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

[jerry33]

Exactly that





Because the alternative is hiring a massive # of people to manually charge cars. Though as I mention they might need to anyway to keep the interiors cleaned out regularly. Then they'd also need some kind of shelter/booth for that person to remain on-site all shift in bad weather. The ideal case is using Optimus I suppose (a weatherproofed one). But none of that is FAST to get done if you're talking deploying millions of cars, and including induction capability means you don't have to RELY on having someone to plug/unplug RTs. Of course it relies on installing a ton of inductive chargers- but you will need to add a ton of chargers ANYWAY to support a fleet of millions.




Not the OP, but...

Alabama, South Carolina, and New Mexico ban not just direct sales, but also service centers. WV, Kansas, Louisiana, Wisconsin, Nebraska, Iowa, Oklahoma, Arkansas and Connecticuit also ban direct sales (CT allows direct leasing however).

Another 9 states allow direct sales, but heavily limit the # of allowed locations to do so.

In a few cases Tesla has done an end-run putting a sales center on a Native American reservation, in others as you suggest they conduct the "sale" out of state and a customer still gets their car-- but at greater effort and cost to at least Tesla, and sometimes the consumer (lots of stories about added time/effort/etc around registration and other paperwork stuff on this in various delivery threads here)- and of course the states that also ban service centers directly worsen customer experience.

And the states that ban sales keep introducing bills to also ban service centres because they've found that just banning sales doesn't stop the flood of Teslas. So far these bill haven't passed.

 

[Ellec]

Kinda wondering if Tesla of late is shaking many of the bulls in this thread....maybe I should re-strategize my DCA approach....seems like we could go quite a bit lower over the next several months....I may need to be buying for many months....

People that are new perhaps, but if you been invested in Tesla for a while this is nothing burger, it's business and what Tesla does when preparing for a new growth stage. Im DCA'ing monthly for the next 3-5 years on top of my shares from before so don't mind if the price coils up for a while, will be good to build up some short interest to fuel the inevitable take off.

 

[SOULPEDL]

Sometimes I wonder if RT is more of a concept than an actual product. ie FSD on your personal car that is legal to take you from A to B w/o a driver.

The product for us will likely be called FSD (Unsupervised), then Sleep-n-Travel feature. (FSD "waves" incoming?)

I don't think it can RT properly unless someone creates a plexiglass separator for the driver seat (like Waymo, nothing new). So like $100 in materials and $500 to install, pay for the license (inspections etc), it's now a Robotaxi. Big maybe on this, but possible.

Please, don't ruin my dreams with more FUD. It makes me want to buy more TSLA!

 

[unk45]

That's very noble of you. I'm pretty sure if I sold all of my TSLA stock I would be hoping I was right. What was your reasoning? Bad customer service experience ala @neroden?

I’ve explained that in recent threads. I simply realized they’ve stopped the superb marketing expertise they once had, displayed to excellent results in China where the practices were most advanced but in much of EU and NA too.

Recent attacks on numerous regulatory and influencers did not augur well, either.

This is nothing like the Neroden personal pique.

 

[jerry33]

I don't think folks get this fully. Chinese BEVs are already well under $25K. No, they can't import to US yet, but the rest of the world is fair game.

Even if they could, most of those cheap BEVs wouldn't pass anyone's safety standards.

 

[2daMoon]

What you are stating was always the plan, but most here seem to think FSD or bust. Elon is NOT building a factory in India to build RTs. But do you know what would sell to the chagrin of many here? A cheaper Gem 3 EV.

https://www.reuters.com/business/autos-transportation/teslas-musk-likely-unveil-2-3-bln-india-investment-during-visit-sources-say-2024-04-17/

This sounds like a good strategy.

India is positioned well as an export hub than would be Mexico. India could serve the Americas well, particularly Central and South America. Africa and the Mid-East are close, and Europe is closer than it is to China.

I think that Mexico has been delayed for several valid reasons. One of which is that an RT/Gen3 will be best deployed in the US first, as this is where most of the training data has come from. Tesla needs to finalize the product in Texas before they can start building in Mexico. There are many, many examples of Tesla having to rebuild parts of a factory shortly after construction, so, pausing on Mexico until the line has been finalized helps avoid some of that.

