A pretty damning read on Tesla

Investment Analysis Clubs Macro Economic Trends and Risks
101

Sure, but Syke6 was making the very valid argument that OEMs need to spend a long time and lots of resources to create a new model, so I think it reasonable to ask whether Toyota using 16 models to get that extra 23% is really cost effective.

ICE is a mature technology so I understand that one of the few ways to expand the market is to go after the niche customers with more specialized models. BEVs are at a very different part of the product development curve. Tesla understands that the most significant factors limiting BEV sales are price, range, and charging times and that what is needed to stimulate sales is the mass production of a cheaper, better battery and autonomous driving software. So those are the areas of focus. Adding more models now would just be a waste of resources.

No offense, but I think you have a naive view of the energy storage market. This is not like the batteries you put in your flashlight or what you use to start your car. The functionality of the Tesla megapack stems largely from the controlling software and the engineering of the packs. This is not to say CATL won’t be serious competition, but commoditized? I don’t think so. It’s like the individual components of the iphone are all commoditized, but the value of the iphone itself is much higher due to the software and Apple product ecosystem.

Tesla has spent a decade or two developing the software ecosystem for energy storage. That’s the real value of the Megapack:

Tesla’s suite of optimization software solutions, Autonomous Control, is composed of machine learning, forecasting, optimization and real-time control algorithms used for utility bill reduction, demand response participation, microgrid control and wholesale energy market bidding. Tesla Autonomous Control algorithms automate the dispatch of energy assets to maximize economic value. Autonomous Control products driving value to customers today include: Autobidder, Opticaster and Microgrid Controller. Tesla has also developed software to enable more renewable generation on the grid with features such as Virtual Machine Mode. https://www.tesla.com/support/energy/tesla-software

Throw in Tesla’s serious investments into AI and supercomputing and there is at least the potential for synergies between Tesla’s core business that could give it a huge advantage over a pure-play battery maker like CATL. Tesla may very well buy batteries from CATL, but the massive energy storage systems will very likely remain a Tesla product.

1 Like
102

It can be. Partially because catching that 23% contributes to the 77% (selling people a cheap first car can build a lifetime of brand loyalty). Partially because several of those models are important products in other markets - you’re not moving a lot of B- and C-segment small and compact sedans in the U.S., but they’re the lion’s share of sedans in Europe and many Asian countries. Partially because some of those models are ones that were more important to the company in other times and are just sunk costs (the two Prius models, for example, analogous to Tesla’s continued manufacture of the S and X). And partially because that’s still a lot of cars - 23% of Toyota’s U.S. sales is nearly half a million cars or well north of $10 billion in revenues, enough to be a Top 15 automaker in the U.S. just on it’s own.

Certainly that brings some value to the table…but a huge advantage? The software that goes into a battery storage system is certainly important - but so are, you know, all the batteries. When the company that supplies the batteries for your Megapack decides to enter the market with a competing product, it should temper your expectations for the role that energy products will play going forward. As well as when your company shifts away from promoting its technological advancement in developing its own batteries to focus primarily on autonomy, robots, and AI.

Again, Musk used to describe Tesla as an energy company focused on the transition from fossil fuels to electrification; he now describes it as an AI/robotics company that makes cars. Battery Day has given way to AI day and Autonomy Day. There’s no doubt they’ll keep plugging away with the Energy division (heck, they still nominally do solar stuff). But the idea that Energy might be a $200 billion segment (roughly what SpaceX is worth) doesn’t seem very likely, especially with CATL and BYD leading the way in reducing battery costs and Tesla using them as suppliers.

2 Likes
103

Tesla ER next week 23Apr, Tuesday, After Market Close.
Gonna vote for Elon retro-pay and re-incorporation to TX.
Lots of associated drama.
Tesla is ALWAYS volatile leading into an ER.

Last ER was “bad” due to deliveries miss.
Elon said all 2024 gonna be lower than “expected”.

Elon hired AI guy from Tesla to xAI. Implied threat to move AGI value away from Tesla.

Elon fired 10% of global workforce. Lots of drama over that.

Optimus.
Dojo.
AI compute infrastructure.

FSD supervised was released widely.
Changes to robotaxi vs 25k compact car build location and development priorities.

LOTS of HARSH negative sentiment around TSLA n Elon.

What pertinent items did I miss?

How much of all that is already baked into the current stock price?
That’s the question.

What do you mean by “speculate”?

Watch the YouTube that @captainccs recommended.

Herbert n Hans suggest the TETAR be tempestuous… For 5-6 years.

IMO there’s more down incoming.
The shortterm, 12-18 months, TETAR outlook is TEMPEST.

