Near-future quarterly financial projections

[schonelucht]

I'm wondering how Deepak's comment about collecting cash from consumer loans should be interpreted. Did he only make a cash flow argument?

Correct. Payables and receivables are (future expected) cash flow items on the balance sheet.

Btw. I suspect your original confusion was due to trying to find reason in shorts. Don't do that. Very few of them are sound but most are even more unqualified and talking even more nonsense than the bulls on this board. With the present quarter behind us, the only way the majority of their theories might work would be if Tesla engaged in outright accounting fraud. (Ok, let's leave a little leeway until we've seen the full 10-Q next week or so to accept that the numbers are really as good as they seem to be from the information released so far). For example booking revenue from a fleet sale of loaners but not booking corresponding cost of goods would be such a fraudulent practice. But because it would have been fraudulent, I makes no sense to reason from that theory with correct accounting rules. Garbage in, garbage out principle.

Reminder it is not impossible that Tesla committed accounting fraud. There are plenty of companies that did just that. But the likelihood is small, very small. Small enough that I don't factor it in my valuation for sure.

 

[EVNow]

You are right, thanks for the correction!

I'm wondering how Deepak's comment about collecting cash from consumer loans should be interpreted. Did he only make a cash flow argument?

When Tesla sells, they increase receivables. That gets converted to Cash once the bank clears. In Q3 since the month end was a weekend they had a lag in clearance, so even though there was no difference in deliveries or profit, the receivables were higher and cash equally lower.

 

[Fact Checking]

tw. I suspect your original confusion was due to trying to find reason in shorts.

Well, the main problem was that I had a major misunderstanding of exactly how and when sales get accounted as revenue, and when @brian45011 first corrected me I doubled down on my flawed thinking. So for my boneheadedness I got some well deserved fact checking and ... hardcore smackdown from @brian45011, quoting the 10-Q.

Garbage in, garbage out principle.

In this case the core of my argument was garbage too - I cannot blame shorts for that.

Anyway, it appears that this particular '1,000 fleet sales' bear/short argument is the following:

• Income from used cars is not accounted as automotive, but as services, which have bad margins anyway and can be used as 'dump bad stuff' business segment to make other parts of the business shine. (BTW., accounting used car sales into new car automotive business segment is entirely fair IMO, they don't want to mix the margins of cars made years ago into the margins of newly made cars.)
• So by 'dumping' 1,000 Model S's Tesla was able to artificially increase their automotive gross margins by not burdening the automotive side with bad COGs, so this particular bear argument goes.

I think that argument is still flawed (if I understood it correctly and applied the accounting rules correctly), because if I'm reading Tesla's rules correctly there's no way for a new car to become used without it either being either written down (for example scrapped if they find an un-fixable frame weakness during testing), or being first sold/leased and then re-added to inventory as a used car. Both of those actions carry CoGs recognition for the new-automotive segment.

I.e. I don't think there's any accounting loophole Tesla could make use of here to artificially improve automotive margins.

Nor would they have any motivation to do so even if we assumed that Tesla's executives all inherited Al Capone's genetic traits: the improvement by these 1,000 used cars is still small and the disadvantages of jail time or being disbarred for life are ... large.

$250m profit and $700m cash flow would still have been a hit-the-ball-out-of-park quarter than the $312m+$881 quarter they reported.

So even under the best (well, worst) assumptions I cannot find any internal consistency in the bear thesis, other than totally dumb accounting fraud and other criminal activity for no good reason whatsoever. (There's also the small matter of the annual external audit of Tesla's finances - who I believe would certainly not sign off on such practices.)

So this is new to me: usually bears/shorts at least try to fake some sort of internally consistent rational arguments - which is useful to understand as a reality check against bullish thinking. They seem to have largely given up that pretense lately.

 

[Analyst007]

Is there already a new model from luvbb for Q4? Would be interesting to see the assumptions. I would agree on a similar conservative approach for estimating cost and income. So there is more room for positive surprises

I guess we should estimate that margins for Model 3 stay more or less the same (a little above 20%). Difficult will be to esimate the raise in SG&A (as Elon is opening a lot more service centers in Q4) as well as the assumption of how many cars will be delivered in Q4, as there is no clear guidance of how many cars per week they are producing.

 

[neroden]

Service centers take time to open and it's already November. I expect very few service centers to open in Q4. We should see some capital expense as sites are purchased, and a little SG&A as people are hired and trained, but I expect the big boom in service centers and the according boost to SG&A to be in Q1 2019 at the earliest.

 

[Webeevdrivers]

Sorry guys. Educate a foreigner. What is a 10-Q

 

[neroden]

Quarterly report required to be filed with the SEC

 

[Robertj]

Sorry guys. Educate a foreigner. What is a 10-Q

Take a look at Tesla website
Investors quarterly reporting
10 q is report required to be filed at end of each quarter with set detailed financial information

 

[adiggs]

Sorry guys. Educate a foreigner. What is a 10-Q

SEC Form 10-Q

As distinguished from a 10-k (the annual report).