India is very "renewable" friendly. They are in the top five of solar power generation nations. They also have a lot of tech-savvy people and rank between China and Brazil for number of STEM engineering graduates. Mexico is ranked further down the list, between Ukraine and France.

India has a strong automotive manufacturing base which is pivoting into the EV sector, so more OEMs who might benefit from the "Tesla Inside" aspect of robotaxi production built with Tesla autonomy components. TATA Electronics are already making noises about chip production and their automotive side might make a good match for partnering with Tesla on two fronts.

Mexico still has advantages with economic ties to both Latin America and their proximity/economic/auto production ties to the US.

Things are looking good for Tesla's expansion into more of the world's markets, as well as for autonomy coming to fruition eventually and being deployed worldwide over the long term. 20M cars by 2030 has always been a good target to aim for.

 

[Mengy]

Not sure if it's been discussed, but how is Tesla planning to grow volumes in Q2 and the second half of 2024? They just ended inventory discounts and raised prices $1k on April 1st. They previously had a demand problem (supply > sales) as @Troy has noted since late 2022. Increasing price only decreases demand if you aren't materially changing the product you're selling.

Any theories on how (or if) Tesla will grow unit volumes the rest of the year?

My hunch is production will go back up to Q4 2023 levels since we're past Chinese New Year and Fremont is producing Highland M3';s now, and hopefully no more terrorist attacks and such.

I think at best we're looking at a 1.9 million delivery year for 2024, possibly 2 million if Tesla really pushes hard.

 

[⚡️ELECTROMAN⚡️]

If RT is announced with a different sensor suite than all other Teslas, more cameras or Lidar, etc. I think we'd see some panic on 8/8.

Panic to sell, or panic to buy? All they really need is maybe a few more cameras to cover the current blindspots, and a way to keep them clean in all weather conditions. It's not comparable to rocket science, like the software successfully being capable of FSD is.

 

[jerry33]

Given that, as an investor (or a test if I really am one), I started to think more specifically about Robo-Taxis and how they might really work in my life and the lives of people I know. It is one thing to parrot "it is a gazillion dollar industry!" or "It will never work" - it is another to try to model it out...not with the goal of proving it works or doesn't, but with the goal of simply trying to see if it fits your use cases or how it could.
My wife and I are older and neither of us is currently working (although if TSLA keeps dropping I likely will be...but I digress), so we modeled for us and then discussed my son and his wife with two little kids (in car seats)....and then larger families. We also discussed the interplay between FSD on your PERSONAL car and Robo-Taxis. Here is some of the stuff we covered in a discussion this morning:

Commuting
Commuting back and forth to work seemed like an obvious use-case. It has a reasonably predicable time each day for coming and going. The challenge - you need a lot of RobotTaxis of course (as noted in other discussions) since commute windows are "peak" utilization.

Grocery Runs
A grocery store run requires two rides - one there and one back. The one there is relatively easy. The one back is different as it requires you load groceries into the trunk/frunk and when you get home, the car has to sit and wait for you to unload it...preferably close to a door nearest to your kitchen. In our case, that would be a door INSIDE our garage. So, for this use case, the car will need to have software capable of:
- implementing a "wait" mode (i.e. don't drive off right after you drop me off)
- park "here" specifically
- implementing an "extend wait for x more minutes" (i.e. little Johnny had to go to the bathroom, need car to wait while I handle that before I unload those groceries) - would text user to let them know "Hey, I am still waiting...do you want to continue or can I leave?", etc.)

Multi-Errand
When out and about we often run multiple errands. In some cases, the car will need to have software modes to just "wait here" while the user goes inside (i.e. to pick up dry cleaning perhaps or a prescription). In other cases, if an errand was going to take a while (like going to the doctor) before the next one, you will be better off terminating that trip. and then summoning another taxi when done (to either take you home or head to next errand (i.e. pick up a prescription after that doctor visit).

One thing that starts to become clear, is that you need a lot of Robo-Taxis in many cases to satisfy the "go here, do something for a while and then go home or to the next place". I do however, think that a Robo-Taxi might just "sit" somewhere (like the grocery store) after it drops you off there...if it gets another call, you will be summoning one to come get you when you are done. If not, it is there ready for you when you come out.

Spur of the Moment Stuff
Lots of errands are spur of the moment or less "scheduled". This can be solved with RTs, but really comes down to the need for a volume of them.