:tornado::cloud_with_lightning_and_rain::tornado::cloud_with_lightning_and_rain::tornado:
ralph

I purposely skipped mentioning X, Boring, SpaceX, Neuralink, etc.

2 Likes
104

Invest based on momentum and sentiment and not fundamentals. I.E. gambling.

Two prior times of direct Tesla investment were based solely on speculation and perception of investor sentiment. First time it worked really well (bought at $300, sold at $380 a few months later), second time it didn’t (bought at $260 and sold at $180). Ended up roughly breaking even.*

*From memory so numbers might be off.

Sentiment is clearly negative right now (hence the falling knife) but absent total bankruptcy, I would expect that to change at some point and the fanbois should push the stock back up with exuberance. I am not a fanbois but I have no objection from taking advantage of such.

1 Like
105

Well, I don’t know about that. I couldn’t sell one to my household because our idea of a perfect body shape is a Saab 9-3 Aero Sport Combi.

3 Likes
106

So I’ve had a thought rolling around which is appropriate for this already overlong thread, but I was reluctant to put it here until I had fleshed it out a bit more.

And it’s this: I’m told that the public will increasingly decide to buy cars (and/or manufacturers will increasingly design cars) by software choices, and that eventually that will winnow the field to just a few competitors, sort of like Apple & Android.

While I guess that could happen, here’s why it’s almost surely wrong.

  1. Why I could happen: If one manufacturer gets hold of and can maintain exclusivity over “a killer app” for a long period of time, this prediction could come true. Full Self Driving, actual Level 5, and at some reasonable price and years of exclusivity would give that manufacturer a monstrous lead and pull millions of customers along with it. See: VisiCalc for Apple ][, or maybe Grand Theft Auto for the PlayStation. There are others, of course, but lots of stuff is written multi-platform, but not infinitely-multi-platform, which is why the underlying OS tends to narrow to a select few.

Other than FSD, are there other killer apps - which cannot be replicated on other Auto-OS platforms - on the horizon? Not that I see, particularly. In fact, looking at some of the new cars out* I’m finding near parity.

My new Hyundai has lane follow, smart cruise control, emergency braking, auto wipers, regenerative braking, and of course GPS, satellite radio and other updates as have become common among all manufacturers these past few years. Someone else posted that their KIA has, uh, lane follow, smart cruise control, emergency braking, auto wipers, regenerative braking, and… well. A buddy of mine has a Porsche, which has lane follow, smart cruise control, emergency braking, and… stop me if you’ve heard this.

I am participating in a Cars & Coffee this Sunday morning*, assuming it doesn’t rain, and will be with some Tesla owners, KIAs and others, although the big attractions are the muscle cars of yesterday, a few new models from all over, and zippy paint jobs and pick-up trucks with 9 foot wheels, probably. I will know more then, but until then:

“Software is at parity.”

That could change, Musks frequent proclamations about how many miles Tesla’s have travelled is not to be taken lightly, but then Google and GM and now Hyundai have their own projects going, and I assume China will not be left out, and perhaps the Europeans too, so … nobody knows where we are, including me. But:

  1. the reason OS’s reduce is because of independent softwares that have to ride atop the OS, and developers don’t want to port for 19 OS’s so they stick to the big two, maybe three, and they’re done. And the 19 shrink to 2 or 3 and the others are done-for.

It seems unlikely, (as I say absent FSD as exclusive) that model will be followed. Lane follow, smart cruise control, etc. seem simple enough to have already been programmed by multiple manufacturers. Trivial softwares for auto headlight dimming or putting AM/FM/Satellite on one screen is too laughable to mention.

In this game of “this car has that feature, so we better have it too” we’ve seen the proliferation of everything from automatic transmissions to power steering to lumbar seats to … well, you get the idea. What possible “killer app” is going to find itself on one and only one platform?

The automotive industry is monstrously big, it’s why every industrialized country wants to have at least one, and preferably more domestic manufacturers. France, Sweden, Germany, Poland, Netherlands. China, Japan, India, the US. Argentina, Brazil, Mexico and others south of us. They all produce home grown brands as well as imports; some good, some not, but they’re already using different softwares for their feature sets and trim lines, no reason they won’t continue - or if they do collapse into fewer, so what? They’re going to be differentiated by styling. By price. By color. By marketing. By distribution. And maybe a little bit, by software features, which will quickly be followed by competitors.

6 Likes
107

But that’s almost certainly not going to happen.

I mean, technology might prevent that from happening (for example, it may be that once overall AI tech gets advanced enough to let one company break through to autonomy, it’s advanced enough that lots of companies can do it to). But the main reason is regulatory.

