Q3 2013 results - projections and expectations

[JRP3]

Since the average purchase price of a vehicle in the US was $30K last year we don't need a $20K car to make a difference. A $35K G3, which is more like a $25K car when you consider total operating costs, will make a huge difference. Other OEM's can tackle the $20K and lower market, the Chevy SparkEV is already $20K with incentives and over 80 miles of range. In fact used G3's should beat your 120mile $20K by 2020 target when they start hitting the market it 2020 with around 200 miles of range. Fact is people struggling financially should not be buying new vehicles of any sort. That's what the used market is for.
As for battery constraints today, we have seen no evidence of such and it conflicts with Elon's statements about constraints with volumes over a couple hundred K per year.

 

[vgrinshpun]

VINs are not produced in sequence.

According to my analysis VINs are produced in sequence, but are not released into production in sequence due to the fact that European cars were assigned VINs in large blocks, but are released into production gradually, to blend in with the production of cars for NA deliveries. There is also a certain disturbance to the sequence of VINs that are coming through the production due to batching as I explained earlier (2Q 2013 Model S Deliveries Potential Surprise - Page 7).

I used this explanation to correctly predict the cars delivered in Q2. My prediction was 5200 vs. 5150 delivered actually, so I am quite confident that this theory was validated. (Short-Term TSLA Price Movements - Page 341)

Regarding the VIN 21,591 reported by CygnusX-1 that was scheduled for the 9/28 delivery, it indicates that deliveries for Q3 will be approximately in the 5,400 range. Bear with the explanation below to see how this prediction was made.

According to the interview Elon did in Europe before the Geneva Auto Show this year Tesla planned to deliver about 5,000, "may be" 7,000 cars. Just recently there were reports on this Forum that European deliveries for this year are already booked. Assuming that Tesla allocated 5,000 spots for this year's delivery to Europe, there were 5,000 VINs assigned to cars for European Delivery. I have used lower 5,000 number because I think that while according the Elon NA orders are higher than expected, the pace of European orders did not pick up yet, so it is logical to assume that greater proportion of cars would be allocated for NA deliveries.

Assume that 5,000 European delivery cars will be allocated evenly: 2,500 for Q3 and 2,500 for Q4. Out of the 2,500 cars produced for European Delivery in Q3 about 1,500 would be delivered in Q3, while remaining 1,000 will be delivered in Q4 due to shipping/assembly delay (assumed about 5 weeks). Based on the above out of 5,000 VINS that were assigned to the European delivery cars only 1500 cars will be delivered in Q3. So the total quantity of VINs that are already assigned but are not going to be delivered in Q3 is 5,000-1,500=3,500.

The total quantity of cars delivered in Q4 2012 and Q1, Q2 2013 was 12,700. Calculating approximate number of predicted deliveries in Q3: 21,591-12,700-3,500=5,391.

Regarding the accuracy of the above prediction, the number could be different due to the following factors:

1. Tesla could have arranged production in such a way that the majority of European Delivery cars were produced in the first half of the Q3, so more than 1,500 of them are delivered in Q3

2. The production is about 600 cars weekly, so about 120 cars are produced daily. Therefore about 85 VINs would be delivered on a given day (600/7). So cars scheduled for delivery on 9/28 could have VINs 21,591+/-85.

3. According to my original theory linked below the most common configuration is S85 with air suspension, and VINs for these cars represent projected deliveries with the best accuracy. The post by CygnusX-1, however, does not mention the air suspension, so VIN 21,591 was assigned to S85 with coil suspension.

All in all I feel that 5,400 number arrived at above has a good chance to be higher by one to several hundreds.

 

[cfOH]

To my eye, VINs aren't assigned, built, or delivered in strict sequence. However, they are approximately sequential in that VIN 20,000 is very likely going to be assigned before VIN 21,000.

That's why I use a regression-line approach to find the approximate VIN assignment rate, realizing that a range of VINs will be handed out on any given day. If you consider each day as analogous to a sample with a distribution, then ordinary least-squares regression will find a reasonable approximation of the line that splits the ranges over any given period of time.

