Third quarter deliveries 2017

[dennis]

I am skeptical about 1500 in Sept. But, (a reminder this is a PRODUCED number 1500) while it is frustrating that we are not seeing high vins, our data sample size might just be too low. With S and X civilians were getting them and were eager to talk about them. With the 3, the opposite is true. The "German tank problem" requires SOME data. We might really just be whiffing on the VIN's out there.

Also, didn't they essentially produce the 1-300 VINS twice? So if the highest VIN we had ever seen was 300 (this is a hypothetical) all we would know for sure is that total produced is between ~50 and 600, since some of those VIN's could have both RC and production VIN plates.

Finally, since the metric is produced cars, they could have somewhere between zero and infinity on trucks, boats and in factory staging areas and we wouldn't know.

None of this proves a good or bad delivery number. But I suspect our DATA this time is unusually useless. And, of course, THIS WAS BY DESIGN. The whole plan was to keep a lid on early deliveries to keep bugs and rampant software updates in the family. The side effect is that we don't have VIN feedback. Tesla would not think this is a bug, since they fully intend to update us Oct 3.

The RC's have the big red Off button so we aren't seeing anything twice.

I walked the entire employee charging area at Fremont last week. I saw 11 Model 3's, all with VINs between 11 and 179. That count is up from 4 that I saw doing the same thing about a month ago. I also observed VIN 310 (in the 3XX range that was previously believed to be assigned to production validation cars produced in July) being delivered to a customer at the Fremont Delivery Center. Cars are being delivered to SpaceX employees in SoCal, Texas and Florida. The highest reported VIN is 259.

There is such intense interest in this at both TMC and Reddit that it is highly unlikely that we have missed the delivery of a Model 3 with a 4XX or 5XX VIN. The data is not useless.

 

[SBenson]

Based on @schonelucht commentary and @vgrinshpun analysis I do believe that S/X numbers will be pretty good. However based on gross margin guidance looks like the financials will still be pretty terrible.

Having said that, I was of the view it is all about Model 3 now. Lack of progress on Model 3 front is not encouraging. Skimming through the Model 3 threads seem to imply that the product is not ready for prime time yet. I am very doubtful that the real production line is up and fully functional.

Stock market has it's own mind, no idea how the market will receive all this. After all this is just temporary in the inevitable road to a great Model 3.

I'm myself staying away from any sort of leverage, despite all the excitement here when sp zoomed into 380s last week. See TT's hyper excitement every time there is a spike up, hope nobody's taking him seriously and over-leveraging themselves. As I see it Tesla the company and it's operation itself is an exercise in pushing the risk to the limits. So any money you put in the stock is effectively already very highly leveraged... Related story: at one time I was talking to a software engineer in the AP team, back in the mobileye's days. So I asked him why is Tesla making the kind of progress that others are not making. His answer was that they are willing to take on much more risk and push mobileye's product (and the company) to the edge. He didn't say brilliant engineers or technical competence or anything like that... Over time I realized that Tesla pushes risk to the extreme in *every* dimension... Hence there will be setbacks every now and then, and that partly explains why the stock is so volatile... My basic point is that the company and stock are so extremely volatile and risky. Do you want to take that up one more level by leveraging up.

Add: That chat with AP engineer was much before Joshua Brown incident by the way.

 

[Out4aDuck]

Regarding Model 3 deliveries:

My complete wild guess is that the production line is not in place yet (or some supply is not in place, could it be the batteries?). As people say the few hundred we saw are most likely on the pilot line. I will only have confidence on the full production line once we see 1XXX VIN and beyond. There is not even a sign of 4XX yet.

Early this year I used to argue that with Model 3 the complexity moved from the car to the production line, as Musk is trying to robot automate ever more steps. So the production ramp is highly unpredictable. It's not like complexity went away with simpler design, it just moved up to the factory line... I stopped arguing that as Musk provided ever more confidence on the ramp over time, going all the way to giving guidance on monthly time frames.

As time is passing by I feel that this too is just another example of Musk's over promise (delusional thinking as he calls it). We are still not past the TE ramp and EAP ramp that he promised way back.

