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

[humbaba]

This is exactly right. Tesla's conclusions about safety have never been supported by the data. I've been disappointed that they've presented their results this way, as it's obviously wrong.

My belief from what I've seen is that auto-pilot is a good thing for safety in its current state, but not much. However it will only get better.

I know a traffic engineer who contended that Tesla's were far less safe than their peers. Because his job is traffic safety he had access to data for making that declaration. However, he made no effort to back the claim up and is an EV denier which makes his claim suspect without supporting evidence. As just a sampling of things he has told me:

1. grid cannot support EV
2. gasoline IC engines have 60% efficiency
3. EV energy cost is more per mile

(and that's just fallacious claims about EV, not other bizarre claims)

I'm not saying any of these claims should be taken seriously (I did some digging on the ICE efficiency, and the best I could find for automotive was 38.5% peak with a theoretical peak limit of 46% that varied with what you took for parameters, but a 60% theoretical limit for turbines) and am using them to demonstrate the grain of salt his claim about safety should be taken with.

Still, he does have access to data we do not and so should not be entirely discounted.

 

[JustAaronnoraA]

Do you mean you think that only the 100D and P100D will be sold?

That would be awfully risky:

• According to @Troy's tracker, the 75D S/X is about 55% of sales. (!)
• Q1 is the seasonally weakest quarter, with the softest demand.
• Speculation has been swirling for months about an interior refresh, about 2,170 adoption. There's been a leak a couple of months ago about the new S/X interior. Everybody expects HW3 inclusion in the Model 3 in March/April.
• The P100D was introduced in the summer of 2016. The competition had 2.5+ years to catch up.
• Most tellingly, Tesla refused to comment about the removal of the 75D. This only makes sense if they have other, not yet announced plans. If dropping the 75D was their sole change then Tesla PR would try to reduce confusion and stop speculation, which speculation causes deferred demand. Instead, Tesla and Elon are 100% mum...

Now, in principle it's possible that they screwed up and just killed demand for the S/X. There's no way they can deliver 22k+ 100D's in Q1 IMO, at current prices.

So I do think the alternative has a higher than 90% probability: soon they'll announce new Model S/X plans and a new product lineup, with new pricing.

What is %posibilities of 90kw and 120kw introduction? similar to the axing of 60-85 to 75-100

 

[KarenRei]

What is %posibilities of 90kw and 120kw introduction? similar to the axing of 60-85 to 75-100

Everyone will give you different answers, and they of course depend on the timeframe.

I'll give you my numbers of a larger-pack Model S being introduced in the next three months: 7%: 4,5% odds of a more energy-dense 18650 chemistry, 2,5% of switching to 2170s. But I'd give 12% odds of a new drive unit in the next 3 months, and 35% odds in the next 6 months.

Also remember: Musk is trying to draw peoples' attention away from kWh capacity and simply toward range. Short range, mid range, long range, etc. Range comes not just from the battery pack size, but also vehicle and drivetrain efficiency.

 

[lklundin]

if they plan to do anything at all, they'll probably wait at least a week or two so that it doesn't seem like a giant middle finger (if it's a good thing they do) to the people who just bought 75Ds.

In connection with past battery capacity upgrades, the last models of the version to disappear were actually of the new and higher capacity, only limited to the old capacity via software that could be unlocked via OTA.

If that would also be the case here then that would soften the blow, to put it mildly.

 

[jhm]

Do you mean you think that only the 100D and P100D will be sold?

That would be awfully risky:

• According to @Troy's tracker, the 75D S/X is about 55% of sales. (!)
• Q1 is the seasonally weakest quarter, with the softest demand.
• Speculation has been swirling for months about an interior refresh, about 2,170 adoption. There's been a leak a couple of months ago about the new S/X interior. Everybody expects HW3 inclusion in the Model 3 in March/April.
• The P100D was introduced in the summer of 2016. The competition had 2.5+ years to catch up.
• Most tellingly, Tesla refused to comment about the removal of the 75D. This only makes sense if they have other, not yet announced plans. If dropping the 75D was their sole change then Tesla PR would try to reduce confusion and stop speculation, which speculation causes deferred demand. Instead, Tesla and Elon are 100% mum...

Now, in principle it's possible that they screwed up and just killed demand for the S/X. There's no way they can deliver 22k+ 100D's in Q1 IMO, at current prices.