 

[brian45011]

"During the third quarter of 2018, we made the decision to utilize some of our fleet cars as service loaners on a long-term basis. As a result, we reclassified $72.8 million of finished goods inventory to property, plant and equipment."

I think this change added the $72.8 million to Cash From Operations and is a one time benefit for that metric.

 

[neroden]

"During the third quarter of 2018, we made the decision to utilize some of our fleet cars as service loaners on a long-term basis. As a result, we reclassified $72.8 million of finished goods inventory to property, plant and equipment."

I think this change added the $72.8 million to Cash From Operations and is a one time benefit for that metric.

That would make sense; it would also be a capital expenditure.

 

[Buckminster]

Is there already a new model from luvbb for Q4? Would be interesting to see the assumptions. I would agree on a similar conservative approach for estimating cost and income. So there is more room for positive surprises

I guess we should estimate that margins for Model 3 stay more or less the same (a little above 20%). Difficult will be to esimate the raise in SG&A (as Elon is opening a lot more service centers in Q4) as well as the assumption of how many cars will be delivered in Q4, as there is no clear guidance of how many cars per week they are producing.

I second this - Deepak is desperate for it as Elon is pushing on whether they can afford to kick-off more projects... I would prefer to be as accurate as possible rather than conservative. Its the principle of it really. Sandbagging is toxic and leads to a "no surprises culture" which inherently stops innovation dead. Over optimism is Elon's Achilles heel but also what has made him.

 

[Singer3000]

The surprise was that Model 3 margins hit 20% as quickly as they did after Q2. Despite tariff headwinds, I’ll predict that they nudge closer to 25% this quarter than 20%, with aggressive volume growth foregone for margin. Deliveries will be “higher” but who knows how much higher. Room for some additional SG&A in any case.

 

[Singer3000]

Capex

From Q3 10Q:

"We currently expect 2018 capital expenditures slightly below $2.5 billion, including the acquisition of land use rights in Shanghai, China to build Gigafactory 3 and for its initial design and construction expenses".

Fremont Model 3 production rate to "approximately 7,000 units per week with only limited additional capital expenditure.. beyond 7,000 units per week with incremental capital".

"We currently estimate that capital expenditures will be between $2.5 to $3.0 billion annually for the next two fiscal years".

And from the Q2 call:
"confident that we can do the Gigafactory in China for a lot less... just a guess but probably closer to $2 billion...and that would be sort of at the 250,000 vehicle per year rate"
---
So fiscal years 2019 - 2020, there is $5-6bn to play with:

• c. $2bn likely ear-marked for Shanghai on accelerated time frame,
• call it $500m for Model 3 Fremont?
• incidental (<$200m?) for service and superchargers.

All of that buys you auto sales of 100k X&S, up to 500k Model 3 at Fremont and up to 250k in China (presumably all Model 3 in this time frame). This is the "somewhere between 700,000, 800,000 seems pretty likely" that JB referred to for 2020 production target on the Q2 call.

Which leaves around $3-3.5bn tops of capex to deploy for Model Y, Semi (with initial megacharging), Giga2/solar tile and then smaller projects of Roadster and the F650 Pickup competitor (likely a by-product of Semi programme in my view).

Feels a little tight but benefit of the doubt and efficiency learnings maybe means that's enough for those projects. But it doesn't leave much room for error, much less for production from a European factory or a mass produced Pickup much before 2022/23 does it?

 

[Fact Checking]

• c. $2bn likely ear-marked for Shanghai on accelerated time frame,

Much less I'd say: $750m from Tesla and $1.5b from Chinese state banks.

call it $500m for Model 3 Fremont?

Less I'd say: 8k was estimated at "minimal" capex, and 10k with "incremental" capex. I'd be surprised if it went beyond $250m: the biggest item is probably that they need about ~20% more paint booths.

incidental (<$200m?) for service and superchargers.

I'd say more.

In 2019/2020 I'd expect capex to be dominated by expansion costs:

• Model Y expansion at 500k/year capacity. $2b with clever engineering?
• Tesla Semi - maybe $500m initially, with lower levels of automation.
• Cell manufacturing expansion: they have to pay Panasonic.
• Also there might be residual Model 3 expansion capex outflow in early 2019.

Also note that their capex guidance range is roughly at 8x their Q3 free cash flow of $881m minus long term debt repayments for FY19-20, which is 8x$881m - $920m = ~$3b annual.

I.e. I'd expect them to be profitable but reinvest almost all FCF.

This also suggests that the $2.5-3.0b range and guidance is not cast into stone: they'll adjust the capex guidance if Q4 or Q1 provides lower or higher cash flow - to actively maintain positive cash flow in every quarter going forward.

 

[Singer3000]

The capex guidance is what it is. I’m taking as a given that they’ll find a way of paying for the $2.5-3bn annual capex they are forecasting (i.e. operating cashflow, debt or fresh equity). Just because part of Shanghai will be paid for by debt, doesn’t mean it gets stripped out the capex forecast.

What do “incremental” and “minimal” mean? How many more lagging capex payments are there for existing Model 3 production capability? We don’t have much beyond guesswork for what remains. Bit disappointing.

