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

[ggr]

I just saw a Highland Model 3 on the road in San Diego. I didn't realize they are delivering them yet. It had the temporary paper plate, so it wasn't a demo vehicle.

 

[Musskiah]

Each car on average could probaby generate $50 / month in subscription fees from FSD. $600 a year.

I think everyone who puts FSD at $50-100 a month woefully underestimates the utility of it. Forget the value of incredible added safety benefits (what is your life and your family's lives worth to you and everyone in their lives?), and the reduced driver fatigue value, and the reduced need to pay attention to where you are going, and they ability to attend to a child in the backseat, and the fact that it is like accident avoidance insurance..and on and on and on... all the other ancillary benefits we can think of. Forget all that.

Just do this little exercise: I think we can agree the average person who purchases a Tesla probably makes over $25 per hour. Right? So if they get 2 hours worth of productivity (answering emails or texts, conference calls, or other productive tasks...just as examples) back from FSD each month, they break even. I get so much work done while on FSD it's insane. FSD at $200 a month gets paid for in literally minutes for me. It's a no brainer just from this aspect alone. It has been since Mobile Eye Autopilot in 2016 on the highway TBH.

The only trouble is, people don't understand how much FSD IN IT'S CURRENT FORM saves lives, hassle, headaches, fatigue, and most importantly productivity. I could justify paying $500+/month for it easily. It's incredible! By 2026-2027, when you can literally watch a movie or actually do work completely uninterrupted...I will laugh hysterically if anyone suggests it's worth less than $200 a month. Some people pay more than that for cable still!

 

[DarkandStormy]

Whomever can see Joe's photos from his latest flyover G-TX, see the number of CTs stored nearby. It is quite something else, compared even to last week. He estimates 1K/Week which is still just a beginning.

Point: Overwhelming majority of those new CTs sold take a bite from other trucks. "Let that sink in"

I don't think the influencer crowd who are early buyers of the CT would have bought another truck, imho.

 

[Tony73]

Some googling suggest Tesla sold 650k cars in the US in 2023. Some people guess FSD take rate is 19%. I am assuming that even given no more progress, FSD V12 doubles that over time.
so 123,500 extra FSD purchases per year, for revenue (its $12k now right?) of $1.48 billion a year. The marginal cost of sales here is virtually zero.
Thats almost exactly 10% of profit for 2023. That assumes FSD is never improved outside the US.
Given demand for FSD in Europe, China etc, I think 2025 is going to be the year we really see FSD hit the financials in a notable way. Luckily between now and then we have growth in profits from energy, and also the cybertruck and semi ramp.

Of course if FSD 13 comes along and is genuinely end-to-end no intervention for 99.999% of drives, then all bets are off and its robotaxi time, but WAY before that I expect FSD to become a huge driver of sales for people cross-shopping between EV brands.

FSD does not have a 19% take rate. It’s somewhere around 7%.

 

[DarkandStormy]

5 million on Sept 17 2023 so close to 2M run rate.

https://twitter.com/x/status/1773775354315841596

I believe this implies ~440k of Q1 production. Inventory likely to grow a bit.

EDIT - they announced 4 millionth vehicle in March of 2023. So they haven't really increased their production rate over the last year.

 

[Singuy]

The summary of this is simple: A long term bull is now realizing Tesla's brand value advantage isn't given it the extra margin cushion it once did, which has long-term implications to its automotive valuation.

He probably didn't expect to rely on valuation Tesla's AI to product potential for stock appreciation previously, which now everyone has pivoted to.

I will say, I think now isn't the best time to evaluate where Tesla is compared to other automakers. All EV demand is getting wrecked, it isn't Tesla specific. The fact is Tesla still has better margins on EVs than anyone else. Them having to using advertising to move cars in this market doesn't mean they aren't superior as an EV brand.

To me the question is less about marketing, and more about how much are people willing to pay for an EV and the associated costs?

Are electricity rates going to go down relative to gas costs in the next few years? Not sure elsewhere, but not in California it seems.

How much are people really going to pay for "supervised" FSD? I don't think most people are going to spend $100 / month, let alone $200.

This is why I model $5 in EPS in 2025 and not more.

