Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

[KarenRei]

1-3 sec: show Cybertruck
4 sec: black screen, silver text: "ok, hummer"
5sec: Tesla logo.

Game, set, match.

They could do so much with those three seconds. Simultaneously have steel balls thrown at the glass (not by Franz! ), the doors shot, the car sledgehammered (maybe even a small wrecking ball), someone fire at it with a flamethrower, and a magnetic crane (being powered by the AC outlet in the bed) drop a tonne of steel scrap into the bed when it's unplugged.

If you can allocate a few more seconds, you could then have it climb a steep incline with a high approach angle and out of frame.

 

[BioSehnsucht]

They could do so much with those three seconds. Simultaneously have steel balls thrown at the glass (not by Franz! ), the doors shot, the car sledgehammered (maybe even a small wrecking ball), someone fire at it with a flamethrower, a crane dump a ton of heavy steel scrap into the bed, etc.

If you can allocate a few more seconds, you could then have it climb a steep incline with a high approach angle and out of frame.

Just split the screen N-ways and show all of it at once.

 

[KarenRei]

Just split the screen N-ways and show all of it at once.

I wasn't picturing a split screen - I was picturing all of that coming from all angles simultaneously

 

[ReflexFunds]

I'm posting here given its a weekend, but if you have comments on the model please reply in the Earnings Projections Thread: Near-future quarterly financial projections

Q419 Earnings Estimates:

P&L
Revenue $7,325m (+$1,022m QoQ, +$100m YoY):

Gross Profit $1,430m
Opex $970m
Net Income $269m (+$58m YoY, +$118m QoQ)
GAAP diluted EPS $1.5, Non GAAP EPS $2.6:

Key Revenue & Margin Assumptions:

In this model I have kept Model 3 like for like production cost flat QoQ. This excludes credit revenue, one off software upgrades (Acceleration boost), deferred FSD recognition and China one off ramp costs which have all been broken out separately.
I think flat production cost QoQ is likely conservative - Tesla has consistently reduced production costs QoQ & > production drives > margins (fixed cost leverage, < staff hours/car, supplier scale saving passed on). Main risks are larger one off GF3 headwind & more +ve one offs in Q

Cash Flow
Operating Cash Flow: $1,476m ($1,031m before Working Capital)
Free Cash Flow: $976m

My cash flow statements use different presentation & include many estimates for past Q line items but Operating Cash Flow & Free Cash Flow for Q4 uses Tesla definition.
My Cash Gross Profit Line = Gross Profit plus non cash cost addbacks for Depreciation, Warranty Reserve and Net Deferred Revenue

Net debt down $871m QoQ to $6,266m.
Net debt/EBITDA down to 2.6x from 4.4x at 4Q18.
Net recourse debt down more than half YoY to $1,765m from $3,702m at 4Q18. This includes $4,200m converts which are all in the money (but not convertible until 3 months before maturity).

Q4 Total Liquidity: $9,666m.
Including $6,233m unrestricted cash and $3,433m undrawn bank lines.
I would expect to see credit rating upgrades after Q4.

 

[ammulder]

I assume the development applications referencing the North and South paint shops mean 2 paint shops are operating, which I take as a proxy for more cars per week.

Like you, it has occurred to me that a Boring Company tunnel or 2 in and out of Fremont might be a good addition... perhaps the cost of funnelling is currently too high...

They will hit a limit when it is easier to start with a new factory elsewhere in the US.

I think the limit works the other way -- as soon as they hit a limit of another gigafactory being affordable, they might want to do it. In order to keep the exponential growth going past perhaps 2023, they will need to plant factories like weeds. As Fremont and Shanghai and Berlin approach their full capacity, there will need to be more factories hitting their stride. (Even three more factories of the 500K/yr to 1M/yr scale would only double max production over the current three, which gives them what, two or three years of this kind of growth? Good through 2025-2026? By then they need another six factories to double again? This exponential thing is crazy.)

