TSLA Market Action: 2018 Investor Roundtable

[neroden]

Any strategy that’s public and works will be copied by others, including high-frequency algos, until all the alpha is arbitraged away. So if it’s the kind of thing that amateur investors know about, then it probably doesn’t work.

While I agree that most technical trading is nonsense, the fact is that arbitrage opportunities are left on the table all the time. Advantages simply are not arbitraged away. The most famous example I can think of is RDS.A and RDS.B stock. You could make money consistently just by shorting one and buying the other, and reversing every so often. Few enough people do it that *you could still make money doing that right now*.

So you are simply dead wrong about this point. You are repeating one of the versions of the efficient market hypothesis, and specifically, one of the false and disproven versions.

 

[neroden]

What metrics should be used, then? Free cash flow? EBITDA?

Proper "business decision" or "management purposes" accounting.

Look, you've probably never read Benjamin Graham's first edition of _Security Analysis_. I strongly recommend it... but don't bother to read the whole thing. It consists of a long series of chapters on how to take the financial statements of the 1930s and recalculate the contents to get a real economic model of the company, eliminating the distortions caused by bad GAAP standards. That's the *entire book*. Over 700 pages of it.

GAAP of the 1930s had different rules and different distortions than GAAP of the 2010s does, which is why learning most of the specific techniques for re-analyzing the published numbers isn't really relevant any more. But the general principle of investment analysis, to re-analyze the numbers to get a real business analysis, remains the same.

One of the ways you can beat the market is if you find a company where GAAP differs radically from accurate business-decision accounting, and you do the analysis, and everyone else is blindly assuming that GAAP reflects reality. Tesla's actually a good example of this, but there are plenty of others.

 

[steve3]

(A) From reading the Model 3 delivery forums -- and I know this is anecdotal -- it looked like more than 2/3 of Performance buyers were getting "P-" (no "upgrade package") rather than "P+"
(B) Given that people seem to overvalue free supercharging, I expect that less than half of the P+ buyers will actually ask for the refund
(C) I thnk that no more than 15000 people have bought *Performance* since Model 3 came out; that would be 15% of buyers which seems high already.

So I'm guessing the exposure is more like 1.25 million dollars. Still.

I get $12.5million...?
15000 * 1/2 * 1/3 * $5000
Might not be right though, watching world series.

 

[oneday]

Neroden is closer on the NN subject than the others here.

I think the NN is going to start getting really good, really fast especially once the Tesla chip is in use. The edge cases of FSD on surface streets and rural environments is what could hold it back in those situations for many years.

 

[neroden]

1. 
   Verified account 11h11 hours ago
   Didn’t you get free Supercharging for life? Current buyers don’t get that.
   
   69 replies21 retweets788 likes
   
2. 
   
   Verified account 11h11 hours ago
   That’s not worth $5K. I opted for my Performance without PUP but I feel bad for anyone who paid. That sucks.

Not worth $5k? Quick math about the value of free SupperCharger here in Germany.

Using actual gas payment receipt in front of me.

This is not the right calculation. I'm not defending Fred, but the correct comparison is between free Supercharging and the cost of paid Supercharging. Most people will charge their Teslas at home (and save lots of money on gas by doing so, yes).

If you take very few road trips, the free Supercharging is worth next to nothing. If you take a lot of road trips, then it's worth a lot. I take maybe 1 Supercharger trip per year; over 5 years, free supercharging has probably been worth about $700 to me.

 

[TradingInvest]

I wonder if we should prepare Elon a list of Twitter handles he can trust.

Elon should consider creating a website similar to Twitter, but figure out a way to keep it a nice place.

We work so hard to make our air clean, lake clean, room clean... why it's is the norm and acceptable to live in a dirty social environment which is so damaging to our mind? If there are lots of smokers in the room and we can't kick them out of the room, the right action is to find/build a cleaner room.

 

[ThereAre4Cars]

I think that the NN is going to start getting really good, really fast especially once the Tesla chip is in use.