What is the associated cost for a Tesla that is more vs the equivalent ICE car? I just calculated that I have saved 12k in gas and 310 file up trips after driving 100k miles. If each trip cost me 10 minutes, I have saved myself over 51hours of time. I swapped out tires every 50k miles and bought some cheap Chinese ones which gave me zero problems in the FL rain(never hydroplaned even once).

Last time I checked, a similar 5 second car from all the premium brands cost the same if not more than a Tesla.

Are people still comparing 100k luxury interiors to a Tesla model 3 and then a hyundia accent's associated cost to a Model 3?

 

[Singuy]

I just saw a Highland Model 3 on the road in San Diego. I didn't realize they are delivering them yet. It had the temporary paper plate, so it wasn't a demo vehicle.

Spotted one in Central FL as well. They are definitely delivering them.

 

[rallykeeper]

Troys current delivery estimate last I heard was under 410k.. and he tends to be very accurate this near to EOQ... so if production is >450k that'd be a massively large glaring gap between the two (largest gap ever)

Agreed. It seems like Tesla's tweet may have been a warning shot of sorts on P&D estimates.

Other analysts seem to be homing in on 425k deliveries.

While it's possible Tesla is ending the quarter with 25-40k vehicles in inventory or transit (assuming no change in Q4/Q1-end inventory), that seems very unlikely. (Our ship-counters in this thread seem to have departed so not sure if that's even technically possible?).

Presumably, Tesla will produce another few thousand cars today at Fremont and Texas (plus whatever cars are produced this weekend). If anything, I'd guess we land somewhere around 455k or so (if not slightly higher).

As to deliveries, my guess is in the same ballpark.

 

[Zaddy Daddy]

I think everyone who puts FSD at $50-100 a month woefully underestimates the utility of it. Forget the value of incredible added safety benefits (what is your life and your family's lives worth to you and everyone in their lives?), and the reduced driver fatigue value, and the reduced need to pay attention to where you are going, and they ability to attend to a child in the backseat, and the fact that it is like accident avoidance insurance..and on and on and on... all the other ancillary benefits we can think of. Forget all that.

Just do this little exercise: I think we can agree the average person who purchases a Tesla probably makes over $25 per hour. Right? So if they get 2 hours worth of productivity (answering emails or texts, conference calls, or other productive tasks...just as examples) back from FSD each month, they break even. I get so much work done while on FSD it's insane. FSD at $200 a month gets paid for in literally minutes for me. It's a no brainer just from this aspect alone. It has been since Mobile Eye Autopilot in 2016 on the highway TBH.

The only trouble is, people don't understand how much FSD IN IT'S CURRENT FORM saves lives, hassle, headaches, fatigue, and most importantly productivity. I could justify paying $500+/month for it easily. It's incredible! By 2026-2027, when you can literally watch a movie or actually do work completely uninterrupted...I will laugh hysterically if anyone suggests it's worth less than $200 a month. Some people pay more than that for cable still!

That is the average value per vehicle I am estimating, as many people won't subscribe at all, while others will pay $100 / month. Few will spend $200 / month and very few will spend $500 / month. You can't assume all the owners will spend the max number of money for the product.

Most people don't have $200 / month lying around. Their monthly payments are already $500-$700.

 

[Tony73]

I could justify paying $500+/month for it easily. It's incredible! By 2026-2027, when you can literally watch a movie or actually do work completely uninterrupted...I will laugh hysterically if anyone suggests it's worth less than $200 a month. Some people pay more than that for cable still!

Has Tesla ever said that they will offer full autonomy outside of a company owned or company operated fleet ? I think what you are describing - privately owned vehicles in full autonomous mode in private operation - is decades away. Sure an individual can choose to sleep while on FSD but I don’t think Tesla is going to market their product as such or accept liability for such operation anytime soon. FSD in private hands will require supervision for a very long time.
Full autonomy will happen only in the context of Tesla robotaxis or private vehicles that are temporarily part of the robotaxi fleet. I see zero chance of Tesla marketing a fully autonomous solution to Tesla buyers in the 2026-2027 timeframe even if they have already started rolling out robotaxis.

However that said, a supervised solution with rare interventions is still easily worth $100 - $200 a month for a lot of people. It will drive a lot of buying decisions.

 

[Troy]

Troys current delivery estimate last I heard was under 410k.. and he tends to be very accurate this near to EOQ... so if production is >450k that'd be a massively large glaring gap between the two (largest gap ever)

I get different results for the 6 millionth car calculation if I subtract all cars produced in all past quarters instead of starting with the 5 millionth car they reported.

