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

[Unpilot]

Thanks for posting these amazing video's.
I feel like I and most here are living in upside down world when I see the flat out amazing progress Tesla has achieved in the face of headwinds pushed by the idiot's on the likes of CNBC.

No mention of AP or amazing Quarter over Quarter growth.

I wonder if GF 4...or 5...or 100 will change the tune.
They have almost reached Kelly Ann Conway levels of alternative fact's.

 

[jhm]

Surprised there is so much pushback on this here. I love it. This one more that the pickup:

The dash is a beautiful work of art to my eyes:

However, I won't be buying.

Lego stylings. Definitely not my cup of tea.

 

[EVMeister]

Lego stylings. Definitely not my cup of tea.

 

[Pras]

Thanks for posting these amazing video's.
I feel like I and most here are living in upside down world when I see the flat out amazing progress Tesla has achieved in the face of headwinds pushed by the idiot's on the likes of CNBC.

No mention of AP or amazing Quarter over Quarter growth.

I wonder if GF 4...or 5...or 100 will change the tune.
They have almost reached Kelly Ann Conway levels of alternative fact's.

Kudos to Chinese folks as well. Respect.

 

[jhm]

This is illustrative of how quickly market share can be gained and lost near 50% penetration. Norway shift 10% points in 12 months. Imagine what a shift of that magnitude in the US or China would mean to the auto industry. Laggards will be crushed.

 

[Chunky Jr.]

I think summon will soon surpass the fart app as the feature that convinces the most people to buy a Tesla

 

[Fact Checking]

Probably a significantly lower SR+ mix in the US, due to SR+ backlog being used in up in Q2 and the new price structure promoting upsell.

For what little it's worth, Troy's survey says an ASP rise this quarter.

True.

Tesla can contrarily be expected to do what they've done every single quarter, which is significant COGS reduction. It's virtually inherent with increasing volumes.

But COGS improvements are logarithmic with volume improvements: Wright's Law estimates a 15% reduction in COGS for every 200% expansion of output.

If Q3 Model 3 production has indeed expanded from 72,500 to say 85,000, a 17% expansion, then that maps to a COGS improvement of about 1.3%.

Significant but certainly not as large of a volume leverage factor as in earlier phases of the Model 3 ramp-up.

A switch from pre-Ravens to Ravens (as has been ongoing this quarter) lowers FCF, not net income. Ravens somewhat increase ASP and significantly increase margins.

What works against this ASP increase is the decrease from having to significantly discount pre-Ravens made in 2018 or early 2019 - more so than in Q2.

I want to be optimistic about S/X sales, but the public data so far is sobering: a further drop in the Netherlands and Norway over Q2.

We also don't know yet whether the Nürburgring Plaid stunt improved or hurt September Model S/X sales... It was risky to show an non-camouflaged prototype, in terms of cannibalization.

I expect expansion this quarter, somewhere between "slight" and significant". Seasonal growth plus a new solar push. I also expect storage to continue its QoQ growth and margin improvement.

I agree that there will be significant expansion in sales and production this quarter - but not necessarily in revenue. The delivery report will tell us more.

I'm not sure what hiring rate increase you're referring to. Stores, service centres, and superchargers are not growing faster than deliveries. Superchargers in particular are in no rush whatsoever (seems to be in somewhat of a deployment lull while waiting for V3)

I was referring to this article:

Hiring at Tesla is up 33% since June

(Note that this is the new hiring rate, which increases opex only with a delay - but expanding service will also have more direct opex and capex effects.)

Extensive internal hardware was already on display in the Q2 letter. It would thus have to at least have already been registered in accounts payable, and thus, would not effect the net income picture this quarter. This quarter at GF3 has been about getting production hardware operational and getting fit right, not about installing said hardware.

Yet Q2 had a further reduction in capex over Q1.

In manufacturing it's customary to phase payments for new equipment based on multiple milestones such as delivery, demonstrated on-site function tests, and actual demonstrated production capacity.

I.e. even just "tests" and trial runs might automatically have resulted in meeting the first milestones and capex payments.

Or not - these details are not disclosed, but I wouldn't expect equipment manufacturers to allow Tesla to not pay for equipment for too long time periods, they have expenses and revenue targets too.

Edit:

Tesla has guided $1.5b-$2.0b capital expenditures for 2019, and has so far only realized a fraction of that: $280m in Q1 and $250m in Q2, which leaves $970m-$1,470m for the rest of the year.

If this is distributed roughly 50%-50% between Q3 and Q4, then that's a Q3 capex of $485m-$735m - a significant increase and a drain on Free Cash Flow.