However I certainly hope that we will see some big Model 3 numbers soon. As this ramp is far more important than any prior ramp as costs pile up very dramatically and revenue needs to catch-up soon.

SBenson, I think you hit the nail on the head. With a few hundred release candidates out there now, there's sufficient numbers to get meaningful feedback. It makes sense to stop the line now and finish tooling it up for volume production.

 

[ValueAnalyst]

Based on @schonelucht commentary and @vgrinshpun analysis I do believe that S/X numbers will be pretty good. However based on gross margin guidance looks like the financials will still be pretty terrible.

Having said that, I was of the view it is all about Model 3 now. Lack of progress on Model 3 front is not encouraging. Skimming through the Model 3 threads seem to imply that the product is not ready for prime time yet. I am very doubtful that the real production line is up and fully functional.

Stock market has it's own mind, no idea how the market will receive all this. After all this is just temporary in the inevitable road to a great Model 3.

I'm myself staying away from any sort of leverage, despite all the excitement here when sp zoomed into 380s last week. See TT's hyper excitement every time there is a spike up, hope nobody's taking him seriously and over-leveraging themselves. As I see it Tesla the company and it's operation itself is an exercise in pushing the risk to the limits. So any money you put in the stock is effectively already very highly leveraged... Related story: at one time I was talking to a software engineer in the AP team, back in the mobileye's days. So I asked him why is Tesla making the kind of progress that others are not making. His answer was that they are willing to take on much more risk and push mobileye's product (and the company) to the edge. He didn't say brilliant engineers or technical competence or anything like that... Over time I realized that Tesla pushes risk to the extreme in *every* dimension... Hence there will be setbacks every now and then, and that partly explains why the stock is so volatile... My basic point is that the company and stock are so extremely volatile and risky. Do you want to take that up one more level by leveraging up.

Add: That chat with AP engineer was much before Joshua Brown incident by the way.

My valuation-hat take is that SP has already priced in a more pessimistic scenario than "setbacks every now and then."

I estimate that what's priced in at $350 SP is 200k in 2018 and 500k Model 3's in 2020 with some but not much TE.

 

[luvb2b]

i did some work estimating q3 deliveries. i posted details in the general thread.

my estimate is 25,739 deliveries of s&x, plus 300 deliveries of model 3. record s&x deliveries, but a bit disappointing for 3's.

 

[mrdoubleb]

So my scientific guesstimate is 28,600 deliveries and close to 30k production

Ok, so I have to temper my expectations a bit. I was really hoping we would hear word of M3 ramp by now, but everything points to a delay. How serious the reason? Don't know, probably not very as Elon seems to be in a good tweeting mood and working on other stuff vs. sleeping on the production floor. So it may be just some supplier causing a few weeks of delay.

Anyway, I am now thinking <300 M3 deliveries and high 25k or low to mid 26k S/X. So I'll go with 26001 total deliveries just for the heck of it.

 

[BLSmith2112]

Would be disappointed if it were a supply issue, seeing as their working with the "A Players" now instead of the C/D/E/F players. Haven't seen anything to suggest their on track ether.

 

[Turing]

Ok, so I have to temper my expectations a bit. I was really hoping we would hear word of M3 ramp by now, but everything points to a delay. How serious the reason? Don't know, probably not very as Elon seems to be in a good tweeting mood and working on other stuff vs. sleeping on the production floor. So it may be just some supplier causing a few weeks of delay.

Anyway, I am now thinking <300 M3 deliveries and high 25k or low to mid 26k S/X. So I'll go with 26001 total deliveries just for the heck of it.

Does the market freak out over only 300ish M3 deliveries or does it react positively to increasing s/x demand?

 

[mrdoubleb]

Does the market freak out over only 300ish M3 deliveries or does it react positively to increasing s/x demand?

I am not sure 1-2k S/X deliveries above expectations will make up for a miss on M3. If Tesla is behind schedule, they better put some encouraging language around it, like a cause they have already fixed or a revised schedule of the ramp that's just a little off plan. A 2-3 week delay is not the end of the world, but if they just don't say anything this will be spinned hard by the bears.

They normally do a very short, factual announcement on production and deliveries, but when X was behind schedule they made an exception and gave some background. So this would not be unprecedented and could soften the blow.