There is a public spreadsheet here that shows Tesla's production history in the past:

Tesla Production & Delivery History, created by @TroyTeslike, last updated on 19 Apr 2024

Deliveries 1 Canada,Canada,Canada,China,China,China,Europe,Europe,Europe,ROW Deliveries,Canada,China,Europe,ROW,USA,Total,Deliveries,Model 3,Model S/X,Model Y,Model 3,Model S/X,Model Y,Model 3,Model S/X,Model Y,Model 3 Year,Canada,China,Europe,ROW,USA,Total,Year,Canada,Model 3,Canada,Model S/X,C...

docs.google.com

It shows 5,572,978 units produced until the end of 2023 plus 2,349 Roadster Original equals 5,575,327 entering Q1 2024. If we subtract that from 6 million, we get 424,673 production so far in Q1.

It was recently reported that Giga Shanghai now works 5 days a week instead of 6.5 days. See the news article here. I think Giga Berlin is also 5 days. Giga Texas Model Y production is probably 5 days too.

 

[rallykeeper]

I get different results for the 6 millionth car calculation if I subtract all cars produced in all past quarters instead of starting with the 5 millionth car they reported.

There is a public spreadsheet here that shows Tesla's production history in the past:

Tesla Production & Delivery History, created by @TroyTeslike, last updated on 19 Apr 2024

Deliveries 1 Canada,Canada,Canada,China,China,China,Europe,Europe,Europe,ROW Deliveries,Canada,China,Europe,ROW,USA,Total,Deliveries,Model 3,Model S/X,Model Y,Model 3,Model S/X,Model Y,Model 3,Model S/X,Model Y,Model 3 Year,Canada,China,Europe,ROW,USA,Total,Year,Canada,Model 3,Canada,Model S/X,C...

docs.google.com

It shows 5,572,978 units produced until the end of 2023 plus 2,349 Roadster Original equals 5,575,327 entering Q1 2024. If we subtract that from 6 million, we get 424,673 production so far in Q1.

It was recently reported that Giga Shanghai now works 5 days a week instead of 6.5 days. See the news article here. I think Giga Berlin is also 5 days. Giga Texas Model Y production is probably 5 days too.

This is helpful. I was using other sources.

Shows why I shouldn't dip into counting!

 

[Krugerrand]

Oh, absolutely it's a contradition. To paraphrase Tesla Sage Earl, It's analogous to 'Appreciating Asset (Depreciating)'. But since 'Full Self-Driving' (ie back-seat riding) was a non-starter when it started (har har), still is, and always will be on current hardware, specifically calling out the supervision-required nature right in the name rather than in the documentation/opt-in screens is a clear improvement. And IMO it would have deflected a lot of the well-deserved flak the company has gotten over the years.

Krugerrand clearly disagrees. That's OK.

I don’t disagree with all of it. And I don’t dismiss ‘your’ personal experience or reaction. I dismiss the notion it makes a difference in general.

If it made a difference we wouldn’t need bathtub instructions for toasters.

I will change my mind if after this headline appearance of the word (supervised) suddenly causes the termination of Tesla drivers to hit police cars, fire trucks, ambulances, buildings, big rigs etc… while FSD is engaged. If suddenly people stop using defeat devices. If suddenly people stop getting into Teslas drunk, high, overtired etc…. AND, I will come here to say I was wrong about people in general as soon as the NHTSA announces this week that Tesla can remove the nag feature completely because the word (supervised) appeared in a headline after FSD.

 

[hobbes]

Contributors to this site often talk of the exponential curve when talking of ramping production/sales. Some times we tend to forget that the high end is a mirror image of the start. Some of Tesla’s products are at that end of the curve, The S, X, and possibly the Y. This makes the small car and the truck all the more important. Yes, there are a lot more markets to enter, but the majors have been entered. (The exception Brazil and India, if they are major. Unk? ) The S and X have been stagnant for years. The last refresh did not increase sales. Tesla is in an in between period. The refreshes should help the Y and 3, but true exponential growth depends on the small car.

