Near-future quarterly financial projections

[EVNow]

That's why I say 3+Y will be 10k/week. We'll find out in January, I suppose.

I posted this in the market thread - with the context being Musk saying GF4 will start with Y.

Does that also mean Y production volume in US will be lower than previously thought ? Already we know they will make Y in GF3 too.

I think we should start thinking about Y production being on 3 continents from get go - may be 4k/wk in US, 3k/wk in China and 3k/wk in EU.
​

I think the 3+Y being 10k/wk is likely correct. They won't necessarily cut 3 production - but Y production will be built out only for 3k. As 3 production for China moves to GF3 (starting Q1) and GF4 comes along (2021 ?) - Y will take up the slack in Fremont. We should expect 3/Y mix in Fremont to reflect demand in NA. So, yes, Fremont will produce fewer 3s than they do now - but overall 3 production WW will expand.

It will be somewhat chaotic and complex as
Q1/Q2 : GF3 gets to volume production, we won't know what the steady state demand will be ...
Q2/Q3 : Y comes online and ramps up. We won't know what the steady state demand will be for a while. Also they will probably only sell Y in US first - and when China starts making Y, they will sell it there.
2021 : GF4 comes online making Y. But will Fremont Y make its way to EU before that ?

Then there is the Pickup & Semi to be still made in US. We'll have the uncertainty around how far FSD can progress in the meantime driving demand and margin.

 

[EVNow]

The chart actually says that an inexpensive Japanese car is the most likely trade-in for the M3.

But they are saying - as % of sales Toyota is not getting hot much, since they sell so many more. But BMW losing so many sales is hitting them.

 

[electracity]

Q1/Q2 : GF3 gets to volume production, we won't know what the steady state demand will be ...

Tesla raising the price of GF3 M3 suggests they are not worried about demand. They are likely most worried about batteries.

 

[Thekiwi]

I posted this in the market thread - with the context being Musk saying GF4 will start with Y.

Does that also mean Y production volume in US will be lower than previously thought ? Already we know they will make Y in GF3 too.

I think we should start thinking about Y production being on 3 continents from get go - may be 4k/wk in US, 3k/wk in China and 3k/wk in EU.
​

I think the 3+Y being 10k/wk is likely correct. They won't necessarily cut 3 production - but Y production will be built out only for 3k. As 3 production for China moves to GF3 (starting Q1) and GF4 comes along (2021 ?) - Y will take up the slack in Fremont. We should expect 3/Y mix in Fremont to reflect demand in NA. So, yes, Fremont will produce fewer 3s than they do now - but overall 3 production WW will expand.

It will be somewhat chaotic and complex as
Q1/Q2 : GF3 gets to volume production, we won't know what the steady state demand will be ...
Q2/Q3 : Y comes online and ramps up. We won't know what the steady state demand will be for a while. Also they will probably only sell Y in US first - and when China starts making Y, they will sell it there.
2021 : GF4 comes online making Y. But will Fremont Y make its way to EU before that ?

Then there is the Pickup & Semi to be still made in US. We'll have the uncertainty around how far FSD can progress in the meantime driving demand and margin.

Good post.

On Europe site, Tesla says 2021 for model Y European deliveries, and 2022 for Model Y SR. So one could reasonably assume that it is intended for Model Y Deliveries in Europe to all be from GF4. Although to be fair US made Ys could still end up being sent to Europe if either:
A) GF4 is significantly behind schedule or
B) US demand OR production capacity evolves to the point where it means delivering high priced Y variants to EU before GF4 opens makes best financial sense for Tesla.

One wonders where the pick up truck will fit into all this. I know many are keen to see a new factory built in US to cope with increasing volumes brought on by the Y & Pickup, but its maybe in Teslas best interest for their 2022/23 production plans AT THIS STAGE to be somewhat cautious in case of a recession. It might turn out that the Pickup could also be produced at Fremont in the event of a slowdown in auto sales from a recession - leaving maybe a 3rd/3rd/3rd split for 3/Y/P in that scenario while both GF3 & GF4 are both up and running for 3 & Y by then to serve local markets. It would actually be perfect to introduce a popular new variant at a time when global auto demand became temporality suppressed due to a recession.