 

[Fact Checking]

The rumored "Phase 2" construction work is a day-and-night effort as well, as can be seen from the floodlights:

​

Much progress with the Phase 2 foundations:

​

The white building in the upper right corner is the 220KV electrical substation.

One of the previous videos called Phase II the "Battery Workshop".

I find it interesting that they'd put the "Phase II" facilities closer to the substation than the rest of the factory - this suggests that it might have a substantial power draw - which cell making equipment would certainly have: ovens and fully automated high speed lines making millions of cells per day.

Or it could simply be a logistics optimization, not based on power draw: main factory was placed closer to the road network that supplies it with parts and employees. Substation would be as close to the incoming electrical lines as possible, regardless of the factory's power draw distribution.

 

[aubreymcfato]

Every West European country should be above Hungary.

That Italy is behind Bulgaria is ridiculous.

Yep.
Unfortunately we're still below the critical mass, you don't see many Tesla around, people are uninformed, they think it costs >100k, they joke about forgetting to plug in the car and being stuck in the traffic.
I think many will change their mind when they understand a 50k€ car is capable of coming out of a parking lot itself, and try AP, and finally get that range is a non-issue.
It will take time: but I'm curious to see latest numbers of Q3, I've seen some improvement.

 

[Artful Dodger]

Even with 100k deliveries Q3 revenues of $7b is IMHO way too optimistic: @luvb2b is estimating $6,480m in Q3 revenue with 99k deliveries.

Ahaha, that's an estimated 2% increase in revenue instead of FactSet's 1.5% increase. Fantastisch!

To go from $6.35B in revenue to $7.0B is the equivalent of delivering 13K additional Model 3s @ $50K ASP. So is that possible from the production standpoint? That would be about 85K model 3s, or about 944 each day over the 90 work days in Q3, vs 72K in Q2. Very possible IMO, indeed likely I think.

With 15K S/X, that's also about 100K total deliveries for the quarter. I think this is realistic.

Further, Tesla has stated it expects to claim about $800M in deferred revenue over the coming 4 quarters. If approx 1/4 of this can be claimed in Q3, then the $7B revenue level could be reached in at least these 2 alternate ways:

1. deliveries of just 81K Model 3s at $50K ASP (an increase of about 9K over Q2)
2. deliveries of 85K Model 3s at an ASP of just $47,650 (a decrease of $2,350/car)

Regarding pt 1, we know there has been a step-wise increase in production. Will those turn into deliveries? Logistics appear smooth, drama is low, hope is high; 4 days to go until we know.

Regarding pt 2, as @KarenRei commented above, Tesla is quite deliberate in the way they reduce sticker prices: we can be confident that across-the-board price decreases are the result of progress in production efficency, which allows Tesla to maintain their gross profit (not just margin), as we saw in the Q2 results.

I think neither of the 2 scenarios above is likely given the guidance provided by Tesla and recent email leaks. I think 100K total cars (incl. 85K Model 3s at ASP $50K) is our best central estimate. Further, with just 5 days remaining in the quarter upon receiving the last bit of information from email, variability is likely to be fairly tight.

How tight? Tesla delivered ~20K in the last 10 days of Q1 (choose a number). Thats a max deliver rate of about 2K cars per day. Tesla might not be flat out right now, but that's likely due to better planning and advancement on the learning curve. My WAG would be +/- 2.5K cars around that central estimate of 100K deliveries.

If you're worried about Shortzes high-balling the estimates so they can claim a 'miss', don't worry. NO MATTER WHAT, they will lie cheat and distort any achievement by Tesla. Its what makes them shortz. And they've paid with their souls (which apparently were not of any intrinsic value or use to their former owners).

If you're worried about Shortzes reading this forum and getting ideas, don't be. The Shorts sold their souls, not their brains. Besides, its the Financial Press reading this forum for ideas due to a lack of native ability (pardon @Curt Renz you know better than I how it is).

Then, we face 3 or 4 more weeks of SP manipulation and spin after the P&D report before the Q3 Earnings Call. Frankly, I don't really care what story the shortzes cook up in the mean time; I'm holding for 7-8 years and I have cash on the sideline to buy the dips.

I will watch with interest however to see Q3 FCF, although even there I have ZERO DOUBT that Tesla could raise another $10B or more in a week if they wanted to do so. Pick a number. Elon could get 10B in a fortune cookie if he wanted. TSLA is the only high-growth, big-cap equity available on US markets, and they're expanding into NUMEROUS separate trillion dollar markets (energy, storage, EVs, autonomy, trucking). They'd have domestic and international investors pounding on the door with cash if they asked.