Panasonic: I might be wrong but I was under the impression that the capex for cell production was incurred by Panasonic and this was stripped out from Tesla’s cashflows and instead incurred as they consume cells and shown on the balance sheet as Fixsd Asset versus Lease lease liability? There’s a detail in the notes beginning several quarters ago that shows how much Panasonic has spent to date. If this is not stripped out, that leaves even less within the forecast capex budget for other projects.

I guess my point is, the more excitable forecasts for production volumes in 2021/2022 should probably be tempered by this $5-6bn capex guidance, as it doesn’t leave room for much incremental production capacity beyond smallish Semi programme, 250k in Shanghai and 250-500k Model Y.

Unless of course, these are deliberately low balled numbers. Management could reasonably take the view that since the market is underestimating medium term cashflows, it would be scared by a 2-year capex number of say $8-10bn. So the true pace/scale of expansion will only become clearer as debt is paid doubt and cash balances tick up in the coming quarters.

 

[EinSV]

Much less I'd say: $750m from Tesla and $1.5b from Chinese state banks.



Less I'd say: 8k was estimated at "minimal" capex, and 10k with "incremental" capex. I'd be surprised if it went beyond $250m: the biggest item is probably that they need about ~20% more paint booths.



I'd say more.

In 2019/2020 I'd expect capex to be dominated by expansion costs:

• Model Y expansion at 500k/year capacity. $2b with clever engineering?
• Tesla Semi - maybe $500m initially, with lower levels of automation.
• Cell manufacturing expansion: they have to pay Panasonic.
• Also there might be residual Model 3 expansion capex outflow in early 2019.

Also note that their capex guidance range is roughly at 8x their Q3 free cash flow of $881m minus long term debt repayments for FY19-20, which is 8x$881m - $920m = ~$3b annual.

I.e. I'd expect them to be profitable but reinvest almost all FCF.

This also suggests that the $2.5-3.0b range and guidance is not cast into stone: they'll adjust the capex guidance if Q4 or Q1 provides lower or higher cash flow - to actively maintain positive cash flow in every quarter going forward.

I think the calculation should start with operating cash flow, not free cash flow, since the $800M in FCF already has CapEx deducted from it so you are effectively deducting twice for CapEx.

Tesla had $1.4B in operating cash flow in Q3. Even with no growth over the next two years (a bad assumption) that equates to $5.6B per year, or $11.2B over two years to fund $5-$6B in projected CapEx. The rest can be used to pay down debt and add to cash reserves and shore up the balance sheet.

And with increasing revenues, improved gross and operating margins and an improving cash generation cycle (which Elon has discussed at length) operating cash flow seems likely to be even higher — much higher.

 

[EinSV]

I guess my point is, the more excitable forecasts for production volumes in 2021/2022 should probably be tempered by this $5-6bn capex guidance, as it doesn’t leave room for much incremental production capacity beyond smallish Semi programme, 250k in Shanghai and 250-500k Model Y.

There will be additional CapEx in 2021 and 2022 for European GF, China expansion, Model Y and Semi ramps, the Pickup, Solar Roof, etc. With cash generation continuing to increase due to improved margins and an improved cash generation cycle Tesla should have plenty of cash to keep investing at the same level as 2019-2020 or higher.

Tesla’s CapEx projections for 2019-2020 seem consistent with continued 50%+ revenue growth per year for the foreseeable future.

 

[Singer3000]

There will be additional CapEx in 2021 and 2022 for European GF, China expansion, Model Y and Semi ramps, the Pickup, Solar Roof, etc. With cash generation continuing to increase due to improved margins and an improved cash generation cycle Tesla should have plenty of cash to keep investing at the same level as 2019-2020 or higher.

Tesla’s CapEx projections for 2019-2020 seem consistent with continued 50%+ revenue growth per year for the foreseeable future.

And maybe that’s right. But clear that things have changed from 18 months ago:
“I will announce locations for between two and four Gigafactories later this year – probably four.”

That’s just not possible with these capex forecasts.

The future we're building -- and boring

It does drive me slightly mad that no analyst asks for detail on the 3-year capex and growth plan and instead we get stupid questions about Terminator cars and customer deposits.

 

[EinSV]

And maybe that’s right. But clear that things have changed from 18 months ago:
“I will announce locations for between two and four Gigafactories later this year – probably four.”

That’s just not possible with these capex forecasts.

The future we're building -- and boring

It does drive me slightly mad that no analyst asks for detail on the 3-year capex and growth plan and instead we get stupid questions about Terminator cars and customer deposits.

The four could be:

China
Europe
Model Y
Semi/Pickup

Many now seem to think the Semi and/or Model Y will be built at GF1 — we’ll see. Other than that and the schedule moving back six months or so (Elon time) I don’t see any significant change in the plan. Elon doesn’t discuss details, which is understandable given how freaked out people were over the Model 3 ramp. And as you say analysts don’t seem that interested in asking.

Next year we will probably find out locations for Y and Semi builds and possibly the European GF.