One could thing about GF4 completion will be that Tesla will be able to serve any country in the world with vehicles from whatever region has the most favourable trade deals.

 

[Doggydogworld]

I posted this in the market thread - with the context being Musk saying GF4 will start with Y.

Does that also mean Y production volume in US will be lower than previously thought ? Already we know they will make Y in GF3 too.

I think we should start thinking about Y production being on 3 continents from get go - may be 4k/wk in US, 3k/wk in China and 3k/wk in EU.
​

I think the 3+Y being 10k/wk is likely correct. They won't necessarily cut 3 production - but Y production will be built out only for 3k. As 3 production for China moves to GF3 (starting Q1) and GF4 comes along (2021 ?) - Y will take up the slack in Fremont. We should expect 3/Y mix in Fremont to reflect demand in NA. So, yes, Fremont will produce fewer 3s than they do now - but overall 3 production WW will expand.

It will be somewhat chaotic and complex as
Q1/Q2 : GF3 gets to volume production, we won't know what the steady state demand will be ...
Q2/Q3 : Y comes online and ramps up. We won't know what the steady state demand will be for a while. Also they will probably only sell Y in US first - and when China starts making Y, they will sell it there.
2021 : GF4 comes online making Y. But will Fremont Y make its way to EU before that ?

Then there is the Pickup & Semi to be still made in US. We'll have the uncertainty around how far FSD can progress in the meantime driving demand and margin.

With Y coming to GF3 at some point and GF4 from the start, it's possible Fremont could only ever make US/Canada Ys. But I doubt it. If Y has higher margins, why hold it back? Export to Europe and Asia from day one. Or at least day two. Y won't have Model 3's huge pent-up US demand, and it can piggyback on 3's global logistics.

I could also see them sending China 3s to Europe. That would let Fremont really focus on Ys. Later, as GF3 starts making Ys and GF4 starts up, Fremont could move a Y line over to pickup production.

Of course I'm thinking from a 500-600k 3+Y perspective, not Musk's 2 million or the similar view of the TMC cognoscenti.

 

[EVNow]

With Y coming to GF3 at some point and GF4 from the start, it's possible Fremont could only ever make US/Canada Ys. But I doubt it. If Y has higher margins, why hold it back? Export to Europe and Asia from day one. Or at least day two. Y won't have Model 3's huge pent-up US demand, and it can piggyback on 3's global logistics.

I could also see them sending China 3s to Europe. That would let Fremont really focus on Ys. Later, as GF3 starts making Ys and GF4 starts up, Fremont could move a Y line over to pickup production.

Of course I'm thinking from a 500-600k 3+Y perspective, not Musk's 2 million or the similar view of the TMC cognoscenti.

In NA 3 is selling at the rate of some 60k/quarter (or 5k/wk). In 2 years time, taking into account some cannibalization, 10k/wk for 3+Y looks like a good base demand number. Its possible it could be a little higher - but then with Pickup, some people may move over from Y to pickup.

But in the interim, I think Tesla will do whatever they can to maximize margin. So if the demand for higher trim Y + 3 outstrips the production, they are going to make higher trim 3+Y. But what they won't do is to build out 7k/wk of Y in Fremont knowing GF3 is coming online now and in 2 years GF4 will be online.

 

[Doggydogworld]

In NA 3 is selling at the rate of some 60k/quarter (or 5k/wk).

It was 45k in Q3. InsideEVs said 44k in US alone, but they were too high (even after revising down).

79.7k Total
-26.6 Europe
-6k China (est)
-2k Australia/NZ/ROW
------
~45k North America

Might be a bit lower as my China number is conservative. They did 52-53k in Q2, thanks to a SR+ surge and new tax credits in Canada. Q1 was a disaster, of course, so YTD is 120-125k with full year 2019 looking like 165-170k.

I believe 3+Y will be 1.5x Model 3 alone, with the split heavily favoring Y. Thus my global baseline of 500k 3+Y. Upside to this comes from::
- Zero 3/Y cannibalization, as we saw with S/X
- Organic sales growth from word of mouth effect
- Huge untapped demand in China holding out for GF3 version

I see all three as possible, but not likely. Tesla's recent moves, e.g. building Y at Fremont and starting GF4 with Y support my view (at least in my mind, ha). I treat the numbers Musk throws out as long-term aspirations, like 105/150 GWh at GF1. I pay more attention to their capital plans and budgeting.