 

[mongo]

Adding the speaker is probably not that hard (likely "not possible" only due to a reluctance to go through the effort to do so), but the easy alternative is to lower the windows and use the interior sound system.

Adding on the pedestrian warning speaker is only (easily) possible if the vehicle's current body module has the additional audio output. Otherwise, you are talking an additional module and wire harness updates to connect it.

1) The buildings on the northeast side that would have been in the way of the Y line? They're being demoed

Not so much demoed as moved to their next job. Those were stacked modular units serving as short term housing for construction workers. Densified mobile homes, if you will. In the Oct 18 video, you can see another set of them at the back of the site, along with the partly deconstructed units at the front.

 

[S3XY]

Hmm, aren't you low-balling that estimate just a tad? Phase 2 at GF3, and Phase 3 at GF4 are all scheduled to be complete by 2023, so thats:

• Fremont: 875k
  • M3: 7k/wk * 50 wks = 400k
  • MY: 7k/wk * 50 wks = 400k

Hmm, I don't think that math checks out.

Baker's dozen.

 

[defc0n]

Perhaps a nothing burger but not good optics for nuclear power again: Crews respond to ‘incident’ at Pickering Nuclear Generating Station

 

[KarenRei]

I'm posting here given its a weekend, but if you have comments on the model please reply in the Earnings Projections Thread: Near-future quarterly financial projections

Q419 Earnings Estimates:

P&L
Revenue $7,325m (+$1,022m QoQ, +$100m YoY):

Gross Profit $1,430m
Opex $970m
Net Income $269m (+$58m YoY, +$118m QoQ)
GAAP diluted EPS $1.5, Non GAAP EPS $2.6:

Key Revenue & Margin Assumptions:

In this model I have kept Model 3 gross margins flat QoQ. This excludes credit revenue, one off software upgrades (Acceleration boost), deferred FSD recognition and China one off ramp costs which have all been broken out separately.
I think flat margin QoQ is likely conservative - Tesla has consistently reduced production costs QoQ & > production drives > margins (fixed cost leverage, < staff hours/car, supplier scale saving passed on). Main risks are larger one off GF3 headwind & more +ve one offs in Q

Cash Flow
Operating Cash Flow: $1,476m ($1,031m before Working Capital)
Free Cash Flow: $976m

My cash flow statements use different presentation & include many estimates for past Q line items but Operating Cash Flow & Free Cash Flow for Q4 uses Tesla definition.
My Cash Gross Profit Line = Gross Profit plus non cash cost addbacks for Depreciation, Warranty Reserve and Net Deferred Revenue

Net debt down $871m QoQ to $6,266m.
Net debt/EBITDA down to 2.6x from 4.4x at 4Q18.
Net recourse debt down more than half YoY to $1,765m from $3,702m at 4Q18. This includes $4,200m converts which are all in the money (but not convertible until 3 months before maturity).

Q4 Total Liquidity: $9,666m.
Including $6,233m unrestricted cash and $3,433m undrawn bank lines.
I would expect to see credit rating upgrades after Q4.

I think if you post the whole thing here, we should be able to offer basic responses as to whether we agree, no? That said, I don't see much to complain about. Perhaps a bit more Model 3 ASP hike, from the combination of both the $400 17 Oct price hike, the $1k 1 Nov FSD hike, and the improved mix in Europe... but only a bit. Not really sure how to estimate margin movement in Q4 because I'm not sure how GF3 will be treated with respect to COGS. Production is up overall, which helps COGS, but not tremendously (though cell output may have improved more than overall automotive, driving more COGS reductions there). No dramatic raw material price swings - nickel and cobalt sulphate a bit up, but not dramatically. Li carbonate/chloride/hydroxide still slightly trending down.

