Long-Term Fundamentals of Tesla Motors (TSLA)

[Krugerrand]

My expectations are based on historical data. This world used to be quite awful place in the past. We made great leaps in many respects. Progress is unstoppable, nothing to do with faith.

There are still plenty of places in this world that are awful, some located smack dab in the middle of those great leaps.

No argument that progress continues. It doesn't, however, ensure a good outcome. History has shown us that too.

- - - Updated - - -

Krugerrand - you might find the perspective provided in "Don't Panic" interesting: http://www.gapminder.org/videos/dont.../#.VDoCUjbn9es

Thank you. I'll check it out.

- - - Updated - - -

now maybe back to Long Term Fundamentals of Tesla Motors?

Nothing has changed. They've stayed the course. Perhaps accelerated it a bit in a couple of areas. All good. Hold. :wink:

 

[30seconds]

Consider this, the Model S already has a software layer between the driver and the mechnical systems of the car. The driver is only manipulating a controller, and software already transforms that input into action. This digital control affords much more that merely adding on safety features. Every aspect of performance can be modified through programming. So for example, dual motors, we now know, can be digitally controlled in such a way that 10 to 30 extra miles of range can be obtained. This involves having an algorithm to optimize the distribution of power from front to back. This algorithm could take in as inputs the current speed, the desired acceleration (from the driver), the angle of the steering wheel (also fro the driver), the curvature of thebroad ahead (from radar, camera and gps), the grade and slope of the road (precice maps and sensors), slippage in the wheels (from sensors), even weather conditions. All this can potentially be taken in to optimize power sent to fron and back motors. Now suppose programmers have just a simple model to work with. They program that up and tweak it until it works great. Future research perhapsnbased on real world data from customer carsnon the road may lead to formulating better models for integrating so many variable and deriving an optimal power assignment. No problem. Programmers gonto work on coding up the new model. It gets tested, and if it delivers better performance, it gets pushed out in a software upgrade. So now any Model S owners with sufficient hardware ge a performace upgrade over the air. This is much more than just delivering a certain safety feature or autopilot capability, this is about delivering high performance. Drivers will want Tesla's autopilot technology not because it lets them go to sleep at the wheel, but because it enhances their driving experience. Imagine being able to drive with the precision of highly skilled professional driver through a winding mountain road. You take each cure with confidence, you accelerate and brake in perfect coordination to make the most of every curve. You have incredible control the the whole way. The car responds impeccably to your slightes suggestion. All the while, it is the software that is intepreting your instruction, road and vehicle conditions and transforming it into something amazing. The potential of the digital car is not only to enhance the performance of the car, but the performance of the driver too. You will want to drive this car because it will feel amazing to be the driver. The P85D is amazing not because some professional driver can make it do 0-60 in 3.2 seconds, but because you can make it do 0-60 in 3.2 seconds.

What's more is that they will / could have access to real time data from the universe of local tesla's to inform real time traffic / road conditions, weather, etc. When going blue sky on autonomous driving I think a network of real time nodes will be more powerful than a collection of independent autonomous vehicles.

probably why elon's so interested in AI - if they got control of Tesla and SpaceX we would all be in trouble :wink:

 

[Robert.Boston]

I find it interesting to contrast how Tesla and Google approached automated driving.

Google is building a "big data/cloud computing" model, consistent with its core expertise. As I understand it, Google's system relies heavily on centralized databases that contain every detail of the road, augmented by on-the-ground observations such as location of pedestrians. There's an implication of enormous bandwidth between car and servers, big servers, and huge (and up-to-date) databases.

By contrast, Tesla relies primarily on "on the ground" data. What can the car see (in one spectrum or another) dictates how it behaves. The GPS link appears to provide only speed limit data. EM didn't mention the 3G connection as part of the system (although we know that the nav system uses information from other Model Ss about traffic conditions).

For my money, Tesla's approach makes much more sense, at least to achieve the somewhat more limited objectives of "autopilot" rather than "autonomous." First, it's clear that Tesla was able to make huge strides quickly, while Google has spent years with only limited results. Second, there is no cost to maintain huge server farms and the comm links between these farms and cars. Third, it's scalable: the car can function pretty much anywhere, whether or not someone has put that geography into its databases, and isn't subject to flaws when on-the-ground changes outdate the central servers.

People drive every day without accessing the master database of all roads; surely a good AI can do likewise.

 

[AlMc]

When EM mentioned on the last conference call that they had significant R&D expenses on 'cards that they were not showing' is it safe to conclude that it was the DA features in the new model S, and the upcoming model X or something else??? My opinion is that it was the DA but am open to other conjecture that might serve TSLA well long term.