 

[EVNow]

79.7k Total

Oh yes, I took total delivery instead of just 3.

I think global y+3 being 500k is a bit low. In 2 years with GF3 and GF4, without Y it will probably be at that number. Basically 50% growth in 2 years. Now if we add Y, with some cannibalization we should see 1M globally. On the lower end may be 750k.

BTW, this will be less than expected production from the 3 factories.

 

[Thekiwi]

Crude attempt at Q4 estimates (Note I’ve combined leasing into automotive sales for now).

Any glaring totally implausible errors?

(Numbers in 1,000s obviously)

 

[The Accountant]

Crude attempt at Q4 estimates (Note I’ve combined leasing into automotive sales for now).

Any glaring totally implausible errors?

(Numbers in 1,000s obviously)
View attachment 477425

Thanks for the Q4 Estimate.
What are you assuming for total vehicle deliveries?

 

[Doggydogworld]

Crude attempt at Q4 estimates (Note I’ve combined leasing into automotive sales for now).

Any glaring totally implausible errors?

(Numbers in 1,000s obviously)
View attachment 477425

I see nothing glaring. I'm pretty close to your Auto Revs. I get 5600m Sale and 240m Lease based on 106k deliveries. Recognizing a large chunk of deferred FSD revenue or emissions credits would push Auto Revs higher. I'm 100k higher than you on Auto COGS, but if they dramatically improve unit COGS again and recognize almost no China COGS then even your 4450m is a bit high.

Your Service Rev/COGS guess is as good as mine - it's a total black box with virtually zero disclosure.

Energy has a ~30m headwind from seasonally lower solar kWh generation (70m vs. 100m). Solar sales are usually down in Q4, too, so maybe they drop from 40m to 30m. 1000 solar roofs per week would change that, of course, but don't hold your breath. I think they recognized the Nantucket Powerpacks in Q3, so I'd guess flat storage revs at 260m. Call it 360m total, with 285 COGS.

Interest expense could drop ~5m as they repaid 566m of convertibles November 1 and will repay 103m next month.

Q4 costs will tell us a lot about how many cookies Tesla pulled from the jar in Q3.

 

[EVNow]

Crude attempt at Q4 estimates (Note I’ve combined leasing into automotive sales for now).

Any glaring totally implausible errors?

(Numbers in 1,000s obviously)
View attachment 477425

I've $100 less in Revenue & COGS - so almost the same eps.

What are your ASP & Margin assumptions ?

Recognizing a large chunk of deferred FSD revenue or emissions credits would push Auto Revs higher.

Unlikely. No new functionality will likely be released in Q4 - we may see Traffic lights in EAP, hopefully. Besides, they probably want to keep the extra FSD revenue for Q1. They are doing a small # of HW3 upgrades now - but they will probably recognize the revenue equal to cogs on that.

Seems to me going forward - 2 things will determine TSLA path - apart from macros.
- Margin in Q4 (are Q3 improvements sustainable?)
- Deliveries in Q1 '20 (will it be significantly better than Q1 '19)

 

[EVNow]

I've $100 less in Revenue & COGS - so almost the same eps.

I meant $100M.

Energy has a ~30m headwind from seasonally lower solar kWh generation (70m vs. 100m). Solar sales are usually down in Q4, too, so maybe they drop from 40m to 30m. 1000 solar roofs per week would change that, of course, but don't hold your breath. I think they recognized the Nantucket Powerpacks in Q3, so I'd guess flat storage revs at 260m. Call it 360m total, with 285 COGS.