Also, since it seems that Tesla is generally increasing their output rate, you can't extrapolate a production rate over too long a period...it's not linear. I think about 2 months is as long as you'd want to include -- it sufficiently damps down the variability while still adequately estimating the current rate.

Finally, not all VINs assigned go to customers -- some are assigned to loaners, test mules, demo units, etc. -- and not all VINs that are assigned to customers get actually delivered (some deals fall through). So, there will be some discrepancy between VIN assignment rates and actual deliveries, and the latter is what Musk reports quarterly.

 

[ppl]

According to my analysis VINs are produced in sequence, but are not released into production in sequence due to the fact that European cars were assigned VINs in large blocks, but are released into production gradually, to blend in with the production of cars for NA deliveries. There is also a certain disturbance to the sequence of VINs that are coming through the production due to batching as I explained earlier (2Q 2013 Model S Deliveries Potential Surprise - Page 7).

I used this explanation to correctly predict the cars delivered in Q2. My prediction was 5200 vs. 5150 delivered actually, so I am quite confident that this theory was validated. (Short-Term TSLA Price Movements - Page 341)

Regarding the VIN 21,591 reported by CygnusX-1 that was scheduled for the 9/28 delivery, it indicates that deliveries for Q3 will be approximately in the 5,400 range. Bear with the explanation below to see how this prediction was made.

According to the interview Elon did in Europe before the Geneva Auto Show this year Tesla planned to deliver about 5,000, "may be" 7,000 cars. Just recently there were reports on this Forum that European deliveries for this year are already booked. Assuming that Tesla allocated 5,000 spots for this year's delivery to Europe, there were 5,000 VINs assigned to cars for European Delivery. I have used lower 5,000 number because I think that while according the Elon NA orders are higher than expected, the pace of European orders did not pick up yet, so it is logical to assume that greater proportion of cars would be allocated for NA deliveries.

Assume that 5,000 European delivery cars will be allocated evenly: 2,500 for Q3 and 2,500 for Q4. Out of the 2,500 cars produced for European Delivery in Q3 about 1,500 would be delivered in Q3, while remaining 1,000 will be delivered in Q4 due to shipping/assembly delay (assumed about 5 weeks). Based on the above out of 5,000 VINS that were assigned to the European delivery cars only 1500 cars will be delivered in Q3. So the total quantity of VINs that are already assigned but are not going to be delivered in Q3 is 5,000-1,500=3,500.

The total quantity of cars delivered in Q4 2012 and Q1, Q2 2013 was 12,700. Calculating approximate number of predicted deliveries in Q3: 21,591-12,700-3,500=5,391.

Regarding the accuracy of the above prediction, the number could be different due to the following factors:

1. Tesla could have arranged production in such a way that the majority of European Delivery cars were produced in the first half of the Q3, so more than 1,500 of them are delivered in Q3

2. The production is about 600 cars weekly, so about 120 cars are produced daily. Therefore about 85 VINs would be delivered on a given day (600/7). So cars scheduled for delivery on 9/28 could have VINs 21,591+/-85.

3. According to my original theory linked below the most common configuration is S85 with air suspension, and VINs for these cars represent projected deliveries with the best accuracy. The post by CygnusX-1, however, does not mention the air suspension, so VIN 21,591 was assigned to S85 with coil suspension.

All in all I feel that 5,400 number arrived at above has a good chance to be higher by one to several hundreds.

You accounted for the European cars in transit at the end of 3 qtr. but I don't believe you included the 500- 700 in transit from 2 nd qtr being delivered now

 

[bonaire]

I am close to his estimate but am thinking 5500-5550 for Q3. They are shoehorning in some fast deliveries as noted in the forum posting. I think you would see more "I got mine" threads here if deliveries were much higher. On the gm-volt.com site, there have been a lot of those threads and august sales will be 3400 or so in the US only and perhaps 4k worldwide.

i think tracking the vins will backfire. Many stock guys are thinking that Q3 will be like 6500-7000 sales. If sales really are 5500 or so, then a bunch of negative feedback may occur. However we wont know until early November ad monthly sales are not announced.

