General Discussion: 2018 Investor Roundtable

[Matias]

I see that even though you’ve gone long (for now), you still don’t understand what is happening nor how it’s likely to progress. But this certainly does explain why you clung to your short position so long.

What you’ve laid out in this post and the previous is not what’s going to happen.

No. I am unable (incapable/not skilled enough) to explain it to you but perhaps someone else will tackle the challenge and have the right combination of words for you.

I have to say, your discussion style is a bit disturbing. It might of course be, that I get your tone wrong.

 

[aubreymcfato]

Posting the VIN numbers has become a full-time job:
Tesla registered 3,609 new Model3 VINs. ~82% estimated to be dual motor. Highest VIN is 92929.
Source: Model 3 VINs on Twitter

 

[JRP3]

I have to say, your discussion style is a bit disturbing. It might of course be, that I get your tone wrong.

I think you have it correct.

 

[SebastianR]

Is there historically any example, where customers have stopped buying some product (for years perhaps), because they know, that there would be better product and waited for that?

If you don't want to go through the tech-sector, let me try to convince you with an automotive argument: We all know that car sales (and the most recent losses of the incumbent car makers to the Model 3) should not be compared month to month but only when respecting the respective car's place in the model cycle, right?

Autotrader - page unavailable

Resale values are another good reason to pay attention to model cycles. While a car's price usually doesn't change much following a redesign, a new model tends to hold its value a lot better than an outgoing version. As an example, consider the Ford Fusion, which was redesigned for the 2013 model year. While the base-level Fusion's price stayed roughly the same between 2012 and 2013, used 2012 Fusion models are available for an average price of around $15,300, which is a far cry from the $19,400 average price of 2013 models.

The reason for this price jump is that the latest body style is almost always more desirable than the outgoing version, so it will hold its value better. If you had known in 2012 that the Fusion was about to be redesigned, you could have waited for the new version and dramatically improved your resale value in the process.

This argument goes 1:1 for EVs. But even more importantly this also goes cross-models/car makers. A consequence of the above is that a "2015 -newer in the model cycle car of car maker A" is retaining more value than a "2015 - older in the model cycle car of car maker B".

So can we test this hypothesis? I think we can: There are analysts who looked into this. If I'm correct the Model S would need to retain its value really well (provided it is a well built car that doesn't break after 50k miles etc. etc.)

So then there is this:
Tesla Model S retains its value better than gas-powered cars in its segment, losing only 28% after 50k miles

Lastly (and totally anecdotally) there is this strange guy called Sebastian (me) who owned and bought German cars for the last >20 years of his life. Instead of plunking down money for the next ICE, he is holding out on his 2012 Audi until the Model 3 is coming to Europe...

 

[avoigt]

Nissan selling the Battery business:

"Nissan improved its batteries, but a study showed that the larger 30-kwh batteries in 2016 Leafs had even higher rates of degradation than earlier 24-kwh batteries.

For its latest 40-kwh batteries, Nissan developed a new nickel-manganese spinel cell that is designed to reduce capacity loss.

Nissan plans to introduce a new, longer-range Leaf e-Plus for 2019 that will use 60-kwh batteries supplied by LG Chem.

Nissan sells its battery business, 225-mile Leaf e-Plus to get LG Chem batteries

 

[abbass2]

But wait, there’s more.

500k vehicles annually in Shanghai. That should hit steady state between 2023-2025. I don’t know the mix of vehicles, but if the average price is $40k, that’s another $20B annual revenues. $100B total.

Roadster should be steady by then, but low volume. 8k/year is $2B, 20k is $5B. Your guess is as good as mine.

If Tesla Energy grows at just 50% per year, it should be at $12.7B by 2023. Services & Supercharging should provide at least another $10B in revenue, but no earnings.

So that’s somewhere around $130B in revenue by 2023 or so. Without consider the Pickup, the Semi, or Tesla Network.

Thanks for the great analysis. What are your thoughts on profit margin? I'm assuming 20 - 30% so at an average of 25 %, this would represent $32.5B in earnings?

 

[Foghat]

According to IBM, exoflop in 2020 is going to need 100 MWs+ of electricity to work. The human brain can do a ~ pentaflop on 10s of watts.

The human brain is 1m:1 more efficient.

The nexus between Tesla’s neural network &chip hardware with vehicle and energy production/storage business can not be overemphasized.

The auto and solar/storage data they have already collected is a significant advantage to developing effective predictive outputs.

To imagine when the spacex sattelite network gets up and running, the sky is literally the limit.