I wonder how much of PowerWall they can install considering the power outage in California must have generated overflowing orders.​

 

[Fact Checking]

Seems to me going forward - 2 things will determine TSLA path - apart from macros.
- Margin in Q4 (are Q3 improvements sustainable?)
- Deliveries in Q1 '20 (will it be significantly better than Q1 '19)

Also, production and deliveries in Q4 are a wildcard, and I think there's an increasing chance that Tesla will be able to surprise us positively in Q4 again:

• Q4 European deliveries are intriguingly high so far, per the graph posted by @KarenRei:
• 
  
  
• This suggests (but doesn't prove) a substantially higher early quarter production rate in Fremont compared to Q3. Just 35 days into Q4 Tesla started delivering at late-Q3 levels (!).
• This graph by @JustMe shows record Q4 units underway to Europe via ships:
• 
  
  
• The rate is significantly beyond any previous quarter: 30-40% higher than Q2, if the "ship loading hours" method is accurate.
• We still don't have an answer to the mystery production increase leak from Jerome, back in July: “While we can’t be too specific in this email, I know you will be delighted with the upcoming developments.”
• Q3 Model 3 production was 79k, up from 72k in Q2 - which I don't think matches the tone of Jerome's email.
• There's the leak to Cleantechnica about 7,000 excess Q3 battery packs sent to China:
• Tesla Gigafactory 3 Has ~7,000 Battery Packs In Stock For Chinese Model 3 Production, Will Use LG Chem Cells In 2020 | CleanTechnica
• U.S. order book appears to be almost 100% full in Q4 already, according to Tesla's own "weeks of delivery" estimates. (Which tend to lag true demand.)
• EU order book has closed for Q4 deliveries yesterday: most configs are for February 2020 delivery only.
• Inventory levels are very low in the U.S., to the extent Tesla allows us to see them.
• The 10-Q has shown a significant increase in non-finished goods inventory. Part of it could be battery packs for GF3 - but maybe they stockpiled parts for Fremont as well, for a full quarter demonstration of maximum sustained production rates?

Those 5-8 independent pieces of data/clues point towards a skillfully masked attempt by Tesla to hit the ball out of the ballpark in Q4: they might have sandbagged Q3 production and deliveries to maximize Q4 results.

If they do then even Q2 margins would be enough for record Q4 revenue and GAAP profits.

And here's a final mystery: in their Q4 report Tesla updated the 360k-400k 2019 deliveries guidance range to 360k only. Everyone, including me, interpreted this as an admission that they cannot hit 400k and can barely hit 360k in the best of cases.

But there's another possible explanation for why they removed the 400k upper guidance ...

Anyway, calling Q4 a record quarter at this point is premature I suspect, but the evidence so far is incredibly intriguing, and there's not a single counterfactual I've been able to find, other than Tesla's track record of punishing our optimism most of the time.

​

 

[hobbes]

• Q4 European deliveries are intriguingly high so far, per the graph posted by @KarenRei:
• View attachment 477890

Just regarding Karen´s plot: this is only NL (as I have mentioned before in the general thread) and not representative for all of Europe. We have three countries where we get live registration data: Netherlands, Spain, and Norway. I don´t find it very scientific to just pick one that is best performing. If you look at the individual countries here, you´ll see there is a steep rise even if you plot the sum of all three countries, but it is only driven by NL. It is still impressive and shows there is a shift in shipping times to earlier in the quarter, but we should not yet generalize this to expect a blowout quarter for Europe sales.

I find the shipping plot much more helpful because that gives us info for totals and not for individual countries.

EDIT: Let´s wait another two weeks until end of November when all European countries report their registrations before drawing conclusions. That said, I feel bullish, too.

 

[Fact Checking]

I don´t find it very scientific to just pick one that is best performing. If you look at the individual countries here, you´ll see there is a steep rise even if you plot the sum of all three countries, but it is only driven by NL. It is still impressive and shows there is a shift in shipping times to earlier in the quarter, but we should not yet generalize this to expect a blowout quarter for Europe sales.

Yeah, I agree, and if you check the 30-40% increase I estimated, that's the more realistic one based on the sum of all three countries. I cited Karen's graph because that's the one I found - but the ship loading days graph is impressive too.

Maybe there's a less optimistic explanation than a 8k/week Model 3 run-rate at Fremont - we'll see!

 

[The Accountant]

Just regarding Karen´s plot: this is only NL (as I have mentioned before in the general thread) and not representative for all of Europe. We have three countries where we get live registration data: Netherlands, Spain, and Norway. I don´t find it very scientific to just pick one that is best performing. If you look at the individual countries here, you´ll see there is a steep rise even if you plot the sum of all three countries, but it is only driven by NL. It is still impressive and shows there is a shift in shipping times to earlier in the quarter, but we should not yet generalize this to expect a blowout quarter for Europe sales.