 

[Mike Smith]

I am not sure 1-2k S/X deliveries above expectations will make up for a miss on M3. If Tesla is behind schedule, they better put some encouraging language around it, like a cause they have already fixed or a revised schedule of the ramp that's just a little off plan. A 2-3 week delay is not the end of the world, but if they just don't say anything this will be spinned hard by the bears.

They normally do a very short, factual announcement on production and deliveries, but when X was behind schedule they made an exception and gave some background. So this would not be unprecedented and could soften the blow.

If the 3 number is low they will definitely describe the problem and provide updated q4 guidance with the release. This guidance will determine how the market reacts, not the q3 number.

 

[Bgarret]

I am skeptical about 1500 in Sept. But, (a reminder this is a PRODUCED number 1500) while it is frustrating that we are not seeing high vins, our data sample size might just be too low. With S and X civilians were getting them and were eager to talk about them. With the 3, the opposite is true. The "German tank problem" requires SOME data. We might really just be whiffing on the VIN's out there.

Also, didn't they essentially produce the 1-300 VINS twice? So if the highest VIN we had ever seen was 300 (this is a hypothetical) all we would know for sure is that total produced is between ~50 and 600, since some of those VIN's could have both RC and production VIN plates.

Finally, since the metric is produced cars, they could have somewhere between zero and infinity on trucks, boats and in factory staging areas and we wouldn't know.

None of this proves a good or bad delivery number. But I suspect our DATA this time is unusually useless. And, of course, THIS WAS BY DESIGN. The whole plan was to keep a lid on early deliveries to keep bugs and rampant software updates in the family. The side effect is that we don't have VIN feedback. Tesla would not think this is a bug, since they fully intend to update us Oct 3.

While I tend to believe the most obvious answer - slow Model III ramp, is also the most likely, there are a couple of things to consider:

1. If the production is for Fremont employees, there is no shipping time and there would be some advantage to holding a good number for extended Quality Control checks and releasing them into the wild at the end of the month.
2. There would also be some advantage to make a rather large batch of initial production for stores/test drives, which would count for production, but not delivery.
3. Because of the anticipated slope of the ramp, 1-2 weeks behind could be 500 cars or more. If Tesla is able to state that they are 1 week behind on the ramp, but see the path forward, that would go a long way towards assuaging any concerns.

I still am tending toward the obvious answer to not seeing Model III in the wild, but have some optimism based on the items above.

My quarterly prediction - 27,725.

 

[neroden]

So here's an optimistic thought. Perhaps they have gotten enoufgh feedback from the first few hundred cars that they are making deliberate production line design changes -- right now. Might as well put the pause in production early, iron out the bugs now rather than when more cars are on the road.

 

[luvb2b]

i think if they miss model 3 by a bit but are moving forward on making cars it'll be fine. that's because there's a long line for model 3, it's not like we think they can't sell them.

on the other hand, if s/x demand is weaker that's a huge problem. there's no way to magically reverse that. and the s&x drag will offset revenue gains from the 3.

so i do think getting 26k s/x deliveries will trump any whiff on the 3. my own modeling indicates revenues will track around the $3 billion level for the whole company if the s/x deliveries are coming in at 26k. that's a 35+% jump from a year ago and should bring net loss closer to $1.50-1.75/share vs something worse than $2/share.

I am not sure 1-2k S/X deliveries above expectations will make up for a miss on M3. If Tesla is behind schedule, they better put some encouraging language around it, like a cause they have already fixed or a revised schedule of the ramp that's just a little off plan. A 2-3 week delay is not the end of the world, but if they just don't say anything this will be spinned hard by the bears.

They normally do a very short, factual announcement on production and deliveries, but when X was behind schedule they made an exception and gave some background. So this would not be unprecedented and could soften the blow.

 

[schonelucht]

so i do think getting 26k s/x deliveries will trump any whiff on the 3. my own modeling indicates revenues will track around the $3 billion level for the whole company if the s/x deliveries are coming in at 26k. that's a 35+% jump from a year ago and should bring net loss closer to $1.50-1.75/share vs something worse than $2/share.

