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

[AcesDealt]

Implied Tesla average COGs for a $35k base car:

• 1Q18 $67.6k ($53.2k ex depreciation).
• 2Q $49.0k ($42.3k ex depreciation).
• 3Q $39.5k ($36.8k ex depreciation).
• 4Q $38.5k ($36.0k ex depreciation) despite c.$1k higher costs from Trump's tariffs.

This is from my reconciliation of Tesla's quarterly accounts and using options mix surveys and other bottom up cost estimates and modelling.

Elon disclosed a base model 3 production cost of $38k in November. I presume this is on an accounting basis (which is how Elon normally talks about margin guidance), but it is possible Elon’s number includes deferred costs (in which case cost would be $36k on my basis) and also possible it doesn’t include warranty (in which case it would be $40.1k on my basis).

I think Tesla likely finished Q4 with the base cost at $37-37.5k.

Much of the easy work on cost reduction was done in 2018, but there is still room for significant cost reduction in 2019.

Roughly I think further base model 3 COGS savings could be achieved by:

• Production ramp from 5k/week to 7/k week, which could reduce depreciation by c.$0.5k per car and staff costs by c.$1.5-2k per car through operating leverage.
• New more efficient battery module and pack designs: $0.5-1.5k per car
• $10-20/kwh reduction in Panasonic cell purchase cost: $0.5-1k per car
• Cancellation of referral program (I'm not sure how this is accounted for, but it may reduce deferred revenue by c.$0.5k)
• Cancellation of Trump's China tariffs: $0.5-1k per car.
• Other supplier cost savings with negotiations and purchasing scale: $0.5-1k
• Other car design improvements: $0.5-1k
• Better production quality to reduce scrap, rework and warranty costs: $0.5-1k

Aside from improving gross margin, Tesla can alternatively try to reduce costs by cutting SG&A. This is the reason for cutting store sales. I roughly estimate this could cut annual SG&A costs by c.$750m, or c.$1.5k per car at annual production of 500k.

So Tesla needed to reduce COGs $7-7.5k to get base model 3 costs to the target $30k or a 14.3% gross margin on the base model (likely corresponding to 25% or so average across all options). If they are now instead reducing SG&A by $1.5k per car, the COGs cost target can increase $1.5k to $31.5k while maintaining the previous operating margin target.

That leaves $5.5k to $6k COGs reduction needed from the 2018 exit rate to get to the initial targeted operating margins. I've listed measures above that i think could potentially reduce COGs $5.5k to $9.5k. Of these, i think the ramp to 7k/week, new module lines and referral cut have already likely been implemented (together $3k to $4.5k COGs savings).

So if Tesla has ramped to 7k week, i think current COGs are likely $32.5k to $34.5k. This makes current gross margins on the base model 1% to 7% per car (I would guess closer to 1% than 7%). I think Tesla should get this base gross margin to c.10% in the US with the average Model 3 margins at 20-25% this year. In China next year base margin should be higher which should increase group overall model 3 margins.

These numbers are all rough but I'll finish a full model with all of the new options pricing and my estimated options mix in the next week or so.

I don't have any major problems with this analysis. However, we are of course hand-waving and taking WAGs at the cost savings on those various line items and I think your ranges tend towards highly optimistic. In particular I don't think $1.5k to $2k further savings on labor for production is very likely if you consider how many hours labor that represents and the amount of time they've already had to refine the process. Nor do I think it's likely that a new module design would come anywhere close to $1.5k which is a massive percentage of the entire price of the module.

At the end of the day of course it matters less where the exact cost of the base model is and more what the long-term demand is for cars in the upper price ranges since the vast vast majority of profit will come from ~45k$+ up configurations. Or one could alternately argue that the only thing that matters is the competitive advantages/disadvantages since pricing is mostly limited by competition. I appreciate the commentary at any rate.

 

[EVNow]

Never seen so many FUDsters over in the comment section of .