 

[EVNow]

One thing to keep in mind. Nissan has started selling CARB credits.

 

[vgrinshpun]

You accounted for the European cars in transit at the end of 3 qtr. but I don't believe you included the 500- 700 in transit from 2 nd qtr being delivered now

There were plans to produce sizable batch of cars for European deliveries in Q2, but the general consensus on this Forum is that these plans did not materialize.

During the Teslive conference there were some factory tours organized by Tesla. Several participants reported seeing congratulatory message on a computer screen during the factory tour stating that 5454 cars were built during the Q2. Since 5150 cars were delivered to customers, and some loaners and demo cars for new stores that were open during the Q2 were also produced during Q2, it is highly unlikely that any sizable batch of cars was put on a boat for European deliveries during the Q2. This is the reason that I did not count any European cars that were produced but not delivered in Q2 towards the Q3 estimate.

 

[ItsNotAboutTheMoney]

One thing to keep in mind. Nissan has started selling CARB credits.

I don't think that's important because:
- Tesla has cash
- ZEV credits can be banked
- the Tesla focus in q3 and q4 is gross margin ex ZEV
As long as Tesla is on target or better (17% ex ZEV will do) Q3 will be considered a success.

 

[sleepyhead]

I don't think that's important because:
- Tesla has cash
- ZEV credits can be banked
- the Tesla focus in q3 and q4 is gross margin ex ZEV
As long as Tesla is on target or better (17% ex ZEV will do) Q3 will be considered a success.

Anything below 19% would be a failure.

They went from -4% to 5% to 13% in the past two quarters and they need to get to 19% to get halfway there. It has to be be higher and it will.

 

[DonPedro]

According to my analysis VINs are produced in sequence, but are not released into production in sequence due to the fact that European cars were assigned VINs in large blocks, but are released into production gradually, to blend in with the production of cars for NA deliveries. There is also a certain disturbance to the sequence of VINs that are coming through the production due to batching as I explained earlier (2Q 2013 Model S Deliveries Potential Surprise - Page 7).

I used this explanation to correctly predict the cars delivered in Q2. My prediction was 5200 vs. 5150 delivered actually, so I am quite confident that this theory was validated. (Short-Term TSLA Price Movements - Page 341)

Regarding the VIN 21,591 reported by CygnusX-1 that was scheduled for the 9/28 delivery, it indicates that deliveries for Q3 will be approximately in the 5,400 range. Bear with the explanation below to see how this prediction was made.

According to the interview Elon did in Europe before the Geneva Auto Show this year Tesla planned to deliver about 5,000, "may be" 7,000 cars. Just recently there were reports on this Forum that European deliveries for this year are already booked. Assuming that Tesla allocated 5,000 spots for this year's delivery to Europe, there were 5,000 VINs assigned to cars for European Delivery. I have used lower 5,000 number because I think that while according the Elon NA orders are higher than expected, the pace of European orders did not pick up yet, so it is logical to assume that greater proportion of cars would be allocated for NA deliveries.

Assume that 5,000 European delivery cars will be allocated evenly: 2,500 for Q3 and 2,500 for Q4. Out of the 2,500 cars produced for European Delivery in Q3 about 1,500 would be delivered in Q3, while remaining 1,000 will be delivered in Q4 due to shipping/assembly delay (assumed about 5 weeks). Based on the above out of 5,000 VINS that were assigned to the European delivery cars only 1500 cars will be delivered in Q3. So the total quantity of VINs that are already assigned but are not going to be delivered in Q3 is 5,000-1,500=3,500.

The total quantity of cars delivered in Q4 2012 and Q1, Q2 2013 was 12,700. Calculating approximate number of predicted deliveries in Q3: 21,591-12,700-3,500=5,391.

