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

[ZeApelido]

Adam Jonas is a chump.

...

Long live Adam Jonas!

 

[Jovian]

Just out, Morgan Stanley’s Adam Jonas is now expecting a recovery in demand for the remainder of the year and into 2020 with a 29% potential upside for the stock price.

“While demand has disappointed vs. our expectations this year, we believe the weakness is largely temporary. In October 2018, we had expected 412k total Tesla units delivered for 2019. By February we had revised our 2019 delivery forecast to 380k units. Today, our 2019 global Tesla unit delivery forecast stands at 347k units. Our current year unit volume has been revised down 16% since October. Over the same period, our 2019 group revenue forecast has been revised down 19%. Looking to 2020, while our forward unit volume forecast has been revised down 11% since October, we still forecast a 39% growth in volume to 482k units. Additionally, we forecast 29% revenue growth in 2020, driven by global availability of the Model 3 in markets like China, leasing and a refreshed Model S and X. We forecast further revenue growth in 2021 driven by the introduction of the Model Y. For the remainder of 2019, we expect a significant sequential improvement in demand driven by: continued European ramp, greater availability of leasing, Model S/X refresh, introduction of lower priced models and a number of price reductions. These drivers confront what we expect to be a continued phase-out of regulatory tax credits, rising availability of second-hand cars and less competitive imports (i.e. 40% import tariff) to China before the start of local production. Heading into 2020, we expect the key demand drivers to be heavily dependent upon start of local Chinese production and the Model Y intro.”

Is this a joke?

 

[RobStark]

On the other hand, Ford has a heavy, HEAVY interest in keeping their F-number series going strong. Why would they put the best effort going forward to steal their own truck sales? I just don't see them doing that, so Rivian may be limited.

Mercedes had a heavy interest in keeping their S-Class and E-Class sales going strong. Why would Mercedes invest in and save Tesla from bankruptcy?

BTW Ford doesn't control Rivian. Ford isn't even their largest investor, Amazon is. Ford has 1 of 7 Board seats at Rivian. Their percentage ownership was not disclosed.

 

[AcesDealt]

I would argue in the case of driving, humans too are actually poor at figuring out how more than 4 or 5 objects (if that) would interact and take timely action to avoid crashes.

Even when you consider the infamous scenarios like the traffic in India, i’m Still only considering a couple of vehicles and a few pedestrians at a time. Or you group them and judge when a set of vehicles are turning or pedestrians as a group are crossing.

Imagine some ad hoc scenario where there are 12 random pedestrians, half look old and half look drunk, some with dogs on leashes and some with children, and 50 cars in various lanes and directions, and a complicated physical intersection.. the human brain is capable of making impressive predictions and navigating through it. Our best tech acts like hodor 'go straight. follow line. No hit things.'

When ai approximates our ability to drive in novel situations we are probably within 10 years of singularity.

Those that generalize the FSD chip to this kind of ability have no idea what they are talking about. It has no role to play at all.

 

[stef]

Just out, Morgan Stanley’s Adam Jonas is now expecting a recovery in demand for the remainder of the year and into 2020 with a 29% potential upside for the stock price.

“While demand has disappointed vs. our expectations this year, we believe the weakness is largely temporary. In October 2018, we had expected 412k total Tesla units delivered for 2019. By February we had revised our 2019 delivery forecast to 380k units. Today, our 2019 global Tesla unit delivery forecast stands at 347k units. Our current year unit volume has been revised down 16% since October. Over the same period, our 2019 group revenue forecast has been revised down 19%. Looking to 2020, while our forward unit volume forecast has been revised down 11% since October, we still forecast a 39% growth in volume to 482k units. Additionally, we forecast 29% revenue growth in 2020, driven by global availability of the Model 3 in markets like China, leasing and a refreshed Model S and X. We forecast further revenue growth in 2021 driven by the introduction of the Model Y. For the remainder of 2019, we expect a significant sequential improvement in demand driven by: continued European ramp, greater availability of leasing, Model S/X refresh, introduction of lower priced models and a number of price reductions. These drivers confront what we expect to be a continued phase-out of regulatory tax credits, rising availability of second-hand cars and less competitive imports (i.e. 40% import tariff) to China before the start of local production. Heading into 2020, we expect the key demand drivers to be heavily dependent upon start of local Chinese production and the Model Y intro.”