You can’t have autonomy without regulatory approval. In the U.S., that means state approval at a minimum - and if you want to make a robotaxi (ie. a vehicle without steering wheel or brakes), you need federal approval as well. There are enormous political barriers to getting that approval in the U.S. - it’s no surprise that the Teamsters file an objection to GM’s petition to the NHTSA to allow the driver-less Origin to be built. The same is true in most other markets.

So Tesla (or any other first mover) isn’t going to be allowed to implement FSD if they’re the only one that has it. There’s no way they’d be able to get it done. The entire U.S. auto industry isn’t going to let one company get a boot on their neck, and the European and Chinese regulators are certainly not going to let a U.S. company get that kind of monopoly. Tesla seems to be walking away from the one strategy that might plausibly do a political end run around that - the Tesla Network - and heading towards robotaxis instead, so they can’t avoid that pinch point.^^

I think that’s a major reason why Tesla keeps offering to license FSD. I think they realize that if they keep it an exclusive, they’ll won’t get to implement it - or at least not until a few other companies have caught up and developed their own autonomous driver software packages.

^^ BTW, I think that’s a smart move. The Tesla Network never made much sense (a Model 3 or Model Y can’t close its or see if there’s trash on floor of the back seat), and wasn’t likely to get approval anyway. The play would be to get enough ordinary folks as stakeholders pushing for your new regulation so that there is grass roots pressure on the local politicians not to ban it. But all the incumbents have learned the lesson from what Uber and AirBnB did to the taxicab and hotel industries, and they weren’t going to get caught flatfooted the way those other incumbents were.

1 Like
108

I don’t think it is “the” problem, but it is likely “a” problem. People in the USA like SUVs, small, medium, and large ones. And you have to make what the people want. I think a nice medium sized Tesla SUV would sell quite well. Heck, I might even buy one to replace our minivan. I was originally thinking of replacing the minivan with a 7-seater SUV, but then I realized that we rarely have 7 people at home anymore, and even when we do, they don’t all come with us places. Then I was thinking of replacing it with a truck (and I have a cybertruck reserved), but a truck is very inefficient and not something I really need, and my wife wouldn’t feel comfortable driving something so big. So now I am considering getting a model Y to replace the minivan, but would rather something a little big larger than that. As an example, I will have to move a kid to an apartment near college later this year, and the minivan fits a LOT of stuff, a model Y very likely wouldn’t be sufficient.

1 Like
109

While these are issues, they aren’t the BIG issues that will prevent all model 3s and Ys on the road today from being a robotaxi. The biggest issue by far is that the vehicles can’t see what is directly in front of it. If there’s a kid playing in the driveway right in front of the car, it can’t see. If it’s at the airport and someone inadvertently left a carryon bag in the pickup lane right in front of the car, it can’t see it. Etc.

110

Criticism is great, but it needs to be connected to reality in some form. Toyota is the largest auto manufacturer in the world. Sales are increasing. Profits are increasing. Good margins. Not only that, Toyota is increasing their lead over the competition. Toyota is the global top dog and by all appearances things are going great.

I assure you, Toyota has warehouses full of accountants doing nothing else but calculating which models are making money for them. I straight up don’t believe your claim is Toyota’s key to success is to reduce their product selection. That claim is simply not credible.

Now, if their margins, sales, market share, and profits, were all falling like–I don’t know, say Telsa–then yes. Toyota should absolutely take a knife to their business model. Because clearly something would be amiss. But in the meantime, your suggestion they stop doing the things that made them successful makes literally zero sense.

8 Likes
111

The underlying premise of Goofy’s nicely made argument is that the next 20 years of automobile transportation is not going to be much different than it is now, just with a few more driver assist and entertainment options. And if that’s true then he’s right, the car market will be not much different because the technological advances we see with EVs and autonomous driving will, by definition, not have been disruptive.

The alternative POV, is that they are disruptive, with the most obvious disruption being that transportation as a service (TaaS) becomes a big deal. So big that it significantly reduces private car ownership. Households might still own one car but feel less need to own more than one. The growing senior demographic will welcome not needing to drive and young folks in school will welcome not having the responsibility or expense of car ownership. And even car owners will do far less driving in this disrupted future, as the car does most of the work.

People using TaaS won’t care about car styling or performance or many of the other hardware distinctions associated with models and trims. They will want comfort, convenience, connectivity, safety, and dependability. Most of this will depend on the software running the vehicle and just as importantly the software by which the robotaxi company is coordinating all its vehicles. Tesla already has this system in place as it has long been collecting driving data from all its cars. It’s supercomputer should be able to direct its robotaxi fleet to areas of high demand and coordinate their recharging and cleaning.