Regarding the accuracy of the above prediction, the number could be different due to the following factors:

1. Tesla could have arranged production in such a way that the majority of European Delivery cars were produced in the first half of the Q3, so more than 1,500 of them are delivered in Q3

2. The production is about 600 cars weekly, so about 120 cars are produced daily. Therefore about 85 VINs would be delivered on a given day (600/7). So cars scheduled for delivery on 9/28 could have VINs 21,591+/-85.

3. According to my original theory linked below the most common configuration is S85 with air suspension, and VINs for these cars represent projected deliveries with the best accuracy. The post by CygnusX-1, however, does not mention the air suspension, so VIN 21,591 was assigned to S85 with coil suspension.

All in all I feel that 5,400 number arrived at above has a good chance to be higher by one to several hundreds.

I won't dispute your estimate of 5,400 - it is common sense based on previous quarters, guidance and other info. However. I don't think your model has anywhere near the accuracy to tell whether the number will end up being 4,500 or 6,000.

Your key assumption is the number of EU cars that have had VINs assigned but which won't be delivered in Q3. You assume 3,500, but this is virtually guesswork. First of all, you assume that 5,000 VINs have been assigned in the EU, while admitting that 2013 production could be 7,000. On the other hand, there are plenty of examples of people in the EU who have been promised delivery in November/December, but who have not yet received their VINs. So the true number could be close to 7,000 og it could be 3,000.

Second, you assume EU deliveries will be 1,500 in Q3. This is high. It could be as low as 500.

If you choose the assumptions 7,000 and 500, then your model spits out a number of 2,400. If you chose 3,000 and 1,500, then your model says 7,400. None of these estimates are realistic, but the variation shows that you can justify almost any estimate just by tweaking the assumptions in the model.

We know too little about the production sequence to use VINs to estimate deliveries.

Assignment of VINs looks more promising to me. I read cfOH's efforts with interest.

 

[Acmykguy]

Vins aside, remember the unbundling and price increase that went into effect in early q3. My car would have been 9,000 give or take more with same config. That means demand is there, they can charge more for the same options all while increasing the gross and net margin. 3 q "should have upside from q2. UNLESS the price increase and the increase profit margin makes up for a decrease in zev revenue that Elon already stated would be light in q3.

 

[blakegallagher]

Anything below 19% would be a failure.

They went from -4% to 5% to 13% in the past two quarters and they need to get to 19% to get halfway there. It has to be be higher and it will.

And on top of this Elon has said that they will make it more than halfway to 25 percent GM in q3 .... so I am expecting 20 percent GM for q3.

 

[Clprenz]

I honestly think we produced about 8,500 cars. Being pessimistic I don't think is could be less that 7,250 delivered. At a ASP of $97,000 due to price increases, Euro sigs, and performance plus's. That's 703,250,000 in car revenue. I would add about 30 million in ZEV credits. Non ZEV credit margin of 21%. That's 147,682,500 in pure non ZEV profit, with ZEV = 177,682,500, CAPEX of 115,000,000. Profit of 62,682,500, EPS: .52 and if converted notes EPS: .46

2013 Price target $220-$236 EPS: $1.69

 

[2Pearls]

I Have heard from an informed source that they are presently manufacturing 700+ cars per week. 500 for US and 200 for Europe. Has anyone else heard this or taken a tour to see what they are manufacturing?

 

[cfOH]

@2Pearls: Well, my regression model puts them currently at about 630 VIN assignments per week over the past 2 months. But, if I look at data from just the past two weeks, we're at around 715 VIN assignments per week, so your source sounds like he is relating a factual estimate. I do not know the breakdown of NA vs EU, but 500/200 sounds reasonable.

 

[2Pearls]

@2Pearls: Well, my regression model puts them currently at about 630 VIN assignments per week over the past 2 months. But, if I look at data from just the past two weeks, we're at around 715 VIN assignments per week, so your source sounds like he is relating a factual estimate. I do not know the breakdown of NA vs EU, but 500/200 sounds reasonable.

cFOH, that is exactly right, this is just for the last couple of weeks that they are at around 700+ per week. And, I also understand that sales are actually increasing from where they were just about 2 months ago...so, business looks very good moving forward.