Quick back of the envelope. $3B revenue. 20% gross margin (guidance is slightly lower) is $600M. Flat R&D + SG&A is slightly over $-900M. Interest expense is another slightly over $-100M. Losses a ttributable to NCI $+65M. Gives at least a loss $-335M or roughly $2/share. Where do you diverge? More income from credits? Positive exchange rate gains ('positive other income' which was $-40M)?

I haven't done the detailing yet but the above suggest to me that a loss over $2 or more per share is quite plausible.

 

[luvb2b]

auto gross margins on s&x should be quite good at that level of production. they will slightly be offset by the 3's, but i still think 20% gm would be way low for that level of production. heck they were several percentage points higher than that last quarter with much lower production (i show total company gross profit at 666.6m vs. revenue of 2,789m or near 24% gross margins for the whole company in q2 17).

Quick back of the envelope. $3B revenue. 20% gross margin (guidance is slightly lower) is $600M. Flat R&D + SG&A is slightly over $-900M. Interest expense is another slightly over $-100M. Losses a ttributable to NCI $+65M. Gives at least a loss $-335M or roughly $2/share. Where do you diverge? More income from credits? Positive exchange rate gains ('positive other income' which was $-40M)?

I haven't done the detailing yet but the above suggest to me that a loss over $2 or more per share is quite plausible.

 

[erthquake]

So here's an optimistic thought. Perhaps they have gotten enoufgh feedback from the first few hundred cars that they are making deliberate production line design changes -- right now. Might as well put the pause in production early, iron out the bugs now rather than when more cars are on the road.

This has been my thought as well. A slightly less-optimistic view is that the dearth of new Model 3 VINs is because they are switching to, or have recently switched from, a pilot general assembly line to the main one.

 

[schonelucht]

auto gross margins on s&x should be quite good at that level of production. they will slightly be offset by the 3's, but i still think 20% gm would be way low for that level of production.

I know, but they did guide for below 20%. I think the offset by Model 3 is going to be horrible. And these are mostly costs that they have regardless of how much they produce. Mostly manufacturing employee costs. It doesn't matter if it's for 400 Model 3 or 1500 Model 3, their salaries need to be paid.

 

[luvb2b]

agree. perhaps i haven't included enough negative margin inclusion for the model 3.

i calculated things another way, using what i view as vin assignments. this quarter has seen the highest level of vin assignments of any since q4 2016. which to me means there's a lot of cars made. in q4 16 you could say they were making some cars for hong kong q1 deliveries. and china had a lot of deliveries too.

but there's no hong kong this time. so where are all these vins going? they don't appear to be sitting in inventory from the inventory counts.

i'm guessing that people are feeling bearish on deliveries because the 8,071 estimated sales to usa and eu in the first 2 months of the quarter is less than the ~8800 seen in the first 2 months of the 3rd quarter of 2016. if you use simple extrapolation it would imply 10% lower deliveries than q3 2016, or about 22.5k for s/x deliveries.

however, that extrapolation i don't believe is correct. it's because the non-usa/eu portion of the sales are much higher now than 1 year ago. so 8,000 usa/eu deliveries could match to 25k global deliveries because there are another 2-2.5k vehicles being sold to china vs. ~2-3k sold to china in q3 2016.

I know, but they did guide for below 20%. I think the offset by Model 3 is going to be horrible. And these are mostly costs that they have regardless of how much they produce. Mostly manufacturing employee costs. It doesn't matter if it's for 400 Model 3 or 1500 Model 3, their salaries need to be paid.

 

[ValueAnalyst]

So here's an optimistic thought. Perhaps they have gotten enoufgh feedback from the first few hundred cars that they are making deliberate production line design changes -- right now. Might as well put the pause in production early, iron out the bugs now rather than when more cars are on the road.

This should have been anticipated by Musk when he tweeted out "1,500 in September."

 

[Carl]

Tesla Model 3 could imho very well break the VW Beetle's record of 21 million cars built. And not over 65 years, but over 15 years or so.

So why the heck should we care about whether Tesla built 174 or 226 or 476 or whatever amount of M3's in September 2017? Tesla will deliver and it doesn't matter when, in 2017, as Tesla is years in advance of competitors anyway!