New PT $12.9 ?

 

[RobStark]

I feel Rivian's situation is similar to NIO. Everything seems great, except one problem: how will they deal with competition from Tesla? Tesla spent 16 years to accumulate experience and knowledge in battery tech, drivetrain, chip, software, production scale, automation... If Tesla offers a great pickup truck at $50k, will Rivian sell well at $70k?

NIO did everything including lying to show they have been selling well with bright future, they can't change the fact Tesla will compete with them head on. The market can see that.

NIO doesn't own an automotive factory. Rivian does.

You are falling into the fallacy that BEVs are a tiny market where all BEV makers are fighting over. North Americans buy over 3M trucks per year. Tesla won't be able to make 3M trucks per year any time soon.

NIO is a robotaxi play. Rivan is a BEV truck play.

Secondly, NIO is design and marketing. Rivian is an engineering firm.

New companies are selling BEVs in China just fine too.

Tesla is planning to offer a scifi bladerunner truck by Elon's admission most won't like.

Rivian is planning a conventional luxury truck with screen instrument cluster and luxurious leather seats. With a unique headlight design, that I almost guarantee most will acclimate to in 1.5 years of seeing them in Trucking and Overland magazines.

Rivian can hire experienced engineers too. This was the same argument used against Tesla, that it would not be able to build gliders. Because it has no experience.

I am pretty sure Rivian won't attempt to build Alien Dreadnought 1.0

 

[Fact Checking]

Is this a joke?

It's part of the regular Wall Street dump & pump scheme that is illegal for everyone except Wall Street analysts, who are exempt from regulations that make it illegal to trade on material information. @neroden called it a 'bear raid'.

Leading up to the Q2 delivery report I expect many of the analysts who (after Q1 consciously and deliberately tried to squeeze leveraged Tesla investors and tried to rattle long term holding investors) to find some excuse before Tesla's Q2 results potentially embarrassingly falsify the nonsense they've been spouting for months.

 

[aagosh]

Is this a joke?

I wonder if he still has any credibility left or at least a bit of self respect. What kind of Jokers MS hires who change their tunes with Atlantic winds? Can't believe people pay to hear this sh**

 

[Singuy]

It's part of the regular Wall Street dump & pump scheme that is illegal for everyone except Wall Street analysts, who are exempt from regulations that make it illegal to trade on material information. @neroden called it a 'bear raid'.

Leading up to the Q2 delivery report I expect many of the analysts who (after Q1 consciously and deliberately tried to squeeze leveraged Tesla investors and tried to rattle long term holding investors) to find some excuse before Tesla's Q2 results potentially embarrassingly falsify the nonsense they've been spouting for months.

Sooooooo the knife stop falling? Great...party is over guys. Hope all them dry powder was put to use.

 

[Tslynk67]

How do you know ?

Anecdotally, people I've talked to use EAP all the time and have figured out at what places they should disengage etc.

In any case, if you're holding the wheel - AS INSTRUCTED! - then if the car does veer unexpectedly, then it disengages quickly, you correct and then re-engage AP. It's not an issue at all and these cases will be resolved as the software iterates - exactly as we've seen improvements with every software update for other issues, lane-bouncing being the most obvious - it just doesn't do this any more.

So people complaining about these edge cases now, just don't understand the iterative process that's going on.

 

[aagosh]

Just out, Morgan Stanley’s Adam Jonas is now expecting a recovery in demand for the remainder of the year and into 2020 with a 29% potential upside for the stock price.

“While demand has disappointed vs. our expectations this year, we believe the weakness is largely temporary. In October 2018, we had expected 412k total Tesla units delivered for 2019. By February we had revised our 2019 delivery forecast to 380k units. Today, our 2019 global Tesla unit delivery forecast stands at 347k units. Our current year unit volume has been revised down 16% since October. Over the same period, our 2019 group revenue forecast has been revised down 19%. Looking to 2020, while our forward unit volume forecast has been revised down 11% since October, we still forecast a 39% growth in volume to 482k units. Additionally, we forecast 29% revenue growth in 2020, driven by global availability of the Model 3 in markets like China, leasing and a refreshed Model S and X. We forecast further revenue growth in 2021 driven by the introduction of the Model Y. For the remainder of 2019, we expect a significant sequential improvement in demand driven by: continued European ramp, greater availability of leasing, Model S/X refresh, introduction of lower priced models and a number of price reductions. These drivers confront what we expect to be a continued phase-out of regulatory tax credits, rising availability of second-hand cars and less competitive imports (i.e. 40% import tariff) to China before the start of local production. Heading into 2020, we expect the key demand drivers to be heavily dependent upon start of local Chinese production and the Model Y intro.”