Tesla slashes retail employee compensation, seemingly pushing them out instead of layoffs

They have removed bonuses. Of course, if they are no longer salesmen, they wouldn't get sales bonuses.

 

[Dan Detweiler]

I think the drop in price of the S and X are explained much easier than many are making it. I think this is the final nail in the luxury sedan and SUV markets. Superior product and now cheaper than the competition. If I were BMW, Mercedes or Audi I would be very worried.

Dan

 

[KarenRei]

Interesting:

New Vehicle Sales Continue to Slip in February, Edmunds Forecasts | Edmunds

That's US sales only.

In December they estimated only 19,3k, which was a low estimate:
New Auto Sales on Track for Near-Record Year in 2018, Edmunds Forecasts | Edmunds

So now it's Edmunds and AlphaHat strongly disagreeing with InsideEVs. InsideEVs may have the same factor that messed with Troy's Q4 delivery estimates (the drop in participation in VIN reporting).

ED: Might as well include January:
New Vehicle Sales Expected to Lift in January, Edmunds Forecasts | Edmunds

 

[Intl Professor]

What is the technique for getting out?

I don't remember. What are you referring too?

 

[EVNow]

No, the life experience of struggling with $1 a day never goes away, it's formative and has an almost universally binary outcome: it either turns people into lifelong sugar-holes ('if I made it you can do it too and you don't need my help'), or results in lifelong compassion and philanthropy ('that guy could have been me with a bit less luck').

This is absolutely true. I see this everyday - since most of the immigrants from India I know fall in this category of barely above poverty when we were growing up. It not only defines the attitude but also the politics.

 

[anthonyj]

Interesting:

New Vehicle Sales Continue to Slip in February, Edmunds Forecasts | Edmunds

View attachment 382425

That's US sales only.

In December they estimated only 19,3k, which was a low estimate:
New Auto Sales on Track for Near-Record Year in 2018, Edmunds Forecasts | Edmunds

View attachment 382424

So now it's Edmunds and AlphaHat strongly disagreeing with InsideEVs....

ED: Might as well include January:
New Vehicle Sales Expected to Lift in January, Edmunds Forecasts | Edmunds

View attachment 382426

Geez. That would clear up a lot of my doubt if Edmunds is right.

Edit: why don’t they include 2018?

 

[Hock1]

Yes. @Hock1 follows this closely, he may wish to comment as well.

I think @lklundin also posted some work in the Fall with data he extracted regarding deliveries. I think his data source is updated mid-month, 2 months in arrears IIRC. Correlating those reports with total volume should produce an interesting statistic. <hint, hint>

BTW, Isn't it the mythical forensic Shorts that are supposed to keep the market honest?

Cheers!

No, there is no way to absolutely prove that the market for TSLA or any other stock, manipulated or not, is filled with undelivered, naked short-sold stock. All of the data feeds on short-selling, undelivered stock, etc., by design, obfuscates any attempt at understanding a snap-shot of the extent to which the concept of "zero-sum game" has been compromised at any point in time. "Short interest" is reported to the SEC twice per month and revealed to the public 10 days after the reporting period; fails-to-deliver reporting follows a similar pattern but is even harder to ferret out any meaningful patterns. I.e., an alphabetical list of ticker symbols is produced on daily FTDs, weeks after the fact. There is no correlation from one day to the next of the FTDs in any particular stock. There is no "running " total of FTDs reported. If the authorities (SEC, FINRA, DTCC) wanted to present this information in an organized, up-to-date fashion they certainly could with the capabilities of today's computers. Why can't one just go to the SI site or the FTD site, plug in TSLA, and get cumulative data through yesterday's trading??? Why?

 

[EVNow]

So now it's Edmunds and AlphaHat strongly disagreeing with InsideEVs....

Only Insideevs takes the trouble of accounting for seasonality. Others tend to take the quarterly figure and just divide by 3.

 

[Artful Dodger]

So, I’ve lightly followed the conversation on this topic. Enough to know that some posters have said naked shorting is illegal (to which I snorted and said to myself - ‘like that’s ever stopped anyone before.’) and insisted it’s not happening/can’t happen nowadays.