 

[ItsNotAboutTheMoney]

When EM mentioned on the last conference call that they had significant R&D expenses on 'cards that they were not showing' is it safe to conclude that it was the DA features in the new model S, and the upcoming model X or something else??? My opinion is that it was the DA but am open to other conjecture that might serve TSLA well long term.

Everybody know they were working on AWD and Autopilot. I don't think it would be that.

 

[renim]

The aim of this model is not so much to predict the future as it is to understand the implications of competitive intensity, to explore sensitivities to the strategic choices of competitors, and to consider strategic options. To that end, I post a working worksheet model for readers to play with and a sensitivity analysis table to summarize outcomes under different strategic choices.

I agree that logistic growth is the correct model. It very well described the transition for horse to car. It currently seems to be describing the transition from oil to electricity for cars.
I disagree that logistic growth is the correct model for plugin hybrid / electric vehicle analysis, Historically the early EV,ICE and steam automobiles had differing market shares based more on other factors - quality of execution?

2013 global plugin vehicle sales were 210,000 EV Sales: January 2014
2014 global plugin vehicle sales to August are 192,000. So the year could end with around 330,000 plugin vehicles sold. So an annual 60% growth rate for the industry seems OK.

With the assumption of 100% marketshare for plugin vehicles (BEVs and PHEVs) vs ICE (Hybrids and plain), Tesla is currently looking around the 10% figure.

Mitsubishi was a pioneer in EVs, akin to Tesla Roadster in timing but 10x more production (due to assembly run continuing) but
Mitsubishi is far more successful in PHEVs than EVs
why, because with PHEV Mitsubishi was able to price their PHEV variant to compete well vs the AWD Diesel equivalent.
remember
battery tech is improving, so both EVs and PHEVs are improving
but
EVs benefit from cumulative improvement in infrastructure deployment.
PHEVs benefit from faster improvement in power and longevity that Li ion enjoys compared to energy improvements.

 

[jhm]

I agree that logistic growth is the correct model. It very well described the transition for horse to car. It currently seems to be describing the transition from oil to electricity for cars.
I disagree that logistic growth is the correct model for plugin hybrid / electric vehicle analysis, Historically the early EV,ICE and steam automobiles had differing market shares based more on other factors - quality of execution?

2013 global plugin vehicle sales were 210,000 EV Sales: January 2014
2014 global plugin vehicle sales to August are 192,000. So the year could end with around 330,000 plugin vehicles sold. So an annual 60% growth rate for the industry seems OK.

With the assumption of 100% marketshare for plugin vehicles (BEVs and PHEVs) vs ICE (Hybrids and plain), Tesla is currently looking around the 10% figure.

Mitsubishi was a pioneer in EVs, akin to Tesla Roadster in timing but 10x more production (due to assembly run continuing) but
Mitsubishi is far more successful in PHEVs than EVs
why, because with PHEV Mitsubishi was able to price their PHEV variant to compete well vs the AWD Diesel equivalent.
remember
battery tech is improving, so both EVs and PHEVs are improving
but
EVs benefit from cumulative improvement in infrastructure deployment.
PHEVs benefit from faster improvement in power and longevity that Li ion enjoys compared to energy improvements.

Renim, thanks for looking at this.

I agree there are fits and starts with this. We could even go back to GM's EV1 and see that didn't lead into persistent growth. So there is a bit of a challenge knowing when something has enough critical mass to take off. It think one of the basic probles is that most efforts to make EVs have simply been for regulatory and PR motives. So there has been no serious strategic commitment to them and they have not been designed to make money on their own without regulary incentives. Naturally those sorts efforts are not likely to be sustained. My view is that the Nissan Leaf has some staying power, but the only company with real strategic commitment to EVs is Tesla. So that is basically my starting point. But who knows, Nissan could pull the plug on the Leaf tomorrow or when the Model 3 comes out. I do think that PHEV are promissing in that they indicate consumer demand for a car that can source power from something other than a gas motor. But again at what point do we cross the line from a halfhearted compliance car to something that really wants to compete in the marketplace.

I don't really see this as a modeling problem for the logistic model. It's more of a problem with initial conditions. It is really hard to be sure that appart from Tesla, EVs have truly begun to compete. One variation of this model would be to build separate models for different kinds of vehicles. For example, we could have a model just for sedans, one for SUVs and another for trucks. We've got initial conditions for sedans, but so far Tesla has no SUV and I don't think there are any EV trucks out there. So these to other models are just going to have to wait before there are initial conditions adequate for logistic growth to commence.