Kid Car Seats
If you have kids or know anyone with kids you can appreciate the challenge of car seats. They are often huge, multi-part (carrier plus base), rear-facing (taking up a lot of space), and unique. Children are in rear-facing seats for YEARS now (this ain't the 60s/70s). In a RoboTaxi model some options might be:
- Rider brings the car seat and installs it themselves - this is practically a non-starter for most parents
- Robotaxis have "convertible" seat options where one or more seats can do some "flip around magic" to have a car seat (or car seat "base" at least) appear at the push of a button or whatever. Volvo has a simple booster seat option similar to this, but rear-facing seats are, of course, more complex.
- There could be Robo-taxis with dedicated child car seats
I do think a "universal" base for car seats would help here as owners might be willing (in SOME cases) to just slot their carrier into a universal base. Note that carriers are typically just used for babies and very small toddlers (i.e. my 1 year old grandson is in a carrier sometimes, but his 3 year old sister is obviously not).

Big Families
A 6 person family where everyone is going to church or a ball game or wherever will obviously need more seats in any vehicle - with some of them potentially being kid car seats, etc. In this case, larger "mini-van" style Robo-taxis will be needed. This seems pretty-straightforward. Ultimately, it will come down to the percentage of people needing X. Mini-van taxis will be produced in the quantities the market requires. Indeed, I could see the RT world with two predominant forms - the 2 seater and the mini-van (call it a 6 seater).

Hauling Stuff
This and other use-cases start to become a little more "one-off" for the AVERAGE person. They could be addressed with dedicated vehicle types or just leave them as rented standalone cars.

Your life
My recommendation is to model trips you have taken over the last month against two things:
- Robo-Taxi
- Full Self Driving in your own personal car

As an example, on Monday nights my wife and I happen to attend two different Bible-studies at our church. Today, we have ONE car (note: my Tesla broke...and then broke again...and again...was compelled to sell it - a story for another time). Right now, her study starts at 6, and mine at 7pm. I was taking her at 6, driving home, driving back at 7 for my study, then (since hers ran longer than mine - she is a leader), either waiting around for her or going home only to turn around and go back and get her. A Robo-taxi would clearly help with that, but SO WOULD FSD IN MY PERSONAL CAR. She could take the car at 6pm and then "send it back home", for me and then I would drive (or it would drive me) it to my 7pm and back home...and then I send it back to the church to get her...basically, it becomes our own personal taxi (but w/o a driver).

Any overlapping event scenario starts to become very interesting for either a personal car with FSD or RoboTaxi world. If I wanted to see one child's school play, while simultaneously needing to pick a second child up from band practice, I could either leverage RoboTaxis OR drive to my child's school in my own car and then SEND my personal car (with FSD) to get the second child. Heck, the latest Uber ads are clearly targeting this kind of thing. When any parent sees those Uber ads they are concerned about the Uber driver (in a world of "don't talk to strangers!") - which is why the ads purposefully talk about only the best drivers, etc. getting these drives. Now take that fear away with no driver in a RT or your own personal car (with FSD).

At first, it seemed like the RT would be a pain in the butt to deal with, but the more I really challenged it the more it seemed to work. Certainly, I started to see complimentary scenarios where both RTs and personal cars with FSD would be great. The combination of the two makes a "one car family" a LOT more feasible than I imagined.

TLDR - lots of Robo-Taxis with specialized seating combined with personal cars with FSD likely works for a huge number of use cases.

Finally, one additional tidbit. Jordan from "The Limiting Factor" recently test drove the Cybertruck (which does not currently have FSD), and posted a short video afterwards noting that he missed FSD from his current Tesla, BUT not just as a "convenience"...he recognized he has become somewhat "reliant" on it and it was a little jarring (my words) dealing with driving without it. Driving may just be something folks just don't want to "deal with" over time. Heck, very few shift their own gears these days. They don't even like to unlock their car or even have to use the fob now!

During the remodel of our house, the X was our bedroom for a couple of months. Saved more than 60 days of hotel bills. And then there are also the many nights of camping.