The winners in this disrupted future will be the companies that can most efficiently make the small number of vehicle form factors best suited for TaaS (IMO the infamous box-on-wheels) and companies with the AI experience and computing facilities to coordinate a complex robotaxi system efficiently. I suspect the latter group of companies will be small in number and dominate personal transportation in urban areas.

In short, those using TaaS could care less about models and trims. Car owners in this disrupted future will be mostly letting the car do the driving and so will be most focused on cars with the best self driving software (e.g., is statistically the safest) and the most comfortable and functional interior. In other words, the car as computer with software being the most important feature.

Apple vs Android or Windows vs MacOS.

2 Likes
112

You are absolutely correct if the world doesn’t change. Unfortunately the world does change, and these days it changes pretty rapidly. Successful companies stays successful by being willing to adapt their business model. Otherwise there is the risk of that Kodak moment.

To be honest I think Toyota has long been shifting to my side of the fence. As you and others have pointed out, it is typical for different “models” to use the same platform and the same engine. This means the the distinction between models is increasingly more about marketing and cosmetics than serous engineering. Toyota is actually making fewer truly distinctive models.

113

While Tesla is or was production constrained it does not make sense to add new models. Once EVs become commodities the competition shifts to this and that feature, i.e. models. This issue should be viewed in terms of where Tesla and the industry are terms of development and adoption.

But there is more, the question is not selling more cars but making decent margins and profits. The auto industry is a phase-transition mode from selling hardware (decreasing returns) to selling software (increasing returns). The recent change of plans at Tesla from the mass market car to the RoboTaxi is an indication of this phase-transition. The mass market car is the right choice to sell more hardware while the RoboTaxi is the choice to sell more software but it has additional risks, will the software be good enough, soon enough to get the necessary regulatory approvals. Besides hardware is more standardized world wide than driving conditions which complicates adoption (localization).

A recent thread mentioned the divergence inside Tesla between the more futuristic Elon and the more conservative executives.

I read this after posting! I see that you are in the same frame of mind, TaaS vs. ownership.

The Captain

1 Like
114

I have so many questions! For instance, software is nothing without hardware. For Microsoft this was great (also great for Quicken, Oracle, et. al.), since somebody else shouldered the cost of designing the hardware, assembling the hardware, retailing the hardware, and finally purchasing the hardware. All they had to do was put the same software in a cardboard box and sell it for an exceptionally high price. Very high leverage business.

But the model you’re proposing for Tesla doesn’t give Tesla “software margins”, because they still are doing all the hardware with all the associated costs. It’s like Apple, in the early days, doing both hardware and software, which Bill Gates advised Steve Jobs was an inferior model.

Unless you’re saying that other car manufacturers are going to make cars which Tesla can then ‘import’ their robotaxi software into. That seems unlikely, since other cars will have different hardware and different sensors, in different locations in the vehicle, with different OS configurations to meet. That’s possible, I suppose, but will require a lot of hand-holding, not to mention regulatory compliance hoops.

As far as TaaS, I wonder, doesn’t that exist already? Can’t people call one of the 5 million Uber drivers, or 2 million Lyft drivers, or in some cities the hundreds of thousands of taxi drivers or limo drivers already? So … they can call a Robotaxi instead? The difference is that there isn’t a driver to pay - but there are still all the associated costs of maintenance (lower, maybe), fuel (lower), capital investment (same), management (more), overhead (same), depreciation (same). So by my very rough back of the envelope calculation we’ve eliminated … the driver, who as most anyone will tell you, is an underpaid schlub possibly making minimum wage. And that’s going to change everything?

I mean, if that’s going to convince people to go from two cars per family to one, why isn’t it happening already? Or is TaaS, in the famous epithet about nuclear power, going to be “too cheap to meter”?

And while I agree that styling would be less important, there will still be a segment for limos, for handicapped, and so on. I mean, who wants to pull up for prom night in a ‘98 Toyota?

How does Tesla’s collection of driving stats from owners have anything to do with where TaaS will have high demand? That kind of data is owned by Uber & Lyft, not Tesla. Maybe they will sell it to Tesla if they want to cut their own throat. By the way, the “charging and cleaning” will surely be less efficient than a human driver, who can gas up or clean up within a couple blocks of the need, whereas a robotaxi will necessarily have to drive to a more distant location, sort of like finding a charging station now. Which, I point out, will have to be manned 24/7, so they are unlikely to be found on every block.

Ah, so many questions. I’m sure you guys have answers. Let’s hear ‘em!