Of course you are also correct that developments in battery tech are fundamental to this transition. It seems that for the most part the auto industry is content to wait for someone else to solve the battery problem for them; whereas, Tesla has been willing to take stuff pretty much off the shelf and make it work. In principle any car maker could do what Tesla is doing; they have even openned up the patents to remove any excuse. From the modeling side it could be interesting to formulate a transition model that based on batteries. One of the special challenges PHEV and small BEV makers may have is trying to increase battery sizes without sacrificing unit sales. If you're growing unit sales by 30% and battery sizes by 20%, then you've got to grow battery capacity by 56%. Tesla is coming in from the other angle. It's leading with big batteries. It does not have increase battery pack size to compete for more unit sales.

In any case the market share model is still useful in that it shows what can happen if there is poor comptitive intensity. If Nissan and others are not really competing for EV market share, the industry is even in a worse situation than I'd give them credit for.

BTW, in my set up I include PHEV with other hybrids. I do not see PHEV as being an aborbing state. If I were to have a separate segment just for PHEV I would still want to assume some net transfer of market share from them to Tesla. Have you played around with what happens if you assume a smaller transfer intensity from H (including PHEV) to E and T? You should be able to make H persist a little longer. Admittedly, I am at a loss at this point to find a reasonble way to calibrate those parameters. We ay get a feel for it when the Model 3 comes out. If the growth of PHEVs slows down at that time, we'll know there is a lack of parity. Another way you could use my model as is, is to go ahead and assume PHEVs are included in E and adjust the initial E level, but set the transfer parameter from E to T to something like 10% or 5%. This way you can see how strong the parity assumption is over the long run.

 

[chickensevil]

Everybody know they were working on AWD and Autopilot. I don't think it would be that.

Actually I wouldn't put it past them that this was most of the money they were spending. This isn't just two motors and a couple sensors. Making it this far, this fast on the autopilot is better than most imagined... We weren't supposed to get this until the end of next year... Talk about one year ahead of schedule.

And the two motors thing is more than just that. They changed the breaks, they changed the seats, and they also managed to tune the car to hit 3.2 seconds... This takes a lot of work to hit those numbers shaving off a full second off their posted time on the old Performance model.

We know they are working the X and they are very much likely working on the 3. So if they are also working on something else that they haven't mentioned then I wouldnt have a clue where to start guessing.

People have thrown out crazy things in the past, but serious claims toward what they could be working on extra would maybe fall under the relm of a yet unspoken fourth model... But I really think it is too early for that.

 

[jhm]

@jhm many thanks for sharing your skilled and insightful model. It's people like you who make my nights interesting!

Like many here, my first reaction was that 2.75% general growth rate is much too high and it will actually begin to decline in a few years. There is already a trend to migrate to cities where a personal auto is a liability. My college-age kids consider cars to be a burden which is pretty much the opposite of how I felt at their age. I don't think developing countries will make up the difference because most of the population and economic growth is in cities. We'll see. In any case, playing with the General Growth Rate in the spreadsheet produces interesting consequences. If I change it to 1% which is too low for the short term, possibly too high for the long term, Tesla and EV Peak sales drop by almost half.

I don't fully understand the General Attrition Rate.

Right if you reduce the general growth rates to 1%, the total number of cars sold in 2050 is about half. But note that the share of EV and Tesla do not change. The general growth rate does not impact market share, just the size of the maket.

For projecting changes to the fleet or parc, I have to account for cars taken off the roads (attrition) and sales. So the change in fleet is new cars minus cars taken out of service. I'm assuming that attrition is 6% per year. This is consistent with an average useful life of 16.67 years (=1/0.06). And at 12k miles/year, this is also consistent with a 200k mile life per car.

Thanks for playing with the model!

 

[Arb1ter]

Tesla has been very cagey about their future battery technology. Given the importance of batteries, and the amount of money they are spending on building them, I'm willing to bet that what Tesla reveals for future products and refreshes is going to be a lot more impressive than most will expect.

 

[30seconds]

Great article on the ongoing IHS teardown of a Model S. It really shows that Tesla had a very deep role in the design and production of all the components (so far), which is unlike traditional automakers. I think this has given the company a number of long-term advantage over legacy automakers such as - 1. rapid iteration & production roll out, 2. diagnosis and problem solving 3. systems integration & control software 4. (my favorite) no way for automakers to cross shop Tesla components.