Speaking of exponential growth - can also happen on a small scale (for a short while that is)...

https://twitter.com/x/status/1773042054894903389

 

[Krugerrand]

Actually you make a perfect point, specifically about Starlink, one fo the more public-facing musk enterprises. You have seen a Starlink ad because you are, as you explain, in a very rural area probably, I surmise, with substandard public internet access solutions. I, on the other ahdn have never seen a Starlink ad (except within an off-site test) because both my residences are within major urban areas with a plethora of fast internet access solutions.
The advertising trick for non-ubiquitous markets/products, is to place the content in front of likely prospects. You, Krugerrand, seem to qualify. I, on the other hand, could buy , but am unlikely to want to deport my 1meg current products.

As in this case, at present Tesla advertising looks a trifle generic to those who see it, but it is not presented to unlikely buyers. To the contrary, among the finest prospects are those who already own an existing product. There Tesla excels by presenting highly attractive trading options, ones I've accepts three times thus far, for example.

The real question is what happens when/if Tesla decided to copy General Foods, P&G or Mars. Then, if it happens, we watch high Free Cash Flow disappear quickly.

You would surmise correctly. I am in the middle of nowhere. Not sure why the retro TV station spot, but I won’t argue there isn’t a very good reason whether I know what it is or not.

I really don’t think Elon would purposely or knowingly let money be wasted on generic advertising trifle. If it happens, I’ll first assume it was a mistake by an intern. Only if it were to continue for a period of time such that it became clear that was the path forward would I pause to consider negative implications for the company.

 

[Knightshade]

Has Tesla ever said that they will offer full autonomy outside of a company owned or company operated fleet ?

Yes. Often. For years. Directly in the product description of FSD.

I think what you are describing - privately owned vehicles in full autonomous mode in private operation - is decades away. Sure an individual can choose to sleep while on FSD but I don’t think Tesla is going to market their product as such or accept liability for such operation anytime soon.

FSD as sold from launch in late 2016 until the big change in ~March 2019 promised buyers exactly that. The vehicle would drive the car on both short and long trips without requiring any action of any kind from a human.

It got even MORE explict if you count Elons own remarks (summon from the other side of the country as a promised ability for example) but even sticking to the literal FSD product description it was minimum L4 and arguably L5 being promised to buyers.

Those buyers are still owed that in fact.... (versus the folks who bought April 2019 onward, where the product description during purchase never promises more than L2)

This has been discussed- exhaustively- in the FSD forum for years now.

 

[Drumheller]

I think everyone who puts FSD at $50-100 a month woefully underestimates the utility of it. Forget the value of incredible added safety benefits (what is your life and your family's lives worth to you and everyone in their lives?), and the reduced driver fatigue value, and the reduced need to pay attention to where you are going, and they ability to attend to a child in the backseat, and the fact that it is like accident avoidance insurance..and on and on and on... all the other ancillary benefits we can think of. Forget all that.

Just do this little exercise: I think we can agree the average person who purchases a Tesla probably makes over $25 per hour. Right? So if they get 2 hours worth of productivity (answering emails or texts, conference calls, or other productive tasks...just as examples) back from FSD each month, they break even. I get so much work done while on FSD it's insane. FSD at $200 a month gets paid for in literally minutes for me. It's a no brainer just from this aspect alone. It has been since Mobile Eye Autopilot in 2016 on the highway TBH.

The only trouble is, people don't understand how much FSD IN IT'S CURRENT FORM saves lives, hassle, headaches, fatigue, and most importantly productivity. I could justify paying $500+/month for it easily. It's incredible! By 2026-2027, when you can literally watch a movie or actually do work completely uninterrupted...I will laugh hysterically if anyone suggests it's worth less than $200 a month. Some people pay more than that for cable still!

Valuing an item or service doesn't mean the person can afford it. Many office workers in the USA work on salary. They are not paid for overtime. Spending $200 / month to get more work done while traveling means they still spend the same amount of time in the office , work extra during your commute, and recieve the same pay.

Also, if you stretched your budget to afford a Tesla, you may just not have the budget for FSD, regardless of its value.

 

[GatorMeat]

(market closed today, so slightly OT)

This surprises me. Asphalt shingles can be blown off in storms far less energetic than a hurricane (I've lost a couple in heavy thunderstorms not approaching hurricane strength). Given the hurricane rating of an asphalt shingle would be a seemingly a low bar, from what we've seen of the solar shingle construction and attachment methods, that seems odd.