Naked short selling is illegal for normal traders. There is an exemption for market makers, based on the fallacy that options makers need the ability to sell short naked in order to "mantain liquidity". This is of course the very way in which they prevent a short squeeze from occurring (see 13-day settlement rule below). This is how market makers are able to cap the share price, while also working hard to scare retail investors and fund managers out of the market:

Morgan Stanley: Tesla Inc (TSLA) Has Become Untradeable as Volatility Explodes
| StockNews.com | Jul 25, 2018

The SEC enacted Regulation SHO in January 2005 to target abusive naked short selling by reducing failure to deliver securities, and by limiting the time in which a broker can permit failures to deliver. The rule included the following exemption:

1. Broker or dealer accepting a short sale order from another registered broker or dealer

Regulation SHO requires broker-dealers to close out open fail-to-deliver positions in threshold securities that have persisted for 13 consecutive settlement days. This is the amount of time brokers have to cover their massive naked short sales. Have you noticed how TLSA stock price occillates in a cycle of high->low about every 10-12 days recently?

By exemption, two or more market makers can collude to conduct unrestricted naked short selling by volume, and constrained only by 13 days to cover. I believe this is what is happening to TSLA, and I'd like the trading patterns investigated of market makers GoldmanSachs and UBS to begin.

Where is the SEC?

P.S. Look at this 2018 case study of naked short selling by market makers. It's a practice that's deeply hidden both by the rules of disclosure and in the data release/delay.

 

[KarenRei]

Only Insideevs takes the trouble of accounting for seasonality. Others tend to take the quarterly figure and just divide by 3.

How do you take the quarterly figure of a quarter that is in progress?

Are you saying they take Q4 figures for Q1? Then how are they talking about monthly trends in the articles? Why are the January and February numbers different?

Remember that AlphaHat is also disagreeing with InsideEVs:

Tesla’s US Q1 Deliveries are Surprisingly High (so far)
Tesla Delivered More Vehicles in January than Reports Suggest

(Although AlphaHat's numbers are between those of Edmunds and InsideEVs)

Someone is going to have serious egg on their face when the final numbers come out, and I'm not sure who it's going to be.

 

[InDaClub]

Anyone coin “pull a Fred” yet?

Yes, however, around here we prefer to call him Dick.

 

[EVNow]

I'm driving right now the development version of Autopilot and it works extremely well in terms of recognizing traffic lights and stop signs and is now starting to make turns effectively in complex urban environments.

This is key. Google's Waymo struggles with this in their geofenced urban area driving even now - though this feature probably doesn't have much to do with geofencing.

Waymo One Experience (Part 2) : SelfDrivingCars

At that point, the vehicle was attempting to make a left turn out of the neighborhood onto a 5 lane with median surface street. Ultimately, the safety driver did place it into manual mode in order to make the left turn and then placed it back into autonomous.

 

[anthonyj]

Johnny Depp Claims Amber Heard Cheated on Him With Elon Musk

Very important

 

[AudubonB]

Yes, Elon doesn't use it "correctly", but there's simply no good word for what he is trying to communicate: the binary filter locking out way too many people who'd like to own a Tesla.

There's basically no other major consumer products firm that is locked out of 80% of people's lives just due to affordability.

Almost every luxury products firm can sell to ~80% of consumers, either via lower quantities, or by tiered features:

• You can buy an entry level B&W speaker, you don't have to buy the $20,000 Nautilus to own a B&W set.
• You can buy a MacBook Air for $999, you don't have to buy a fully configured $10,000 iMac Pro.
• You can buy a Mercedes or BMW in the $30k-35k MRSP range.
• Except Tesla: a new Tesla used to start at $80k, locking out ~90% of consumers.

With the $35k Model 3 this has changed - it should be within the financial reach of about ~60% of new car consumers.

I.e. theres a big difference between people choosing other products because they prefer them to Tesla's products (regular definition of demand), versus not choosing Tesla only because they truly cannot afford any of Tesla's products.