Source please! That needs to be promoted ASAP and his change of heart needs to be explained

 

[LN1_Casey]

Mercedes had a heavy interest in keeping their S-Class and E-Class sales going strong. Why would Mercedes invest in and save Tesla from bankruptcy?

BTW Ford doesn't control Rivian. Ford isn't even their largest investor, Amazon is. Ford has 1 of 7 Board seats at Rivian. Their percentage ownership was not disclosed.

At the time, 2008, the Model S hadn't been released, which was the vehicle to upset the most Mercedes sales.

They also purchased motors and batteries from Tesla for a number of years thereafter for their own vehicle. The $50 mil saved Tesla, but is a middling amount for a large corporation such as Daimler. Perhaps they expected to possibly turn Tesla into a supply making only company of theirs, if Tesla continued to struggle to make it's profit? IDK. I'm not an exec.

Now, for Ford-Rivian; I just don't see Ford being ok with enabling a company to unsurp their Truck strangle hold. Perhaps they bought in just enough to get a hold of the IP.

 

[AcesDealt]

It's part of the regular Wall Street dump & pump scheme that is illegal for everyone except Wall Street analysts, who are exempt from regulations that make it illegal to trade on material information. @neroden called it a 'bear raid'.

Leading up to the Q2 delivery report I expect many of the analysts who (after Q1 consciously and deliberately tried to squeeze leveraged Tesla investors and tried to rattle long term holding investors) to find some excuse before Tesla's Q2 results potentially embarrassingly falsify the nonsense they've been spouting for months.

You are making a falsifiable prediction here -- do you want to stick to it or call it a poetic extravagance? I find you an interesting fellow because your mind summons evidence impressively but usually for obvious and predetermined ends and the conspiracy talk is frequent and always lingers nearest the shadows of most hopeful idols.

 

[Fact Checking]

The ~650 difference is correct if Tesla ever makes 90k+ Fremont Model 3s in a quarter.

So it's not a "game of pennies", right? Which was my point.

I disagree that GF Shanghai will only add 20m/quarter of fixed depreciation.

I went by their expectation of ~$500m capex. If it's higher the $20m figure becomes $25m or even $30m - it doesn't change the end result that the per unit depreciation of Shanghai made Model 3's is going to be significantly lower than the per unit depreciation costs of the Fremont made Model 3's. (Which was more expensive than necessary, both because it was rushed, because Tesla was inexperienced, but also because it was in reality R&D cost of building the Dreadnought.)

Because we're talking about valuation, not near term liquidity.

Firstly, even in valuation terms Tesla's current "depreciation" rates are vastly overestimating the true maintenance capex they would experience in steady state - as @neroden has pointed it out recently. I.e. in steady state this sum would go away with time.

Secondly, "depreciation" as a word is a misnomer: while there's equipment wear & tear, not much truly gets 'depreciated' in the usual meaning of the word, instead Tesla's current high depreciation rate is in large part a "cost of rapid growth", the already paid for capex of the Fremont Model 3 expansion, a fictitious non-cash expense distributed over dozens of quarters. It has no place in a discussion about the long term break-even point of Model 3 ASP.

In fact mixing it into a discussion about Model 3 break-even margins in the context it was done here is borderline intellectually dishonest.

Most stock comp is in opex, it's almost a rounding error for COGS.

But warranty expenses aren't a rounding error, nor other cash allocations that have historically been generous.

Depreciation is a real expense, just time shifted. They spend real cash on buildings, equipment and tooling.

See above and see @neroden's reply, you seem to be ignoring the fact that depreciation is in significant part a time shifted investment expense, not a running expense which is normally used for break-even analysis. Someone in this thread estimated the per quarter 'baseline' maintenance depreciation rate to be around $30m per quarter - which is much, much lower than the ~$120m they are currently depreciating. So the 'true' equipment depreciation expense is more like in the ~$460/unit range, at a 5k/week run-rate.