But I don't think they'll need cash. They're growing as fast as their rapidly expanding engineering expertise allows, but no faster. Kinda reminds me of one of one of my favorite saying from Uncle Albert: "Everything should be made as simple as possible, but no simpler."

Einstein was making a statement about how to conduct science. Tesla is adapting his 1st principle of physiks to business and engineering: "Make everything as efficient as possible, but don't break the laws of physiks while trying." And remember "when you fail, next time you'll do better!"

Cheers! Prost! Gānbēi!

 

[KarenRei]

But COGS improvements are logarithmic with volume improvements: Wright's Law estimates a 15% reduction in COGS for every 200% expansion of output.

Not because of Wright's Law.

Wright's Law is based on the premise that you can justify automating more of your production, the higher the volume you get. But the Model 3 was designed for the beginning for automated production. The increases of production aren't simply from "duplicating lines", they're from getting lines to run faster and more automated, as was intended from the beginning. It's extra output with proportionally very little cost added.

Which is why Tesla's COGS reduction has been way faster than Wright's Law, and can be expected to continue as such.

Or to put it another way: see any new GA tents recently? Now how much production are they getting out of Fremont since their first quarter operating with GA4?

What works against this ASP increase is the decrease from having to significantly discount pre-Ravens made in 2018 or early 2019 - more so than in Q2.

More written down than Q2? I see no evidence of that. And again, unless there's an inventory writedown, inventory sales don't affect net income. Selling more Ravens - e.g. vehicles with higher margins and ASPs, however - has a dramatic positive impact on net income. And that's exactly what they're doing.

What I also see is a significant shift away from inventory to Ravens. Even in Europe. For example, yesterday in Norway, they delivered 37 dozen Model X. The number that weren't "Model X LR" (aka, Raven)? One (Model X 75D). Last quarter they were overwhelmingly inventory pre-Ravens.

I want to be optimistic about S/X sales, but the public data so far is sobering: a further drop in the Netherlands and Norway over Q2.

And up in the US, so?

Is this going to be a blowout S/X quarter? No, of course not. Looks to be a roughly even or slight growth quarter globally for S/X. But it's also predominantly a Raven quarter rather than an inventory quarter, at least from what can be seen. And while that's bad for FCF, it's good for net income.

We also don't know yet whether the Nürburgring Plaid stunt improved or hurt September Model S/X sales... It was risky to show an non-camouflaged prototype, in terms of cannibalization.

That's more of a Q4 issue, due to delivery times.

Hiring at Tesla is up 33% since June

​

And the breakdown says "Service & Energy Installation" is the top category, which undercuts the notion of a weak Energy quarter. Although I don't know the timing between hiring and actually getting to work, so it might just be anticipatory.

#2 is Engineering (R&D), which the market would actually like to see more of, and #3 is manufacturing.

Note that hiring being up by 33% does not mean # of employees up 33%, or even remotely close to that.

Yet Q2 had a further reduction in capex over Q1.

Tesla guided for capex being much lower for GF3. Apparently even longs didn't believe Musk when he said that.

In manufacturing it's customary to phase payments for new equipment based on multiple milestones such as delivery, demonstrated on-site function tests, and actual production capacity. I.e. installation and "tests" might automatically have resulted in meeting the first milestones and capex payments.

Maybe. But either way, given that they already have had to book at least part of all that production hardware, plus all of the Phase 1 building, I do not expect GF3 to contribute to a significant capex increase this quarter, and rather, expect a significant GF3 capex decline. Concerning MY, I'm more uncertain of that. I suspect that it's going to hit more in Q4 than Q3, because we're seeing lots of permits filed recently, and you can't build before filing permits. But I really don't know. GF1, capex has been dropping, and there's no sign that they've started adding new lines yet. But when they do, that will add more capex. Then again, who knows what they're doing in the background.

Edit: Tesla has guided $1.5b-$2.0b capital expenditures for 2019, and has so far only realized a fraction of that: $280m in Q1 and $250m in Q2, which leaves $970m-$1,470m for the rest of the year.

Interesting, what guidance are you looking at? Either way, as you note... that's FCF, not net income.

The short of what I'm seeing is that the differences vs. Q2, overwhelmingly, are adverse to FCF but favourable to net income.

 

[JRP3]

I think summon will soon surpass the fart app as the feature that convinces the most people to buy a Tesla

I'm afraid we're going to see a lot of parking lot collisions, mostly the fault of human drivers, but Tesla will be blamed. Here is a near miss where the other car was going too fast but the Tesla would probably been at fault if a collision had occurred: Roddie Hasan - راضي on Twitter

 

[larmor]

Watch the narrative flip on cnbc once the Shaghai factory starts producing cars. They will be falling all over themselves lining up for GF4 site and funding, ie their sponsors-investment banks etc. Factory as a product and investment with ONE year timeline is a lucrative investment and unreal timeline....