Well, he could use "desire" and, in fact, I would strongly argue that is the perfect word in English. "Desire" would, unfortunately, come across as more than a little weird and therefore, even wearing my economist hat, I advise him to keep on using the word "demand".

 

[Hock1]

Naked short selling is illegal for normal traders. There is an exemption for market makers, based on the fallacy that options makers need the ability to sell short naked in order to "mantain liquidity". This is of course the very way in which they prevent a short squeeze from occurring, and how market makers are capping the stock, while working hard to scare retail investors and fund managers out of the market:

Morgan Stanley: Tesla Inc (TSLA) Has Become Untradeable as Volatility Explodes | StockNews.com

Look at this 2018 case study of naked short selling by market makers. It's a practice that's deeply hidden both by the rules of disclosure and in the data release/delay.

The SEC enacted Regulation SHO in January 2005 to target abusive naked short selling by reducing failure to deliver securities, and by limiting the time in which a broker can permit failures to deliver. The rule included the following exemption:

1. Broker or dealer accepting a short sale order from another registered broker or dealer

Regulation SHO requires broker-dealers to close out open fail-to-deliver positions in threshold securities that have persisted for 13 consecutive settlement days. This is how long brokers have to cover their massive naked short sales. Have you noticed how TLSA stock price occillates in a cycle of high->low about every 10-12 days recently?

By exemption, two or more market makers can collude to conduct unrestricted naked short selling by volume, and constrained only by 13 days to cover. I believe this is what is happening to TSLA, and I'd like the trading patterns investigated of market makers GoldmanSachs and UBS to begin.

Where is the SEC?

The SEC, in their infinite wisdom and their proclivity to avoid work, depend upon FINRA and DTCC to report any infractions of trading rules to the SEC. Only problem is: GS, UBS and others literally own and operate FINRA and the DTCC. Unfortunately, the only way I can see for this farce to be revealed is not a picture I'd like to describe. Also, given its experience to date with TSLA, it is hard to imagine the SEC foregoing their pettiness in order to actually enforce the law.

 

[AudubonB]

Regarding information, the following has come up already, but I felt it's worth pointing out in this context:

View attachment 382331
[2017 Data]
Source [December 12th, 2018]: "EV Sales Capital Of The World" Is Shanghai | CleanTechnica

I always wondered how Tesla was going to counter VAG's enormous dealer network and their millions of customer leads. However, not sure the virtual approach works so well for luxury vehicles. As others have pointed out, Tesla and Elon Musk adapt.

As an aside, Changsha had 7.432 million inhabitants in 2015. VAG has become the world's #1 automotive manufacturer due to their success in China - where they are being forced to step up to the EV plate.

Wow: Oslo and Bergen certainly are punching way above their weight. Does anyone see any other real outliers?

 

[AudubonB]

Given age, girth, overall stiffness, I had to get in and out of an M3 before final purchase. I had already driven a Model S where the same issue existed. Now I'm "relearning" how to twist after side-saddle sit. Getting the last leg in is still a problem but a hand helps.

If I can do it almost anyone can. When TSLA hits $4,000 if I last that long (and still driving) will probably upgrade to an MX if refreshed.

In that case, I hope you get a good night's sleep before that so that you indeed are refreshed.

 

[Carl Raymond]

I mostly agree.

But those employed at Tesla giving Test drives would mostly disagree.

There is a price to the chaos.

Including the feeling that maybe Tesla doesn't offer careers but only temporary employment. Is this 3 or 4 rounds of job cuts at Tesla in the last 12 months?

True, but they also keep employing. One hopes they always offer redeployment where possible, and should an opening arise later I would expect that a prior history at Tesla would rank pretty highly.

 

[Artful Dodger]

Stores would not be required at all.

Neither would Delivery Centers be require. Or a fleet of delivery trucks and trailers. Just leave the back gate open at the Shanghai factory, and let those 3SRs go straight to work.

CHA-CHING!