 

[vgrinshpun]

I Have heard from an informed source that they are presently manufacturing 700+ cars per week. 500 for US and 200 for Europe. Has anyone else heard this or taken a tour to see what they are manufacturing?

The 200 cars/week designated for European delivery over 6 month matches up with Elon projection made during the interview at Geneva's car show when he mentioned that about 5,000 cars will be slated for European delivery in 2013. He actually said 5000, "may be" 7000. As I explained in my previous post, Tesla settled on 5,000 because demand in US picked up nicely, while reservations in Europe so far stayed relatively flat.

It is also interesting that although Elon mentioned during the Q2 call that wait time in US will increase in second half of the year, so far Tesla was able to avoid increased wait time by increasing production to match the increasing flow of orders. I have a feeling that they will ramp up to 800 cars a week much sooner than end of 2014 that was mentioned by Elon during the Q2 call. Q4 2013 seems to be more reasonable prediction.

 

[justdoit]

The 200 cars/week designated for European delivery over 6 month matches up with Elon projection made during the interview at Geneva's car show when he mentioned that about 5,000 cars will be slated for European delivery in 2013. He actually said 5000, "may be" 7000. As I explained in my previous post, Tesla settled on 5,000 because demand in US picked up nicely, while reservations in Europe so far stayed relatively flat.

It is also interesting that although Elon mentioned during the Q2 call that wait time in US will increase in second half of the year, so far Tesla was able to avoid increased wait time by increasing production to match the increasing flow of orders. I have a feeling that they will ramp up to 800 cars a week much sooner than end of 2014 that was mentioned by Elon during the Q2 call. Q4 2013 seems to be more reasonable prediction.

I understand the whole set low expectations and over perform but a whole year ahead of schedule? I'm assuming ramping up from 700 to 800 will be harder than 500 to 700 just like 800 to 900 will be harder than 700 to 800. Improving something that's really inefficient is a lot easier than improving something that's already pretty efficient. I think Elon knows this so that's why I'd be really surprised if they reached 800 by end of 2013. I mean, that's a whole year ahead of schedule! Maybe early to mid next year.

 

[vgrinshpun]

I understand the whole set low expectations and over perform but a whole year ahead of schedule? I'm assuming ramping up from 700 to 800 will be harder than 500 to 700 just like 800 to 900 will be harder than 700 to 800. Improving something that's really inefficient is a lot easier than improving something that's already pretty efficient. I think Elon knows this so that's why I'd be really surprised if they reached 800 by end of 2013. I mean, that's a whole year ahead of schedule! Maybe early to mid next year.

You've got me all confused, which schedule do you mean, the low expectations, or the real one? :smile:

Seriously, though, initially Elon mentioned 800 cars/week by the end of 2014 in the interview he gave before the Q2 call. He was then asked during the call about the background info on this number. Elon responded that it was based on reasonable assumption on demand at that time. So it is clear that both ramp-ups (to 600 cars/week, and now to 700 cars/week) are closely tied to the adding of production for European cars, and increase in rate of US orders. If this pattern continues, they will have to go to 800 cars/week in Q4.

Another thing to consider is that at these rates they already utilizing second shift. In order to hit the 25% margin, though, they will likely need to be at 800 cars/week for a two shift operation. So this metric indicates that they should be at 800 cars/week in Q4.

500/200 split is also interesting and have significance IMO. The 500 probably roughly matches current rate of US orders, while 200 covers 5000 cars promised for European market. I think that rate of incoming European orders will likely increase as first European cars start to arrive. This will require ramp up, as it seems that Tesla is trying to hold the line on the wait time.

 

[bonaire]

If you are timing interviews and car build counts out into late 2014, would some of the assumptions that were stated including Model-X builds? With someone here saying that Model-X was said to be slipping into 2015, then all of 2014 would be all Model-S. Do we really think Model-S will be at a 800/wk demand through the end of 2014?