I find the shipping plot much more helpful because that gives us info for totals and not for individual countries.

EDIT: Let´s wait another two weeks until end of November when all European countries report their registrations before drawing conclusions. That said, I feel bullish, too.

I agree that it's difficult to support a Q4 story by only looking at one country or a cluster of countries. However, partial data can give us good clues on assessing the quarter. For international shipments, I've used the same data set as @JustMe and instead of presenting it in a graph format, I have used a table. The table below indicates that international Q4 sales will be a record and perhaps as high as 28% above Q3. Shipping activity in month 3 of a quarter is typically low and we can assume a good portion of those shipments do not get delivered in time and thus spill over to the next quarter. So the first 2 months of shipments in each quarter is key. And as you can see from the table there have been 36 hours of loading time in the first 2 months of Q4 (Oct-Nov) much higher than the 28 days of loading time in Q3 (July-Aug) - a 28% increase to Q3.

Unless we see a huge drop in US sales to offset this international upside, Q4 is looking very good.
p.s. - I estimated the current ship loading now (the Dong A Glaucos) at 2.7 loading days.

 

[EVNow]

If they do then even Q2 margins would be enough for record Q4 revenue and GAAP profits.

I think the market definitely expects a record quarter in Q4 - otherwise you can't hit 360k for the year. So, the question for Q4 is one of margin.

For Q1, it is about deliveries - since that tells whether the Q3/Q4 being very good and Q1 being poor is a pattern that repeats from last year or not.

Once we cross these two quarters - it is looking pretty good with fewer uncertainties. Assumption being Tesla will manage the GF3 and Y ramp-ups much more smoothly than the 3.

 

[dc_h]

I agree that it's difficult to support a Q4 story by only looking at one country or a cluster of countries. However, partial data can give us good clues on assessing the quarter. For international shipments, I've used the same data set as @JustMe and instead of presenting it in a graph format, I have used a table. The table below indicates that international Q4 sales will be a record and perhaps as high as 28% above Q3. Shipping activity in month 3 of a quarter is typically low and we can assume a good portion of those shipments do not get delivered in time and thus spill over to the next quarter. So the first 2 months of shipments in each quarter is key. And as you can see from the table there have been 36 hours of loading time in the first 2 months of Q4 (Oct-Nov) much higher than the 28 days of loading time in Q3 (July-Aug) - a 28% increase to Q3.

Unless we see a huge drop in US sales to offset this international upside, Q4 is looking very good.
p.s. - I estimated the current ship loading now (the Dong A Glaucos) at 2.7 loading days.

View attachment 477951

Great stats on loading. Converting Q3 days to roro cars is about 1171. That doesn’t count Australia and other container delivery countries. The current 36 days would translate to 42,171. It’s likely the load rate will be different, but similar. Add at least one more ship, 3 days, and containers to Australia NZ and close to 50,000 cars exported. If USA is steady, it seems like 110,000 is more likely then 105,000. I think auto revenue is more likely 6.1-6.2 billion then the 5.8-5.9 consensus. If cash balance is up after the bond payment the short theme is really tarnished, since quarterly revenue is enough to fund debt payment and capex growth.

 

[kbM3]

I think the market definitely expects a record quarter in Q4 - otherwise you can't hit 360k for the year. So, the question for Q4 is one of margin.

For Q1, it is about deliveries - since that tells whether the Q3/Q4 being very good and Q1 being poor is a pattern that repeats from last year or not.

Once we cross these two quarters - it is looking pretty good with fewer uncertainties. Assumption being Tesla will manage the GF3 and Y ramp-ups much more smoothly than the 3.

View attachment 478007

Zach in Q3 conference call:

“In the immediate term, we're focused on increasing production of Model 3 and Model S and Model X as quickly as we can. ”

Don’t you think only 20k SX deliveries is conservative considering this statement? They also showed a current annual production capacity of 90k / year for SX at Fremont or 22.5k/quarter. Couldn’t this be a decent q4 upside surprise? Especially considering q4 deliveries tend to outpace production?