So I do think the alternative has a higher than 90% probability: soon they'll announce new Model S/X plans and a new product lineup, with new pricing.

Yep, they're up to something.

 

[SageBrush]

Exactly. Intermittency is reduced by a larger area.

Also ..
I'm not quite sold on the notion that 'baseload' is a an invention of the nuclear and coal power industries but I accept that consumer behaviour was affected by those industry's power production characteristics. As intermittent power production becomes an even larger fraction of the power mix the consumer will (happily) adapt with appropriate price signals. Between EVs and hot water there is a *huge* time flexible energy sink that will not only dampen any generation/consumption imbalance but also (I suspect) reduce grid transmission bottle-necks and load costs.

I'm not trying to extrapolate my home to the country at large but it is a bit remarkable to think that due to home PV and EVs my inflexible grid load is around 0.5 kW and my time inflexible daily use about 0.5 kWh a day.

 

[Cherry Wine]

IMHO, the most likely change to the 100D and P100D would be a price reduction. Which would boost demand.

They've got very nice margins on them. 75Ds were dragging the gross margins down (but even then they were solid)

Let's do some math. Let's say that the 75D's margin is 12%, which is dragged up by the 100D and P100D. So that's $76k (for the Model S) * (1-0,12) = $67k production cost. Now let's say that their 18650s cost them $150/kWh at the cell level. Using 25kWh more per pack is just under $4k, so let's say $71k. But they currently charge $94k for the 100D, which would mean a 32% margin** on it. I bet they could trim the margin down to 15% or so on the 100D (with the P100Ds pushing the average margin up to 25%). So Model S would then start at ~$82k.

Basically they'd be cutting $12k off the price of Model S and X.

** ED: The calculated 32% margin on the 100D suggests that the 75D's margin is even less than my guestimated 12%.... they may have a sub-10% margin on them.

Unfortunately, this got me thinking about how the market reacted to the last price cut on Tesla vehicles.

 

[ReflexFunds]

IMHO, the most likely change to the 100D and P100D would be a price reduction. Which would boost demand.

They've got very nice margins on them. 75Ds were dragging the gross margins down (but even then they were solid)

Let's do some math. Let's say that the 75D's margin is 12%, which is dragged up by the 100D and P100D. So that's $76k (for the Model S) * (1-0,12) = $67k production cost. Now let's say that their 18650s cost them $150/kWh at the cell level. Using 25kWh more per pack is just under $4k, so let's say $71k. But they currently charge $94k for the 100D, which would mean a 32% margin** on it. I bet they could trim the margin down to 15% or so on the 100D (with the P100D pushing the average margin up to 25%). So Model S would then start at ~$82k.

Basically they'd be cutting $12k off the price of Model S and X.

** ED: The calculated 32% margin on the 100D suggests that the 75D's margin is even less than my guestimated 12%.... they may have a sub-10% margin on them.

Looking just at Model S for now, I think Tesla can afford to reduce S100D price from $97k to $87k and PS100D from $135k to $130k while maintaining the Q4 average gross profit per car (which is higher than the current level post Jan price cuts). I don't think this will be enough to sustain demand at c.50k annually though, $87k is still a large step up from the Q4 $77k S75D base price.

The S100D price can be reduced further (probably to around $83k) if they introduce a new S120D top of the range option.

All of these estimates are without any cost savings from a move to GF1 cells or motors, either of which could help bring the pricing down while keeping a constant average gross profit per car.


Model S 4Q18 estimate:

Model S with just S100D available:

Model S with new 120D option:

 

[lklundin]

My view is that it's much more likely that they'll import Model 3's drivetrain for one of the motors in order to extend range rather than to scrap a huge investment in 18650 production (possibly even freeing it up some of the world's best cells for the competition!) and spending a fortune to build another 10GWh/yr at Giga (plus all of the investments at Fremont).

* ~10% range boost
* ~10% charge rate boost (less power used per kWh)
* Lower home charging costs
* 10% fewer cycles on the battery = less warranty costs and/or they can be more aggressive with the cells
* Drive units are a much cheaper part than a battery pack. Even where capital is needed, you're talking much less
* Drive unit production is the smoothest, most automated production process Tesla runs at Giga, and possibly in the entire company. Very few humans involved. Readily scalable.
* Drive units are not structural elements (battery packs provide critical stiffening to the frame). You're not having a meaningful affect on your crash test data by changing drive units.