Energy sales, I'm a bit more optimistic, but not as optimistic as I'd previously been. We've been seeing signs of quite strong sales (and seemingly, corresponding production) on grid-scale products, and powerwall and conventional solar definitely seems to be up, but I haven't seen as much signs of solar roof ramp as I had wanted to see. Still, signs all point to solid TE growth. Margins might be a bit down due to discounts and solar roof ramping, however.

All in all, seems more or less plausible.

 

[winfield100]

Just have the car ask riders to face the cameras for biometric verification.

@BioSehnsucht
A good beta test for early versions of neuralink/neuralace

 

[JRP3]

Yes - Agree the earnings report is unaudited.
I worked at one of the big Public Accounting Firms (not PwC but one of the other recognizable firms). So this is my experience:
PwC signs off on the earnings press release. The reason for this? No one wants to see the Earnings Release financial numbers not match the 10K numbers - which will get released 2 weeks later. So even though the ER is not audited, PwC will sign-off and for this their audit needs to be substantially completed.

One of the bear talking points has been "Just wait for the 10k" as if that's going to expose something previously hidden.

 

[Fact Checking]

Not really sure how to estimate margin movement in Q4 because I'm not sure how GF3 will be treated with respect to COGS.

According to @The Accountant the impact of Q4 GF3 production and deliveries on margins is probably going to be negligible, because the 15 GF3 cars delivered in Q4 have a CoGs impact of less than $1m.

There might be a small increase in SG&A ($2m-$3m max), but SG&A isn't even included in margins.

So Q4 margins almost entirely depend on trends in Fremont.

 

[KarenRei]

One of the bear talking points has been "Just wait for the 10k" as if that's going to expose something previously hidden.

It's funny how obsessed they are with 10Ks Because it's all a big fraud / smoke and mirrors, dontchaknow, and they're going to SURELY find something in the 10k to prove it!

 

[Fact Checking]

It's funny how obsessed they are with 10Ks Because it's all a big fraud / smoke and mirrors, dontchaknow, and they're going to SURELY find something in the 10k to prove it!

"Wait for the 10K" is also a delaying tactic: the annual report is more complex and takes longer to prepare, because there are external auditors who'll be signing off on it.

 

[JRP3]

It's funny how obsessed they are with 10Ks Because it's all a big fraud / smoke and mirrors, dontchaknow, and they're going to SURELY find something in the 10k to prove it!

As soon as the auditors go over everything!!!111Fwauuudd!!11

 

[kbM3]

This has been discussed pretty extensively. Everything is optimized to its particular use case. There's under $2k manufacturing hardware depreciation per unit on Model 3 (we were told by Deepak it was under $2k nearly a year ago), so the potential capital savings are limited. You hurt your drag coefficient compared to current Teslas with the "cyber" approach. Musk thinks that with "great effort" Cybertruck might hit 0,30; by contrast, Model 3 is 0,23. So probably a bit over 0,30 - probably approaching a 50% worse Cd than Model 3. A "cyber-Model-3" would probably end up with ~30% more energy consumption at high speeds. Possibly worse, as it won't be able to have as long, shallow of a rear taper as Cybertruck to encourage flow reattachment, and/or a larger rear cross section after truncation (although the lack of the need for a bed may allow for some improvements on that end).

More energy consumption per unit distance equals a worse vehicle in so many different respects. Home charging cost, supercharging cost, supercharging infrastructure (less efficiency = more time charging = more stalls taken up), range (for a given battery size), mass (if battery size is increased to compensate), handling (if more weight, e.g. if more batteries), max production capacity (if battery size increased), # of cycles put on the packs per unit distance (if battery size not increased), depth of average daily cycle (if battery size not increased), charging taper hit at an earlier # of miles (if battery size not increased) and on and on.