You got to be careful who you listen to. This is wrong.

The edge cases of FSD on surface streets and rural environments is what could hold it back in those situations for many years.

And this is right.

I've been following Karpathy on Twitter and he is overly impressed with topics of optimization and it makes me nervous actually. The real bottleneck on FSD is if Tesla hires and relies on someone that doesn't actually know what he's doing because he might stonewall Elon even. As impressed as people are with Elon, he generally just repeats tropes from various tech subcultures that are 10+ years before him. Very little he says on Rogan etc is novel. The only difference between Elon and these people is Elon actually gets *sugar* done. And that's not a small difference.

Anyway that might be because that's all Karpathy happens to be allowed to post about.. I don't know. I think he posted another optimization related tweet today. This is almost never the bottleneck.

You have as a general stack:

Hardware
Perception
[localization]
Planning
Business Specification (this is where Neroden shines and it's the most important)
User Training

roughly, the hardware's impact stops early in the stack. The hard parts are planning and above.

 

[Singuy]

(A) From reading the Model 3 delivery forums -- and I know this is anecdotal -- it looked like more than 2/3 of Performance buyers were getting "P-" (no "upgrade package") rather than "P+"
(B) Given that people seem to overvalue free supercharging, I expect that less than half of the P+ buyers will actually ask for the refund
(C) I thnk that no more than 15000 people have bought *Performance* since Model 3 came out; that would be 15% of buyers which seems high already.

So I'm guessing the exposure is more like 1.25 million dollars. Still.

Except now that P- people are also crying for a refund or a "free upgrade" to P+ since P+ people will have their cars costing the same as a P- with the upgrade installed after refunds.

Elon should have to said "we are trying to make our cars as affordable as possible so price will be fluctuating constantly. People who got the P+ will not only have supercharging for life, high speed LTE for life, but also guaranteed to have receive their federal tax credit. I thank all the early adopters for supporting my company and our mission". Done and done.

 

[ThereAre4Cars]

Is there a button for, "i only want to read just this one persons posts for awhile, or maybe 2-3 folks, PLEASE!!!! DEAR SWEET MOTHER OF GHOD

Yeah the funny part is most of that was wrong.

 

[neroden]

Except now that P- people are also crying for a refund or a "free upgrade" to P+ since P+ people will have their cars costing the same as a P- with the upgrade installed after refunds.

...that's on Tesla for immediately releasing the P+ at the former P- price. I told them back in 2014 not to do that. They should have discontinued the P- *at the P+ price*, then waited a few weeks, *then* cut the price in stages.

Yes, I am saying that Tesla has been cutting prices too fast and should price-gouge more.

It will make more people happy (in this sort of case), perverse though this sounds -- *and* it will make Tesla more money. They should do it.

 

[graphix25]

Debt Picture from 10-Q in Q2.

I really would like to see 4 profitable quarters before they plow ahead with big investments for the Model Y, Semi, Pickup and the $25k ICE destroyer. It's pretty amazing that ICE competitors let them pretty much have the market to themselves since 2012. I don't see any meaningful competitive response till 2020 or 2021. Typically in business at best one gets a two or three year lead with a new major product offering. The only other new product offering I can think of that had this kind of lead was Amazon AWS. It was 7-8 years before Microsoft, Google and others had a competitive response to AWS. At that time AWS was so far ahead it's very difficult to catch up even with the deep pockets and resources of MSFT and GOOG. Tesla looks to be following the same path to take 50-60% of the EV market over the long hall.

 

[neroden]

I’m invested in Tesla primarily because of self-driving (even though I can’t say with certainty it will arrive; no one knows the future) so my view is kind of the opposite.

Oh boy. OK, this explains your disconnect from reality. Have you actually driven an electric car yet?

Maybe Tesla will have outsized market share of the global car market. Okay, but why? What’s to stop incumbents like GM, Volkswagen, Toyota, et al. and upstarts like Lucid, Faraday, BYD, NIO, Dyson, and possibly Apple from **eventually** catching up on batteries and electric drivetrains?