Plus lots of other things I'm forgetting.

I find the reasoning regarding the 18650 production convincing.

To reconcile that with my wishful thinking that Straubel et al have managed to fit the 2170-cell into the Model S/X pack, I like my own speculation that the 100 kWh pack will remain unmodified (at a reduced price) and that a new, higher-capacity 2170-based pack will be introduced soon.

The competition is talking about cars that will arrive in a couple of years, yet match Tesla Model S/X as they were a couple of years ago, so for Tesla to offer e.g. a 120 kWh(*) pack will really make the competition sweat.

Btw, you have likely forgotten more about Tesla than most people know...

(*) PS. If the 2170 fits, the Model S/X pack can likely hold more than 120 kWh, but then that can come as upgrades later on.

 

[scaesare]

If we just look at two engineering masterpieces of Tesla which is the heat pump or bottle that Munro showed...

I don't believe any Tesla models incorporate a heat pump.

 

[Zaxxon]

What could they even announce? Would Tesla be able to set up new battery pack assembly for Model S/X at GF1 using 21-70s without anyone hearing about it?

Doubtful.

No, there's not been some secret 10GWh extra capacity installed at Giga (in addition to everything else that would be required at Fremont). And don't hold your breath waiting for some grand announcement at all when 75D sales end. At the very least, if they plan to do anything at all, they'll probably wait at least a week or two so that it doesn't seem like a giant middle finger (if it's a good thing they do) to the people who just bought 75Ds.

I agree that we're unlikely to see something an hour after they nuke the 75s. That was me being facetious.

IMO the termination of the 75D is purely to rationalize their line up. Put simply the 75D was too close to the M3 -- Tesla would rather sell either a higher end S or X, or a higher end M3. This supports high production levels of M3 to keep its costs down while maintaining the S & X as premium vehicles. It makes no sense to keep it in production. But Tesla does want to sell out the units so some "get 'em while they're hot" rhetoric is only to be expected.

Part of me agrees--certainly they'd rather sell the high-margin, high-end vehicles. But the removal of the 75s without any other change would introduce a rather large gap into their lineup. Today, Tesla has the ability to sell you a vehicle at any price point you desire between about $40k (after tax credit) and $150k. If all they do is nuke the 75s, they introduce a gap from about $75k (maxed-out 3) to $95k (barebones S 100D). That's... not a great marketing strategy. I think we're bound to see something happen other than the removal of the 75s in the very near future. Either a reduction to the 100D prices, or some new mid-range battery (and possibly a better top-end).

(Somewhat related, I'm not convinced the MR has a long life expectancy for similar reasons -- it was released simply to get something lower priced as a stop gap for the SR not being ready yet. This is well demonstrated by its release being a surprise. Once the SR is being sold there's not much reason to keep the MR around.)

Agreed, although if it's selling well and the margins are good, I also wouldn't be surprised if they work out a way to keep it.

And, at least for now, I think Tesla needs a simple line up with clear choices in order to maximize sales.

I agree with this sentiment, although again I have to mention that leaving a $20k gap in your lineup means leaving money on the table. Unless there'a s large production efficiency / margin reason to do so, I don't think that $20k gap will be long-lived.

 

[MP3Mike]

1. grid cannot support EV
2. gasoline IC engines have 60% efficiency
3. EV energy cost is more per mile

(and that's just fallacious claims about EV, not other bizarre claims)

I'm not saying any of these claims should be taken seriously (I did some digging on the ICE efficiency, and the best I could find for automotive was 38.5% peak with a theoretical peak limit of 46% that varied with what you took for parameters, but a 60% theoretical limit for turbines) and am using them to demonstrate the grain of salt his claim about safety should be taken with.

You can make an ICE look much more efficient if you make the assumption that the passenger wants the cabin to be 100 degrees all the time. (I.e. you reduce the waste heat by using it.)