 

[Fact Checking]

Interesting, what guidance are you looking at?

The short of what I'm seeing is that the differences vs. Q2, overwhelmingly, are adverse to FCF but favourable to net income.

From the Q2 10-Q:

"Capital expenditures in 2019 are projected to be approximately $1.5 to 2.0 billion, to continue to develop our main projects as planned including further capacity expansion and automation for our current Model 3 production, Gigafactory Shanghai, Model Y and Tesla Semi, as well as to further expand our Supercharger and vehicle service and repair networks."​

$280m and $250m capex in Q1 and Q2 is from their financial reports as well.

Last year Tesla has reduced capex rather drastically - that will have to bounce back to support future projects, and just this summer, after Q2, in knowledge of their Q2 capex, in full knowledge of how fast GF3 was progressing, Tesla guided for another $970m-$1,470m capex for the remaining two quarters of the year.

So I expect Q3 capex to increase significantly over Q2, due to which FCF will drop but might still be positive. @luvb2b recently guesstimated $300m.

 

[ZsoZso]

Smart Summon may be a game changer. It's hard not to notice a car with no driver...

But it is also hard not to notice how much it hesitates and delays, slowing down other car movements in the parking lot. This is a double edged sword. General public may end up hating Teslas operating in Smart Summon mode just as much as they hate the slow traffic-hogging Waymo cars in their public test areas.

 

[AlMc]

The rumored "Phase 2" construction work is a day-and-night effort as well, as can be seen from the floodlights:

​

The white building in the upper right corner is the 220KV electrical substation.

One of the previous videos called Phase II the "Battery Workshop".

I find it interesting that they'd put the "Phase II" facilities closer to the substation than the rest of the factory - this suggests that it might have a substantial power draw - which cell making equipment would certainly have: ovens and fully automated high speed lines making millions of cells per day.

.

Ovens, we don't need ovens. (Said in the Back to the Future cadence of 'Roads, we don't need roads')

 

[Fact Checking]

Ovens, we don't need ovens. (Said in the Back to the Future cadence of 'Roads, we don't need roads')

Yeah, they won't need ovens under the Maxwell dry cell process - but they won't need ovens if these are "only" battery module assembly lines, and if the first couple of thousand battery packs come straight from GF1.

 

[Fact Checking]

Not because of Wright's Law.

Wright's Law is based on the premise that you can justify automating more of your production, the higher the volume you get. But the Model 3 was designed for the beginning for automated production. The increases of production aren't simply from "duplicating lines", they're from getting lines to run faster and more automated, as was intended from the beginning. It's extra output with proportionally very little cost added.

Which is why Tesla's COGS reduction has been way faster than Wright's Law, and can be expected to continue as such.

Or to put it another way: see any new GA tents recently? Now how much production are they getting out of Fremont since their first quarter operating with GA4?

Even under ideal scaling they'd be able to take advantage of levering out more of the fixed costs of Model 3 production, which are estimated to be around $200m per quarter.

That was a big deal when production went from 1k/week to 3k/week then 5k/week. When going from 6k/week to 7k/week the effect is about $360 per unit made, which is not zero but still only 0.7% of gross margin improvement at $50k ASP.

The rest of the costs of production are mostly variable: more line workers, more parts, more raw materials needed. Even with improving supplier price tiers I'd not expect more than around 2% of gross margin improvement over Q2, all other things equal. But not all other things are equal: any FSD recognition would be a tailwind, lower pre-Raven write-offs would be a tailwind as well, and there's the always mysteriously timed EV credits business.

(@Doggydogworld and @EVNow might want to correct my figures and my logic.)

 

[hobbes]

The Tesla not only had the right-of-way, but it was at a complete stop just before the impact. And it didn't do anything too abruptly. This illustrates how bad human drivers are overall.

Human drivers are not bad overall. However, there is a huge spread between good and bad drivers and even for one driver, depending on how tired etc. he is, there is a huge spread how well he will drive. I say 9 out of 10 humans driving the Tesla would have avoided this accident, even it the Tesla was less at fault than the car backing out. In Germany there is the concept of being partly guilty, and that would have applied here IMHO - is this a thing US jurisdiction, too?

 

[Christine69420]

It was nice to see some trailers parked at the loading docks, lights in the windows and a couple of buses parked outside. Looks like a working factory. And I wonder who those buses brough in?