 

[Christine69420]

I like what you're suggesting, but felt Tesla did something similar with 4680s. Didn't they originally aim to have Berlin, and then Austin build MYs with structural 4680 packs, with a risk-avoiding strategy of non-structural 2170s as the "backup" plan. I don't know if I'm misremembering or have been misguided by FUD. I'm just left with the feeling that Tesla was overly hopeful on the 4680 timeline. I think all MYs are currently 2170 cells? If true, I'd prefer a parallel play for FSD and gen3. Nonetheless, I just put in another buy order, and will continue to DCA more at these levels...just with less confidence than usual.

Tesla gutted the Berlin battery factory when the IRA invaded the US. So those batteries will be made in Austin since they will get inflation reduction money from the US Government.

I'd guess that new battery production equipment for Berlin is on order somewhere.

Several other battery factories in Europe were cancelled for the same reason.

 

[EQC_]

...

Also, I can't stand how most want to have their cake and eat it too. They want to own TSLA shares that nicely appreciate each year and Elon should focus on more car models, a cheaper car etc, AND evolve FSD , but risky strategies that involve producing the necessary batteries (4680's) or progressing towards FSD (building out compute, sensor suites and inventing cutting edge NN's/training methods) are a big no no. Wut?

...

It's also difficult to separate wants as a consumer from "best for the company" as an investor.

Personally, as a consumer, *I* want to own a Model 2. I've got a Model Y, because that was the cheapest EV that fit my needs (7 seats / 4 kids). But I don't need or really even want much of what is included in the Model Y -- don't need big wheels or low profile tires or a 0-60 below 8 seconds or a top speed over about 90 mph or leathery seats, or a sexy exterior, etc.

Knowing that consumer-me wants to be able to buy a cheaper simpler Tesla, it's difficult to convince the investor-me that the company delaying that is a good thing.

 

[jschwefel]

I disagree with you that going all in on autonomy is a risk free strategy, as FSD is not solved yet.

IMHO, the better strategy would be to produce the consumer variant of the Gen3 car FIRST with manual controls (for which the potential market is HUGE) while FSD is being solved over time, and THEN begin making the RT on the same lines as the Gen3 once FSD is ready. I agree with Franz and Lars on this one.

If these rumors are correct then Elon is doing this backwards, in an extremely risky decision, because if the manufacturing lines get ready to produce the RT and FSD is still not solved, well what then? Time to retool to make the Gen3 for consumers, which could have easily been done in the first place.

Agreed. While I’m happy to see Tesla aggressively pursuing autonomy and AGI, the idea of delivering a true RoboTaxi (i.e. a vehicle without foot controls and a steering wheel) seems pretty far off to me.

Even if all the technical pieces were in place, I simply don’t see regulations allowing it. Tesla currently has to deliver vehicles with rear-view and side-view mirrors, despite the fact their camera system is superior. Tesla has to include mirrors because of industry regulations, and I don’t see this changing anytime soon, especially with regard to vehicle controls. Of course none of this is new, and I’m sure Tesla knows regulations change at a snail’s pace. I have to believe a consumer version (includes manual controls) of RT is still in play. This seems like the most reasonable solution until regulations catch up.

The RT reveal can’t come soon enough. I can’t imagine another four months of speculation.

 

[jerry33]

Do people really do single store "Grocery Runs"?

Yes. I seldom go to more than one store. 95% of shopping is done online.

 

[MD70]

Not sure if it's been discussed, but how is Tesla planning to grow volumes in Q2 and the second half of 2024? They just ended inventory discounts and raised prices $1k on April 1st. They previously had a demand problem (supply > sales) as @Troy has noted since late 2022. Increasing price only decreases demand if you aren't materially changing the product you're selling.

Any theories on how (or if) Tesla will grow unit volumes the rest of the year?

I don’t see any proactive plan to do this, some think interest rate cuts will help but I suspect in a few months incentives and discounts will be back.

 

[jerry33]

Well, it could make more sense to get used EV that was $25k when new. But I agree...if RT were really reliable, then 2 cars between 3 or 4 people could be doable and may make sense for a lot of people. But they would have to be much more reliable compared to what Ubers/Lyft is like in my area.

100% agree. Almost every time we've called Uber they never showed.

 

[FreqFlyer]

Bodes well for Q2

https://twitter.com/x/status/1780604590213501167

 

[MD70]

A few thoughts...mostly that there are so many variables that are tough to quantify but often get completely ignored or downplayed when assigning "blame" for Tesla's timing.