 

[winfield100]

[QUOTE="mrdoubleb, post: 2929190,
At the same time, economies of scale will continue to favor Tesla, as in 50%+ of EV cell production being in 1 hand.
As for the solid state battery breakthrough: waiting for the magic fairy dust is a big gamble. Sure, it may pay off, but I would imagine Tesla is not sitting idle either. What`s the guarantee VW gets there first? What`s the guarantee even if they bring it out first it will be competitive with Tesla`s 2025 cell technology? What`s the guarantee Tesla can`t replicate that technology quickly without braking patents? It`s not like the entire Gigafactory cost goes down the drain if they have to switch to solid state.[/QUOTE]
I've been doing simple calculations
My internet crashed (again) and this is from faulty memory.
In 2017 Tesla had around 2/3 of battery capacity installed in EV's (excluding China), around 4.5-4.8 gigawatt hours.
In 2018, through July, 7/12ths of the year, Tesla had around 74% (up 8%) of battery capacity installed in EV's, heading using simple linearity, and ignoring the 3 more lines that will increase output another 30%, so at least double of around >9 gigawatt hours, factor in another 30% and well over 12gWh and probably much more as lines capacity etc accelerate.
I continue to suggest it's an S curve, more vertical, then flattening in terms of Tesla making batteries.

 

[Mader Levap]

In any case, in the best case scenario Tesla sells 1% of new cars after few years. So 99% are sold by traditional manufacturers. So it is mathematically impossible, that they’re all too late.

That would require two things:
1. Tesla never will sell more than 1% of new cars. In other words, their growth will stop.
2. Everyone else will grow unimpeded.

To be clear, I talk about EVs in both cases. No one here cares about ICEs.

I find point 1 exceedingly unlikely. I also find betting on traditional manufacturers growing faster than Tesla risky, but certainly more realistic.

In fact, if Tesla remains premium brand selling only luxury cars (yes, Model 3 is "entry-level" luxury car that majority of people never will be able to afford), they will be minority in EV market sooner or latter. I doubt they would mind that, both because of mission (EVs everywhere) and fact that premium brand means very nice profits, despite small volume.

But right now? There will be years of growth before they hit wall.

 

[winfield100]

@Matias

I have to say, your discussion style is a bit disturbing. It might of course be, that I get your tone wrong.

Look at a fossil fuel car, any one.
Extremely complex dance of moving parts.
Camshafts, lifters, valves sucking and blowing air, radiators to keep it cool enough not to melt, pumps and hoses to lubricate so there is just enough but not too much friction. (You would be amazed at hot an engine can get without coolent)
Vs a big battery and an engine the size of a watermelon. Simplicity on just the first or second iteration. The technology has caught up with the idea and instantiated it.

 

[jasonb]

Maybe Los Angeles is atypical but a rather large percentage of Tesla sales.

is true? i hardly see any when visiting down there. but LA is bigger than bay area i suppose. i would have thought never having to sit in a hot car would be enough alone to get wide adoption...

 

[RobStark]

is true? i hardly see any when visiting down there. but LA is bigger than bay area i suppose. i would have thought never having to sit in a hot car would be enough alone to get wide adoption...

True.

San Jose Bay Area and Greater Oslo have even more disproportionate Tesla sales.

Kinda like LeBron James telling Kevin Durant he is not a very big guy.

 

[FirebirdAlpha]

When will the Lathrop lot be of concern to the bulls?

New video this morning shows a few things:

1. Still no outdoor lighting, which means they can't do distribution at night.
2. Continue pile-up of cars on the East end. It looks like most of the cars on the East side haven't moved.
3. Construction of a new, large lot South of the lot.
4. Tesla appears to be using a new parking lot for another 300+ cars.
5. The logistics situation looks horrible. The cars are really packed in there now.

In total, Lathrop is now up to 3800 cars outdoors (and likely a lot more indoors). That's up from 2700 just 2 weeks ago.

That's a week's worth of production that doesn't appear to be moving quickly.

 

[SteveG3]

These market share projections and how incumbents will continue to sell ~100 millions of ICE cars per year have one very basic error:
They assume people will totaly ignore there are EVs and will continue to buy ICE in exactly the same numbers.

This will not happen, ICE sell will fall of the cliff. Total number of cars sold in a year will fall off a cliff.
If there is not enough EVs, people will rather wait a little more, save a little more and then buy a better EV.

1M tesla EVs might mean 2% of total market.
And 10% of total profit.

It makes sense that we would see this, or some of this.

That said, autonomous driving, may in a sense save ICE makers from seeing there sales fall off a cliff due to the dynamic you described. Maybe it quite substantially mitigates such a potential dive, maybe not.