Tesla says on their website regarding the solar roof:



And according to this asphalt roofing resource , they agree that class F is the highest rating:

When I looked at Solar Roof I believe the wind rating was only 140 MPH. Miami Dade County I believe moved to 185 MPH and I don't think Solar Roof has their approval. Believe it or not, new asphalt roofs fared ok in 2018's Hurrican Michael. Some believe metal roofs are best, but even those got peeled back or badly damaged by debris. The Solar Roof would have to withstand get blasted by debris also. I'm not sure the California designed Solar Roof has much real world hurricane experience, but maybe Miami Dade is approving them? But Solar Roof still costs 3x-5x of asphalt/metal plus solar panels.

 

[GatorMeat]

When I looked at Solar Roof I believe the wind rating was only 140 MPH. Miami Dade County I believe moved to 185 MPH and I don't think Solar Roof has their approval. Believe it or not, new asphalt roofs fared ok in 2018's Hurrican Michael. Some believe metal roofs are best, but even those got peeled back or badly damaged by debris. The Solar Roof would have to withstand get blasted by debris also. I'm not sure the California designed Solar Roof has much real world hurricane experience, but maybe Miami Dade is approving them? But Solar Roof still costs 3x-5x of asphalt/metal plus solar panels.

From Owens Corning, it looks like the F rating is one of two wind rating tests. It is a fan induced rating for steep slope roofs. The second rating is for uplift force winds which is important for hurricanes, and Solar Roof doesn't appear to have a rating I could find for that type of wind. Asphalt shingles have an adhesive strip that seal the leading edge to the shingle beneath it. I wasn't willing to spend 5x to pay a company to drive 5 hours from Tampa for the Solar Roof unless they could guarantee it's worth it in a hurricane and could promise I could get insurance. Many people even here in the FL panhandle have trouble getting covered with solar panels.
"

ASTM D3161​

ASTM D3161, or the Standard Test Method for Wind Resistance of Steep Slope Roofing Products (Fan-Induced Method). This testing method measures a shingle’s ability to withstand fan-induced wind speeds and includes three classifications:

• Class A for shingles that passed testing at 60 mph
• Class D for shingles that passed testing at 90 mph
• Class F for shingles that passed testing at 110 mph

ASTM D7158​

ASTM D7158, or the Standard Test Method for Wind Resistance of Sealed Asphalt Shingles (Uplift Force/Uplift Resistance Method). This standard assigns a classification to individual shingle products based on their resistance to uplift forces when tested at specific wind speeds:

• Class D resist uplift at wind speeds up to 90 mph
• Class G resist uplift at wind speeds up to 120 mph
• Class H resist uplift at wind speeds up to 150 mph

Asphalt roofing shingles in the market generally all carry the same wind resistance classifications: ASTM D3161 Class F and ASTM D7158 Class H. However, warranted wind coverage varies by individual product, so be sure to check the manufacturer’s product information, warranty details, and installation guide."

 

[Blue horseshoe]

I get different results for the 6 millionth car calculation if I subtract all cars produced in all past quarters instead of starting with the 5 millionth car they reported.

There is a public spreadsheet here that shows Tesla's production history in the past:

Tesla Production & Delivery History, created by @TroyTeslike, last updated on 19 Apr 2024

Deliveries 1 Canada,Canada,Canada,China,China,China,Europe,Europe,Europe,ROW Deliveries,Canada,China,Europe,ROW,USA,Total,Deliveries,Model 3,Model S/X,Model Y,Model 3,Model S/X,Model Y,Model 3,Model S/X,Model Y,Model 3 Year,Canada,China,Europe,ROW,USA,Total,Year,Canada,Model 3,Canada,Model S/X,C...

docs.google.com

It shows 5,572,978 units produced until the end of 2023 plus 2,349 Roadster Original equals 5,575,327 entering Q1 2024. If we subtract that from 6 million, we get 424,673 production so far in Q1.

It was recently reported that Giga Shanghai now works 5 days a week instead of 6.5 days. See the news article here. I think Giga Berlin is also 5 days. Giga Texas Model Y production is probably 5 days too.

This is informative and fact-based with associated links to back up the post.

I find these types of post very valuable; versus the posts where subjective opinion is often injected into the post and written as fact.