Brief recap of an investment thesis which has been explained here many times.

* The incumbents don't want to -- Innovator's Dilemma. They'll run out of cash before they want to.
* The upstarts mostly aren't capitalized -- they'll run out of cash before they can get started.
* Any company *starting now* will take a minimum of *five to ten years* to catch up with where Tesla is now.
* The companies which have started, BYD and BAIC, have demonstrably inferior technology and demonstrably higher structural costs, so even *they* are behind.

Obviously Tesla won't be the market leader forever, just like GM wasn't the market leader forever, but in the long run, we are all dead, as Keynes said. If Tesla has a 20-year run that's plenty.

Repeated bailouts by governments are not really a business model and leave the bailed-out companies in a weak position with damaged brands and the same mismanagement they had before.

No stock is literally a forever stock. ExxonMobil was a great investment from the 1950s through the early 1990s, and then became a trash speculation. GE was a great company for over 100 years and is trash now.

Apple could lose its entire smartphone business practically overnight -- it's *easily* disruptable, no moats except the brand, and the brand already has a lot of negative associations! Amazon could lose market leadership in eCommerce if Alibaba improves its site, and could lose leadership in the "cloud" to any of a dozen competitors very quickly if they manage to beat AWS. Amazon and Apple will probably remain the market leaders for years, though.

 

[neroden]

I guess we will see. How long after these purchases does Tesla have to make the SEC filing? (A filing is required right?)

Form 4 must be filed within 2 business days. So if Musk buys stock on Monday it must be filed by Wednesday at the latest (earlier is OK).

 

[neroden]

Yeah, I agree with the idea that faster-growing companies in the same sector should have higher multiples than slower-growing ones. In a few quarters, we can look at Tesla’s PEG ratio. Then the question is just what the growth ceiling is. And once it hits a growth ceiling, that’s when it should be valued at similar multiples to other companies in the same sector.

*Sigh* So Tesla's growth ceiling is probably the growth rate of the world electric car business, i.e. 50% per year. And if you want to value by PEG (which is mathematically unsound, but whatever), all non-Chinese stock-market-listed car companies are clearly *shrinking*. You can give it a shot, but a DCF analysis makes more sense.

 

[Zhelko Dimic]

Oh boy. OK, this explains your disconnect from reality. Have you actually driven an electric car yet?


Brief recap of an investment thesis which has been explained here many times.

* The incumbents don't want to -- Innovator's Dilemma. They'll run out of cash before they want to.
* The upstarts mostly aren't capitalized -- they'll run out of cash before they can get started.
* Any company *starting now* will take a minimum of *five to ten years* to catch up with where Tesla is now.
* The companies which have started, BYD and BAIC, have demonstrably inferior technology and demonstrably higher structural costs, so even *they* are behind.

Obviously Tesla won't be the market leader forever, just like GM wasn't the market leader forever, but in the long run, we are all dead, as Keynes said. If Tesla has a 20-year run that's plenty.

Repeated bailouts by governments are not really a business model and leave the bailed-out companies in a weak position with damaged brands and the same mismanagement they had before.

No stock is literally a forever stock. ExxonMobil was a great investment from the 1950s through the early 1990s, and then became a trash speculation. GE was a great company for over 100 years and is trash now.

Apple could lose its entire smartphone business practically overnight -- it's *easily* disruptable, no moats except the brand, and the brand already has a lot of negative associations! Amazon could lose market leadership in eCommerce if Alibaba improves its site, and could lose leadership in the "cloud" to any of a dozen competitors very quickly if they manage to beat AWS. Amazon and Apple will probably remain the market leaders for years, though.

