the year before.

Of course I do. This demonstrates nothing.

Let me repeat my point one more time: the Company over the past year increased their deliveries over the past year 50%, while managing to keep rate of incoming orders just ABOVE the rate of production, consistent with what they kept saying to people like you that insisted that demand is not there.

The company again is saying that the rate of incoming orders is up 45%, enough to meet the guidance.

Your response: the rate of incoming orders had barely kept up with the increasing production. The thing is that you refuse to acknowledge is that the rate incoming orders was increasing FASTER than production and deliveries. What is your reason for saying it is NOT continue to happen, when demand is up, and when the company explicitly said that it IS happening? Your desire to get back in at SP=$120? Forget it, it is not going to happen, your heroic efforts notwithstanding.