Circa 2025, a family with a 10 year old ICE car might have some inclination to hold onto it a few years longer if that's what it takes to replace it with a well done EV rather than an ICE. Depending how autonomous driving tech progresses, in capability and regulatory approval, and depending on how widely available it becomes, that inclination could become irrelevant for most in the context of their preference to buy a new vehicle with autonomous capability, ICE or EV, that is 10X safer for their family than the 10 year old ICE car.

Lots of variables here... will the tech get approved by then, will it be available to most ICE mfgs, will this lead to sales dropping heavily anyway because it turns out car ownership is not worth it to a large part of today's market once autonomous fleets exist.

So much is up in the air... but, I've yet to see a compelling scenario where Tesla is not continuing to grow revenues at a torrential pace (30-50% per year) from now into the 2030s. I think that would take their soundly failing in the autonomous vehicle market. I see that as unlikely to happen, and, even if it did, they would still have a good opportunity to navigate to very strong growth on the back of their outstanding position to succeed as a supplier of electric vehicles.

 

[TintDepot.com]

But wait, there’s more.

500k vehicles annually in Shanghai. That should hit steady state between 2023-2025. I don’t know the mix of vehicles, but if the average price is $40k, that’s another $20B annual revenues. $100B total.

Roadster should be steady by then, but low volume. 8k/year is $2B, 20k is $5B. Your guess is as good as mine.

If Tesla Energy grows at just 50% per year, it should be at $12.7B by 2023. Services & Supercharging should provide at least another $10B in revenue, but no earnings.

So that’s somewhere around $130B in revenue by 2023 or so. Without consider the Pickup, the Semi, or Tesla Network.

China is a premium auto market; Teslas average price should be $50,000-$65,000, so your numbers will be even higher.

 

[jbih]

This is a really good video even for people that follow Tesla every business day.

What a great video! Didn't want to watch it first because I thought the title is click-baity. Glad I did.

 

[Esme Es Mejor]

at an average of 25 %, this would represent $32.5B in earnings?

25% gross margin on $130B in revenue would represent $32.5B in gross, but net profits would be a lot lower. Recent earnings for BMW and Toyota have been around 8-9% of revenue.

Assuming Tesla can achieve 10% net, that’s more like $13B in earnings. $10-11B, if you assume no earnings from the service/supercharger revenue.

If they achieve large-scale revenues with the Pickup or autonomy— or grid services on the Energy side— margins would be much higher, but success in those areas is much more speculative at this point.

 

[SteveG3]

Has anyone here done a back-of-napkin calculation of what Tesla earnings might be when the company is actually able to supply enough product to meet the market demand for its products?

Let’s say that in 5 years Tesla has finally built enough manufacturing capacity to satisfy sustainable annual demand for MS, MX, M3, MY, Roadster, Semi, pickup, Powewalls, solar roofs, etc. What might a reasonable estimate of total annual revenue look like? And then applying reasonable margin and cost assumptions, what amount of annual earnings does that imply?

I realize this requires making assumptions with enormous margins of error, but I think it’s even more erroneous to make assertions about a company’s valuation based on projections that do not go out long enough to take into account meeting a company's reasonable sustained annual demand.

I posted a basic model for this in the thread below in April '17. The basic mechanics of the model still work, but, lots of my assumptions look real conservative now (ASPs for several vehicles, and my Roadster margins were very very low compared to what we are likely to see). I think we now know enough about Tesla semi, and it's likelihood for at least some basic success, to add it into a model as well. US corporate tax rate has now been lowered. Tesla has strongly indicated there will be less dilution than it looked like when I assumed 235 million shares in 2026.

skip to the bolded, "On to the model section" if you want to skip the commentary,

Google+ video hangouts (TSLA & other investments)

 

[Esme Es Mejor]

China is a premium auto market; Teslas average price should be $50,000-$65,000, so your numbers will be even higher.

I certainly hope so. But like I said, I have no idea what the mix will be, so I chose what I think is a very conservative number. If the number turns out to be much higher, I’d be very happy.

 

[Mike Smith]

Posting the VIN numbers has become a full-time job:
Tesla registered 3,609 new Model3 VINs. ~82% estimated to be dual motor. Highest VIN is 92929.
Source: Model 3 VINs on Twitter

The Bloomberg Tracker is still stuck at 75k VINs, which makes it a pretty lousy tracker. Even worse is that it specifically claims to be up to date to Aug 6.

The good news is it should spike higher whenever they do update it.

Edit: They updated it as soon as I wrote this comment. It's often many days out of date, yet 'last updated' seems to be automatically set to the current day.