1) I think part of the "delay" in the Model X was actually that the Model S had much higher demand than expected. Tesla and Elon have never made this connection, but I think the early expectation was that the Model S would have demand for about 10,000 per year. Model X was revealed in February 2012, before any S's were delivered -- the first S deliveries were in June of 2012 (at least according to some quick google work). The first Model X was delivered in late 2015, but in that in-between time, the Model S production was ramping up well above that 10,000 per year number. I suspect that, in real-time, Tesla was moving resources and plans around to ramp up the Model S processes and put fewer resources into the Model X. No real sense in spending money on the difficult task of adding a second model when you haven't yet done the somewhat easier work of making production meet demand on the existing model.

2) The timing of the Model 3 was still reliant on battery availability...which really meant Giga Nevada and Panasonic. We know that during the Model 3 ramp, there was some sloshing back and forth, with Tesla's production hell sometimes being the limiting factor, and Panasonic's ramp of batteries sometimes being the limiting factor. Of course, I have no idea how things would have played out if the Model X had not come before the Model 3...but Giga Nevada was truly needed before the Model 3 would make economic sense.

3) I think sortof similar with Cybertruck and Model 2 -- the Model 2 also needs even more massive quantities of cheaper batteries, and preferably US-made for the tax credit reasons, at least in the US market. So, yes, there was some resource draw to the cybertruck, but I also don't ignore the fact that even if Tesla had put everything into Model 2 a few years ago, it's not like there were enough batteries available to make a million Model 2's in the US in, say, 2022 or 2023. Obviously, several years ago, Tesla started work to design/build their own cells, while simultaneously saying they would buy any and all qualified sells from suppliers too. So, what more could they do to get more cells in 2022 or 2023?

With regards to point 3, as soon as the model Y scaled up, Tesla could have made arrangements with the battery companies for more battery’s in a few years whilst working on designing a more “normal truck” And a smaller/cheaper vehicle.

In such a timeline we might have had both vehicles selling in large numbers now and ramping up to meet demand now.

 

[2daMoon]

In any "long wait" scenario the Robo-taxi would not likely hang around for you (and you wouldn't want to pay for that), but (much like an airport) it could hang around for the next fare - imagine a giant Costco parking lot on a Saturday with RTs coming and going (dropping person A off while picking up person B).
Regardless, the taxi might just sit idle (w/o anyone paying) until called for in very low traffic times. As to the model where you are sitting in the car, doubt that would be feasible. In the interest of my own example of that, we have often done multi-store runs where I am little more than the, well...taxi-driver for my wife (i.e. taker her to Ulta, then to Soft Surroundings, then to get prescription, etc.)! In those cases, heck, I'm not even needed...Robo-Taxi to the rescue. I assume later Optimus will replace me entirely...

Consider the logistics of several shopping stops from a frame of reference of what to do with the goods purchased at each stop.

Will you remove them from the RT and carry them with you? Maybe, sometimes. Other times it will be best to leave them in the vehicle and have it wait for you as you bring more cargo back.

 

[2daMoon]

I would expect Tesla to do more with inductive charging so a RT could easily charge itself.

Tesla really should just find a company that has figured that out and purchase them.

Oh, wait, they already did.

 

[Knightshade]

Panic to sell, or panic to buy? All they really need is maybe a few more cameras to cover the current blindspots, and a way to keep them clean in all weather conditions. It's not comparable to rocket science, like the software successfully being capable of FSD is.

I'm torn on this.

On the one hand it'd be very bullish because Tesla is finally admitting the 2016 config of cameras is insufficient for reliable >L2 operation and perhaps THIS TIME they actually know what IS needed.

On the other it's a tacit admission of the by-then-6-million Teslas on the road, 0 will ever be robotaxis and there'll be some significant liability overhang there... (I think once litigated it'd be only the pre 3/19 buyers actually owed something so the financial hit will be insignificant compared to upside of a working RT, but the PR will be pretty awful and a lot of existing owners are likely to... change their opinion of the company)

Bodes well for Q2

https://twitter.com/x/status/1780604590213501167

Why? Multiple thousands in inventory discounts didn't seem to help Q1 any- so why would a $30 monthly lease discount, which is only $1,080 total discount over the 3 years of the lease.