For example since automakers have basically stopped designing transmissions the ZF variants are now used across brands. What was once a competitive differentiator in the industry is now more of a commodity. The easiest example is that Tesla batteries are completely controlled by tesla and not available to other automakers, but the same can be said for the invertor, motor, display systems, cooling system, superchargers, etc. Of course some components are sourced from the same third party vendors (electric steering I believe), but overall this is a Tesla effort. This becomes even more important when you consider how Tesla has been able to integrate the entire car via their software layers.

http://www.streetinsider.com/SI+Newswire/Is+it+a+Car+or+an+iPad%3F+/9906767.html

 

[eepic]

Here's an excellent article on Forbes about the shift towards autonomous vehicles in the coming years, and Tesla's likely role in being at the forefront because of its pace of innovation. Much less outlandish tone than Adam Jonas (Morgan Stanley).

Driving Disruption: Tesla is Building Cars Out of Software - Forbes

 

[neroden]

I find it interesting to contrast how Tesla and Google approached automated driving.

Google is building a "big data/cloud computing" model, consistent with its core expertise. As I understand it, Google's system relies heavily on centralized databases that contain every detail of the road, augmented by on-the-ground observations such as location of pedestrians. There's an implication of enormous bandwidth between car and servers, big servers, and huge (and up-to-date) databases.

By contrast, Tesla relies primarily on "on the ground" data. What can the car see (in one spectrum or another) dictates how it behaves. The GPS link appears to provide only speed limit data. EM didn't mention the 3G connection as part of the system (although we know that the nav system uses information from other Model Ss about traffic conditions).

For my money, Tesla's approach makes much more sense, at least to achieve the somewhat more limited objectives of "autopilot" rather than "autonomous." First, it's clear that Tesla was able to make huge strides quickly, while Google has spent years with only limited results. Second, there is no cost to maintain huge server farms and the comm links between these farms and cars. Third, it's scalable: the car can function pretty much anywhere, whether or not someone has put that geography into its databases, and isn't subject to flaws when on-the-ground changes outdate the central servers.

Yes. Tesla's autopilot also is quite capable of turning off -- saying "I'm confused by this road suddenly turning into an unmarked dirt road covered in traffic cones, driver needs to take over" -- just like the autopilots in airplanes to. And I think this will continue to be necessary in cars pretty much forever. Google is fantasizing about fully autonomous cars, which will never happen, because there is simply too much variation in conditions -- the whole advantage of cars over trains is that they can be driven through these random conditions (I park on grass fields fairly regularly). Tesla is implementing what can actually be implemented now, not fantasizing about automation.

- - - Updated - - -

Actually I wouldn't put it past them that this was most of the money they were spending. This isn't just two motors and a couple sensors. Making it this far, this fast on the autopilot is better than most imagined... We weren't supposed to get this until the end of next year... Talk about one year ahead of schedule.

And the two motors thing is more than just that. They changed the breaks, they changed the seats, and they also managed to tune the car to hit 3.2 seconds... This takes a lot of work to hit those numbers shaving off a full second off their posted time on the old Performance model.

We know they are working the X and they are very much likely working on the 3. So if they are also working on something else that they haven't mentioned then I wouldnt have a clue where to start guessing.

Well, they're also definitely working on battery process improvements and the Gigafactory. And although they've talked about the location of the Gigafactory, they've been quite cagey about the design manufacturing changes planned for the Gigafactory. They're also planning to source the minerals from North America using ecologically sound methods, and frankly I think R&D is going to be required in order to do *that* cost-effectively. The pickup truck was casually mentioned once as well (and it would be a coup to be able to send the rangers out in Tesla trucks). They could also be keeping a new convertible design under wraps (they mentioned this before but then went quiet about it). They could be doing R&D on *car factory processes*, which they had quite a struggle with at first; that could cost huge amounts of money. And there's also Supercharger R&D; there are probably improvements to be gotten out of that. And the battery-swap operations, which they've also mentioned. Those are just things they've *mentioned*.

Then there's the delivery system for the cars; currently it is not even as efficient as the standard automakers' and a lot of R&D could be spent on process efficiency. They have to throw money into license compliance for their software whether they like it or not, which is going to be listed under R&D.

I don't think I've been very speculative here, but that's a hell of a lot of things to burn R&D money on.

 

[austinEV]

Yes. Tesla's autopilot also is quite capable of turning off -- saying "I'm confused by this road suddenly turning into an unmarked dirt road covered in traffic cones, driver needs to take over" -- just like the autopilots in airplanes to. And I think this will continue to be necessary in cars pretty much forever. Google is fantasizing about fully autonomous cars, which will never happen, because there is simply too much variation in conditions -- the whole advantage of cars over trains is that they can be driven through these random conditions (I park on grass fields fairly regularly). Tesla is implementing what can actually be implemented now, not fantasizing about automation.