5 Likes
115

I think this is part of the reason we keep having these conversations. Cars can share a platform and engine, and still be very different from each other. For example, the Toyota Camry, Rav4, and Highlander are all built on the same platform. But they’re very different cars, appealing to different consumer preferences. The Camry is a mid-size sedan, the RAV4 is a compact two-row SUV that seats 5, and the Highlander is a mid-size crossover three-row SUV that seats 8.

That’s not just marketing or cosmetics. I’ll grant that there’s not a ton of serious engineering that goes into switching an SUV body instead of a sedan body on the same platform (or the larger SUV body instead of the smaller SUV body) - but that doesn’t mean that the result isn’t a very different product for the consumer.

I think you just seriously underestimate the degree to which consumers genuinely differ on their preferences between cars like the Camry, RAV4, and Highlander - which all have different sizes and attributes and body types and price points, even as they are built on the same platform. It might be unimportant to you whether you’re driving a Camry or a Highlander, but most consumers aren’t like that.

3 Likes
116

Nowadays it seems more like hardware is nothing without software. And not only nowadays, I’ve given this example before - betamax was better “hardware” than VHS, but VHS won the market because of better “software”, it had the killer app.

Or a more recent example. Nvidia has been making graphics chips for a long time, and they did make ones that the market wanted more than most other ones. BUT they only took off once the killer app came along, and that killer app was crypto creation. And THEN, they got even luckier when a SECOND killer app came along, and that second killer app was AI. Now they are attempting to cement their position by actually owning a lot of the underlying software that makes their hardware so valuable.

1 Like
117

Smarter people than you (and I) have made such calculations.

image
image833×625 55.8 KB

Because TaaS today is more expensive than owning a car. Do you not understand the argument? If autonomous driving doesn’t reduce the cost of TaaS then you’re right, nothing much changes. If it does significantly reduce the cost of TaaS, then there is a real possibility of significant disruption. Owning a car may no longer be the cat’s meow for most people as it is currently to you.

Teslas are sufficiently popular that where Tesla drivers go is probably indicative of overall transport demand.

Just speculating, but one possible answer is suggested by the Tesla deal with BP to place superchargers at BP stations. As the EV revolution continues gas stations are going to be facing declining gas sales. They can compensate by acting as charging and cleaning stations for robotaxi fleets. Would be a bit of poetry if a business that started out as full service at the beginning of the ICE age ended up returning to full service as the ICE age ends, with gas sales essentially subsidizing the Robotaxi service.

1 Like
118

You left out the Lexus ES. Yes, that’s a very different car, built on the same platform, appealing to a very different consumer.

Working in a plant that makes both the Camry and the ES I feel qualified to answer. While yes, the cars share a platform and a great deal of parts, they are far from being the same car. We make a Camry every 68 seconds on my line, and last I knew they made a Lexus ES every 4–5 minutes. The people assembling the Camry know one or two jobs and that’s about it, whereas the people assembling the ES are more highly trained and have a greater number of jobs they perform. While Toyota’s build quality is baked into each process, they pay extra attention to each Lexus they assemble.

Getting more into the cars themselves…

The Lexus uses much more sound deadening than does the Camry, and higher quality interior materials. They are equipped with more technological features and even the paints are nicer, both in color and finish. Some parts of the chassis are revised to provide a better or quieter ride. It’s rumored among the facility that even the engines going into the Lexus get more attention than do the ones going into the Camry, even though they carry the same model designation.


https://www.quora.com/Is-there-a-shared-platform-between-Lexus-and-Toyota-cars#:~:text=Are%20Lexus%20cars%20basically%20the,from%20being%20the%20same%20car.

119

The reason we keep having these conversations is that you keep stretching my position to the extreme to make your point. I’m not saying we should have only one model, I’m saying companies would be better off with fewer models. Small and large versions of a sedan, SUV, and pickup will cover the great majority of one’s customers. That’s six models.

On the contrary, most consumers are like that. That’s why only a handful of model types account for the large majority of car sales.

2 Likes
120

Of course I understand the argument. I just don’t believe that it will reduce costs enough to get people to abandon private vehicle ownership. Mass transit certainly doesn’t, and it’s the cheapest of all. (Yes, less convenient, too.) Sure, there will be some who eschew the car for affordability reasons - temporarily. College kids, elders in nursing homes. But the mass of civilization? Na gonna happan.

This is surely wrong, almost spectacularly wrong. Tesla’s are bought by wealthy people, the least likely to use TaaS, except perhaps on New Year’s Eve. TaaS is most likely to be used by economically deprived (see: college students & other young), drunks on a Friday night, and urbans without access to parking. If I was to draw a Venn diagram among “current Tesla owners” and “people who use Uber” I suspect there would be a microscopic intersection, at best.