Anyway, I primarily replied to your false "game of pennies" interpretation, which is false. There's still significant economies of scale to gain, and if the news about GF1 mandatory overtime are true then it means that at least in June Fremont is able to sustain a production of ~1k/day ~7k/week.

 

[Fact Checking]

You will frequently see on Reddit or Twitter references to the sr+ being effectively nil margin, or they will reference declining ASPs which is also a reference to same problem.

I might have missed your reply among the many responses, but could you outline this "disappointing" Model 3 SR+ margin story in actual numbers, I simply don't think it's there at all:

And crickets ...

@AcesDealt, is there any reason you are throwing out baseless figures and innuendo about SR+ margins, and then don't back it up?

 

[RobStark]

EV Sales: Sweden May 2019

 

[Fact Checking]

It's empirical. No one knows including Elon. A specific kind of Brute force algorithm worked for chess but the algorithm was insufficient for Go or poker or etc.

You really don't know what you are talking about there. Contemporary AI's:

• are self-learning in less than a week to play better chess than the best human chess players,
• are self-learning to play better GO than the best human GO players,
• are self-learning to play better poker than top human poker players.

Literally 100% of the examples you listed are false. You could have listed some lesser known game that nobody has bothered to cover with an AI yet - instead you had to mention GO and poker, the games were AI's are already (crushingly) better than humans including those sector experts who spent a lifetime playing only that single game ...

Not sure what you meant under 'Brute force algorithm', but unless you back it up with links I'll just assume it's gibberish too.

State of the art neural nets are not capable of implementing human level abstract intelligence, which comes into play when contemplating how various objects are about to interact with each other in a novel scenario.

False.

it is unknown whether current narrow AI logic is sufficient.

While of course we don't know the future, we do know that your facts, your premise and most of your conclusions are false.

The worst mistake is to think the hardware chip is a big differentiation. It just makes the system behave poorly in corner cases faster.

Apparently Intel thought that MobilEye's custom ASIC that has a fraction of the NN inference performance of Tesla's new AI chip was enough of a differentiation to pay 15 billion dollars for.

But yeah, I guess this is the false narrative you wanted to inject into the discussion. Is this another burner account, or are you genuinely this confused about all this?

 

[Fact Checking]

You are making a falsifiable prediction here -- do you want to stick to it or call it a poetic extravagance? I

I'm comfortable joining @neroden calling what is going on currently a 'bear raid', and I'm on record calling Tesla seriously undervalued at $350, let alone at the current $170-$180 levels, but I never make absolute predictions, especially about the stock price. Note the key 'potentially' qualifier in my comment. How long the barrage of disinformation that Wall Street and the mainstream media is spewing, and which you are in part uncritically repeating here is going to last, I have no idea.

Not advice, as usual.

 

[dqd88]

Adam Jonas back to his narrative of monetizing Tesla’s strategic value: “A valuation of Tesla’s autonomous assets and technology similar to GM Cruise could be worth $100 per Tesla share, if placed in a separate entity.” He believes such a strategy would also help Tesla attract and retain top talent.

I have thought that it could be a positive strategic move if Tesla actually did something like that. Spin off the autonomy piece. Remove that cost from Tesla Inc. Put in into a private company. The private co still does what it does today. Only difference is that the costs are under a private co that is valued at $20B+.
It really should be this way as autonomy is a separate and independent venture from electrification. But now we have the public company Tesla having to incubate (absorb the risk and costs) of the "autonomy startup."

 

[AcesDealt]

And crickets ...

@AcesDealt, is there any reason you are throwing out baseless figures and innuendo about SR+ margins, and then don't back it up?

I am confident in two things:
1) anyone remotely curious and looking at the financials will quickly discover the COGS for the model 3 is remarkably high, and in line with all the claims I have made with little math required or little assumption required. It takes almost no time to investigate so I recommend you do it yourself.
2) that factchecking is one of the least competent and most diversionary members of this board specifically because he operates not from evidence to conclusion but the reverse almost every single time, not to mention his incredibly fast movement to conspiracy talk whenever threatened which is a low probability conclusion. I highly recommend both neroden (although I find flaws also), and doggydogworld (who so far seems pretty spotless).