I think society will meet automated cars halfway. If you talked to someone in 1920 and said "I think someday urban streets will have cars driving 40-50 mph", they might have said "that's impossible. what if a pedestrian, child, or horse stepped into the roadway? That speed wouldn't allow you to stop in time." The cars didn't change, everything else did. We just understand from a young age that roadways have high speed vehicles moving on them and we cannot go there. We move pretty quickly to remove road debris for the same reason. Roads became these high-speed corridors that are for 50mph cars and for nothing else. In the future this trend will continue. Road crews won't just put out cones and assume things will be ok. They will file a plan with some mapping clearing house so that the computers know to look for the change, and then the cones may not even be necessary (or they would be designed such that cars' vision systems can pick them up and drivers can quickly see that they are the right such cones). Similarly if you leave an urban area there would just be a clear demarcation of where you need to drive. Your car would say "autopilot ending in one mile" and force you to drive before that happened.

In other words, we will make the roadways match the new cars. The cars won't have to figure out everything.

 

[Zzzz...]

I think society will meet automated cars halfway. If you talked to someone in 1920 and said "I think someday urban streets will have cars driving 40-50 mph", they might have said "that's impossible. what if a pedestrian, child, or horse stepped into the roadway? That speed wouldn't allow you to stop in time." The cars didn't change, everything else did. We just understand from a young age that roadways have high speed vehicles moving on them and we cannot go there. We move pretty quickly to remove road debris for the same reason. Roads became these high-speed corridors that are for 50mph cars and for nothing else. In the future this trend will continue. Road crews won't just put out cones and assume things will be ok. They will file a plan with some mapping clearing house so that the computers know to look for the change, and then the cones may not even be necessary (or they would be designed such that cars' vision systems can pick them up and drivers can quickly see that they are the right such cones). Similarly if you leave an urban area there would just be a clear demarcation of where you need to drive. Your car would say "autopilot ending in one mile" and force you to drive before that happened.

In other words, we will make the roadways match the new cars. The cars won't have to figure out everything.

Yes. Furthermore no need to solve 100% of driving problem right away. Even if self driving car would be able to drive around Mountain View only - it will have it use cases. Add new cities - even more use cases for autonomous driving.
No need to create car that would drive in any season/weather condition and on every road possible anywhere in the world. Start with one small city, daytime in good weather driving. It would be useful as is. And from that point add new cities, highway driving, nighttime one. Then add winter/snow condition, rain etc.

 

[green1]

The problem with Google's self driving car approach is they've targeted all or nothing. Tesla has done this right, introduce the parts that work as you have them, and don't worry about the others until the technology gets there.

 

[jhm]

+1 It's as if what they are really after are spy drones that cruise around your neighborhood collecting data. Their vision is too much about putting everything on the internet rather than truly autonomous solutions. That is they want cars and everything else fully dependent on what they provide through the internet. A truly autonomous vehicle would not need to be in constant contact with the Internet to get from point A to point B. It would have localized intelligence and not need anything from Google.

 

[eepic]

Can anyone explain this price action? Going steadily downwards, against the NASDAQ, on very low volume. This on a background of strange Adam Jonas pronouncements, divesting from major holders, and the occasional very large lot coming out of nowhere, like the one today around 11:52.

I feel like the poker player trying to understand who the sucker is, and I don't like the feeling.

Just saw this article come out about Porsche building an electric car to compete with Tesla. In the medium term, I think it's good more traditional auto companies are hopping on the EV bandwagon as we'll get to critical mass for the ICE car death spiral sooner. Maybe it's headline reaction causing the price drop.

Porsche to take on Tesla with electric version of all-new model | Autocar

 

[anticitizen13.7]

Can anyone explain this price action? Going steadily downwards, against the NASDAQ, on very low volume. This on a background of strange Adam Jonas pronouncements, divesting from major holders, and the occasional very large lot coming out of nowhere, like the one today around 11:52.

I feel like the poker player trying to understand who the sucker is, and I don't like the feeling.

In my experience, individual investors almost always end up holding the bag when it comes to the short term game. Between attempts of larger players to manipulate the system, the unpredictable results of bot bashes, and sheer irrationality, the short term game is a lot like gambling.

I say trust your instincts and don't play their game.

 

[Chickenlittle]

Can anyone explain this price action? Going steadily downwards, against the NASDAQ, on very low volume. This on a background of strange Adam Jonas pronouncements, divesting from major holders, and the occasional very large lot coming out of nowhere, like the one today around 11:52.

I feel like the poker player trying to understand who the sucker is, and I don't like the feeling.

I suspect people nervous with high fliers net flix and Amazon drops. I do not link tesla to either of these but people get nervous