 

[M3Rider]

What is %posibilities of 90kw and 120kw introduction? similar to the axing of 60-85 to 75-100

I keep thinking in this direction too. Maybe 75->100 is too big of a step. Maybe the new low is 85 or 90. And the price for this would be pretty close to 75, which is why they wouldn't want to talk about it until they are done with 75 deliveries - or people will start demanding cancellations/discounts.
And then afterwards they could possibly add 110/120. I have my doubts though if there's really any opportunity for such increase in the battery capacity w/o some significant design improvements/upgrade of the 18650 cell and so far I'm not sold on the idea that they'll switch to 2170s in S/X. Maybe they keep prices a bit elevated for a while not to piss off people who just bought a smaller battery and decrease prices few months down the road?
If the "very long range" is indeed 20%+, then it's likely a 2170 switch and a redesign to account for the cell height difference.

 

[gringotuanis]

Anyone on here noticing how many model 3 vin's are being registered. It seems like since Jan 1 we have had something like 35 - 40k registered. Front end loading production for Europe and China? This seems like a hyper pace.

 

[ReflexFunds]

This is exactly right. Tesla's conclusions about safety have never been supported by the data. I've been disappointed that they've presented their results this way, as it's obviously wrong.

My belief from what I've seen is that auto-pilot is a good thing for safety in its current state, but not much. However it will only get better.

What else can they do? No other premium automaker discloses their safety data, or even has an ability to track it, so there is nothing they can compare it against other than the market average.

My analysis is for non Autopilot Tesla's, but every Tesla does still make use of the free Autopilot safety features.

I'm sure demographics play some role in Tesla's 4.4x fewer crashes per mile, but I highly doubt it is the main driver. I'm not sure how much I buy that rich people don't drink drive. A quick search says 12% of car accidents are caused by teen drivers who presumably can't afford a Tesla, but in many accidents there is another car and the driver could be any age.

There are also trends acting against them - I presume Tesla cars are disproportionately in city areas where a disproportionate % of accidents occur etc.

 

[CorneliusXX]

It's very close to 85% for the last 3 quarters if you correctly identify the production quarter boundaries for VIN allocations - which are NOT the same boundaries as the financial quarter boundary. 'Next quarter' VIN allocations IMO start sooner than the end of the previous quarter.

See this post where I used 85%.

Any thoughts on why they use 85%. Do you think there's any ratio to order cancellation or some sort of production metric?

If the idea was to just obscure production numbers then it seems too simplistic.

 

[DurandalAI]

An additional thought on killing off the 75kWh S&X and rebranding the 100 series to something that doesn't denote kWh capacity is that by raising the ASP of the S&X series, they will also help preserve the residual value of currently existing S&X vehicles. I believe that the 3 has had an impact on S&X residuals. Anecdotal example, but if I can choose between a 2015 Model S with AP1, a 240 mile range for $45-49k or a 2018 Model 3 with 260 mile range and AP HW2 for $49k, which one am I going to pick? With the entry price being higher on the S&X it will retain exclusivity. (Not to mention all the reasons others have mentioned already regarding margins on the 75 vs LR AWD 3, etc.)

 

[humbaba]

You can make an ICE look much more efficient if you make the assumption that the passenger wants the cabin to be 100 degrees all the time. (I.e. you reduce the waste heat by using it.)

Funny but true. Although from what I can tell you can't recover that much of the loss -- keep in mind the engine is not designed for this purpose resulting in some inefficiency -- and even so the heat differential isn't that much. There are some nice diagrams breaking out all of the losses, though sadly that was browsing from a different computer so I can't readily provide citations.

 

[Cherry Wine]

ER date?

Median Q4 earnings report date announcement is January 29th.

 

[AudubonB]

I know a traffic engineer who contended that Tesla's were far less safe than their peers. Because his job is traffic safety he had access to data for making that declaration. However, he made no effort to back the claim up and is an EV denier which makes his claim suspect without supporting evidence. As just a sampling of things he has told me:

1. grid cannot support EV
2. gasoline IC engines have 60% efficiency
3. EV energy cost is more per mile

(and that's just fallacious claims about EV, not other bizarre claims)

I'm not saying any of these claims should be taken seriously (I did some digging on the ICE efficiency, and the best I could find for automotive was 38.5% peak with a theoretical peak limit of 46% that varied with what you took for parameters, but a 60% theoretical limit for turbines) and am using them to demonstrate the grain of salt his claim about safety should be taken with.

Still, he does have access to data we do not and so should not be entirely discounted.

Gawd. Your acquaintance also, however, has access to data that we do have and, apparently, with those data he has come up with the clunkers you cited.

So I'm going to commit a Fallacy By Association and say that his other claim therefore carries no weight.