Agreed with everything, except Amazon cloud.
I am on my second company leading AWS migration, and no way in hell would I ever consider moving somewhere else, once settled. It's a very, very sticky service. Stickier than the database(Oracle) or OS(Windows) choice.
Many reasons: mindshare in developers community, availability of talent and their desire to work in AWS, it's a big corporate decisions, i.e. slow and exposed to many decision makers, cost of switch is horrendous, no two clouds are the same, so you're redeveloping/redeploying... Amazon has a momentum too. AMZN is the only other company I'm invested in, and I actually feel more comfortable about AMZN than TSLA. Once TSLA crosses $400, I'll probably reposition 25% of TSLA into AMZN. I am certain AWS becomes 10x of what it is now, maybe 100x. Having said that good chunk is already priced in...
Oh, and cyber-criminal is scary. I'd rather have Amazon continue to spend billions securing their stuff than being directly responsible. You will soon start to hear "No one gets fired for choosing AWS'

 

[ThereAre4Cars]

View attachment 347562

I have to say I honestly didn't get this meme. Is it just 'gotcha!'

Whitfield is Brantley! Brantley is Whitfield!

 

[ThereAre4Cars]

While I agree that most technical trading is nonsense, the fact is that arbitrage opportunities are left on the table all the time. Advantages simply are not arbitraged away. The most famous example I can think of is RDS.A and RDS.B stock. You could make money consistently just by shorting one and buying the other, and reversing every so often. Few enough people do it that *you could still make money doing that right now*.

So you are simply dead wrong about this point. You are repeating one of the versions of the efficient market hypothesis, and specifically, one of the false and disproven versions.

This leaves a bad taste in my mouth. Plus I don't know why you are going after this guy so hard. I don't actually know this specific case but I suspect it doesn't actually generalize as well as you hope it does. Sorry. You definitely had me nodding along with 'While I agree most technical trading is nonsense' because I swear to god this is as close to true as it needs to be to shut people up about it.

 

[neroden]

LG will take profits.

Further to that: who's making batteries?
-- Panasonic, practically captive to Tesla
-- Samsung, low volume and also a Tesla supplier
-- LG, who is squeezing its customers for profits
-- CATL, who is struggling for financing
-- the captive battery manufacturer of BAIC, which is locked in a bad chemistry by Chinese regulation
-- the captive battery manufacturer of BYD, which is locked in a bad chemistry by Chinese regulation
-- VW has grandiose plans

This isn't a long list.

 

[neroden]

(A) From reading the Model 3 delivery forums -- and I know this is anecdotal -- it looked like more than 2/3 of Performance buyers were getting "P-" (no "upgrade package") rather than "P+"
(B) Given that people seem to overvalue free supercharging, I expect that less than half of the P+ buyers will actually ask for the refund
(C) I thnk that no more than 15000 people have bought *Performance* since Model 3 came out; that would be 15% of buyers which seems high already.

So I'm guessing the exposure is more like 1.25 million dollars. Still.

*cough* and I can't do arithmetic today. 12.5 million.

 

[neroden]

This leaves a bad taste in my mouth. Plus I don't know why you are going after this guy so hard. I don't actually know this specific case but I suspect it doesn't actually generalize as well as you hope it does.

It generalizes perfectly. This isn't a matter of hope, this is stone-cold fact. You can find academic papers on the topic; I've read several.

There's a chapter in the best investment book ever, _Efficiently Inefficient_, discusing arbitrages. They're one of the reliable ways to beat the market. The interesting question is why there are always so few arbitrageurs that the arbitrages persist. (Personally, it turns out I don't like doing arbitrages; they stress me out. Dunno why.)

Arbitrages persist for literally decades. Please look up this particular example; it's actually famous and there are papers specifically about it. RDS.A and RDS.B stock are *exactly the same* except that one of them produces Dutch-source dividends and the other produces British-source dividends. For third-country investors, the British-source dividends are better because they aren't subject to the Dutch dividend withholding tax. They're also just as good for Dutch investors. Despite this, half the time the Dutch-source-dividends stock trades higher.

I respond to people who believe common lies about the stock market. Facts matter. It's not about the person. The guy I'm responding to has been demonstrably wrong, a *lot*. Maybe he'll learn something.