The main thing that allows Tesla to make Cybertruck awesome for its price point is simply that it's two years off, and its batteries are priced at their estimated prices two years from now. Cybertruck's design is the result of relentlessly optimizing to a particular goal. They wanted to have it armoured. That means something heavy on the outside. Which you can't stamp. So, to keep it light: origami frame, which equals polygonal. And sure, sometimes things that are invented for one particular goal end up proving useful toward other goals. But it's not so simple. Efficiency matters a lot. Its importance decreases slowly with time as batteries get cheaper and more abundant and more energy dense, but at present it's still very important. By and large, it's Tesla's efficiency advantage that tends to give its vehicles so much better stats than everyone else, which in turn translates to its perception as an EV sector leader.

Also, for a truck, they're already pretty doomed when it comes to efficiency regardless. You could put them on low rolling resistance tires and give them a tapering tonneau and a beautiful aero profile, and your new owners will immediately put mud tires on them and draggy external accessories and drive around with the tonneau up and objects sticking out of it and tow a draggy trailer and on and on. Speaking of the latter: I haven't seen a simulation yet, but I have a strong suspicion that if you tow a tall trailer with the Cybertruck, the flow detachment at the pointed peak of the Cybertruck won't reattach, but will rather flow in a bubble over the trailer, and recirculate between the two (good for the combined drag coefficient of the truck-trailer system). If you tow with something like the Model X, you're basically trying to stick the flow to the back of a smooth, gentle rear taper only to ram it straight into a vertical wall that you're dragging behind it.

Back to cars: as discussed previously, if the goal is simply "cyber-aesthetics", you could create some rather interesting smooth-polygonal hybrids. Parts of the vehicle that already have flow detachment / turbulent flow face little impact from "cyberization", so you can keep the car smooth where important and "cyberize" other parts of it, and blend the two together, for what would probably be quite an interesting aesthetic.

My understanding is that CT’s advantages go beyond just saving fixed cost.

Take the Alien Dreadnought perspective and Elon’s insight that “The best part is no part and the best process is no process”:

Thought of in this way, the paint shop removal makes GFs cheaper, smaller, more productive and quicker to build.
Replacement of the giant stamping machine(s) with Origami Folders may give some of the same advantages?

I would also guess the bulletproof capability is just a happy result of the above, plus making the Exoskeleton strong enough for better than F150 payload and towing capacities even after removing the frame.removing the frame.

 

[SOULPEDL]

Just split the screen N-ways and show all of it at once.

No drama, just all data, so Karen.
Gotta build emotion... in 5 sec, lol.

 

[Zaxxon]

It's funny how obsessed they are with 10Ks Because it's all a big fraud / smoke and mirrors, dontchaknow, and they're going to SURELY find something in the 10k to prove it!

...this time. Amazing how those folks keep failing yet expecting that next time is gonna be the one where it all falls apart.

 

[KarenRei]

BTW, since it's a slow day, and a weekend, a minor OT. I was in the middle of having my mind fried when I discovered the concept of "real-world cel shading" - painting things (or even people) matte with stark colour contrasts and solid lines at creases, to make them look like they're from a video game or cartoon:

.... and I discovered that someone had done this to an early Model S with a vinyl wrap!

Too bad the non-wrapped parts kind of ruin the illusion. Still, that's a really neat concept if anyone's looking for a stand-out look for their car

 

[GaryW]

More energy consumption per unit distance equals a worse vehicle in so many different respects. Home charging cost, supercharging cost, supercharging infrastructure (less efficiency = more time charging = more stalls taken up), range (for a given battery size), mass (if battery size is increased to compensate), handling (if more weight, e.g. if more batteries), max production capacity (if battery size increased), # of cycles put on the packs per unit distance (if battery size not increased), depth of average daily cycle (if battery size not increased), charging taper hit at an earlier # of miles (if battery size not increased) and on and on.

All of this is true when comparing a model 3 to a model y, although the difference is less. For these reasons, it’s harder for me to guess just how much model y will really cannibalize model 3 sales. My model S is bigger, more comfy, and has free supercharging. I still have to think each time we go on the road. The model 3 charges so much faster because it goes so much farther on a KW of electricity. The model 3 is truly an amazing vehicle.