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

[TheTalkingMule]

I couldn't possibly care less about S/X demand, it's all about the main product to me. Today that's the M3 and demand still appears to be far stronger than their ability to deliver. Pump em out. Once the Y comes out, I will rejoice in M3 demand falling off a cliff as the newest product takes over the world.

To me, that should be Tesla's role in the automotive world. Pushing marketplace evolution, not looking to incrementally dominate every category. From an investment standpoint, we will truly begin to profit once Tesla is the largest energy company in the US. There's no real money in just making cars.

 

[Right_Said_Fred]

Bought 66 more at 250. My average is now below 300 again, at 294.

 

[MarcusMaximus]

All nice and good, but there's the issue of the stationary fire trucks, while fully focused all the time.
I am talking about soft target, pedestrians and cyclists. How are Tesla going to sort interaction with those?

Perhaps Level 5 will be homologated, but only for a cherry picked set of streets within an area, making RoboTaxi more of a city bus than a personal car. Then there are unexpected road closures. I sometimes get to consult a road worker or street sweeper. Oh well, people here are planning their robotaxi drives for Christmas already

FWIW, stationary fire trucks(or any other stationary obstacle) don’t seem to be an issue anymore, and haven’t been for a while. EuroNCAP 2018 testing results showed an AP2 Model S can safely stop for a fully stopped obstacle at the upper limits of their testing speed(~85mph as I remember).

 

[neroden]

Either someone didn't do their homework at all, because it should be easy to see that if you stop selling your highest volume configuration you'll lose a bunch of sales, or they expected Panasonic to deliver batteries so that the M3 sales would make up for any loss. Neither happened, so the results are as stated.

Or they expected to get the updated motors out sooner. If the 370 mile S 325 mile X is released in January or February, they probably sell at least 5000 more.

I think they expected to get the updated motors out in February and they expected Panasonic to deliver the batteries they were supposed to in January and they expected the Euro-spec parts to arrive on time in January. None of the above happened. Dangers of optimism...

 

[dc_h]

I say pana effed up...
I say buy them out already.

I think Elon is putting the hammer down. No new infrastructure headline was Tesla telling Pana to fix the processes in Sparks. I think Tesla and Grohmann are likely now involved with Pana in fixing assembly issues and reducing production losses and getting volume up to potential. If they are planning 90-100k deliveries in Q2, that means they most have already solved some cell production issues.

 

[jerry33]

I couldn't possibly care less about S/X demand,

Given the profit in S and X, and that people buy top down (that is they buy the lower priced cars because of the halo effect of the premium cars), it's very important to care about the demand.

 

[KSilver2000]

The Tesla truck reveal can’t come any sooner.
The Ford F and Ram sell far more than any other light vehicle in the US. And with Ford’s investment in Rivian, Tesla needs to hurry.

 

[DurandalAI]

Then I missed that.

Yeah, the topic of which way a pedestrian was facing, if they are looking down at a cell phone, etc was mentioned. Essentially all those visual cues that we inherently pick up, the NN can also pick up in time. (Observe all children and their projected vs actual trajectory, base future predictions on lessons learned.)

 

[neroden]

Or Tesla asked the impossible. The truth may be in the middle.
When Honda was disappointing McLaren, it seemed very one-sided forcasual onlookers. After the fact the real situation became more clear.
I don't understand, if they made margins in the past, why they didn't order over-capacity in cell lines.

Arguably, they did. A shortfall from 35 GWh of production capacity to 24 GWh of production is a pretty extreme scrap/downtime percentage. I think they probably "needed" about 30 GWh to do 7K/week on Model 3 and ramp up Powerwall/Powerpack, which still allows for a large rate of scrap/downtime.

Panasonic screwed up bad.

 

[anthonyj]

The Tesla truck reveal can’t come any sooner.
The Ford F and Ram sell far more than any other light vehicle in the US. And with Ford’s investment in Rivian, Tesla needs to hurry.

Tesla truck will be a niche product. Elon said this many times. It is going to be extremely large, weird/cool, and probably very expensive

 

[KSilver2000]

Gonna stop looking at my portfolio for awhile, for my health and well being.

Then you’ll miss opportunities to buy in even lower like today.

 

[elasalle]

Interesting, NOVA not showing any M3 inventory with 200 miles search

 

[EVNow]

Interesting thing about Model S/X deliveries.

Q1 '18 : 21k
Q4 '18 : 27k
Q1 '19 : 12k

So, if we assume we had even 6k pull in to Q4 (27k-21k) - and add that to Q1 '19 - we get to 18k. Down slightly from Q1 '18.

But there were so many other headwinds as well
- Demand shift to 3
- Anticipation of refreshed S/X
- Removal of SR S/X

BTW, it was funny how Tesla team tried to say there is no Obsborning because the trade-ins of S/X is only 3%. But, it is not 3% of S/X normal sales, but 3% of a much higher 3 sale.

3% of 50k = 1,500 = 12% of 12k.

Ofcourse, the trade-ins for 3 is hardly a good measure of demand shifting from S/X to 3. How is it compared to trade-ins for S/X, for eg ?

 

[Ratanpara]

Hang in there all long term bull, here is how I look at it, market or us even long term bull will have difficulty believing guidance of 90-100 K unit, stock is priced at missing this estimate, If they miss by little current price is ok, if they miss by big margin further downward pressure could come, perhaps nothing new ,Amazon , Netflix they all went through this period, But looking out 5 years span Tesla will kick ass.

 

[Spacemanspliff]

For S & X, there was clearly some drop in demand (perhaps just the tax credit cliffs in US and, was it the Netherlands?), but an awful lot of it seems to have been the planned line changeover. And there seem to have been Euro/China delivery problems again. It looks to me like the line changeover took longer than it was supposed to.

As far as I am concerned, the biggest miss in recent quarters that Tesla has given up on 10K/week out of Fremont. That was truly bad news and the drop from $360 to $250 is partly justified by that. Getting only 5K/week in Q1 rather than the new target of 7K/week was also bad news and accounts for more of the drop. Our resident institutional investor says that his sources say that production is going up so the battery cell bottleneck hopefully really is fixed.

But essentially the bad news was in the delivery report; the earnings report has a lot of goofy wildcard stuff going on which doesn't matter (they do seem to have thrown in the kitchen sink) but the bulk of the loss is from just not making enough cars. IMO.

On S&X I agree on most of the points you stated, but also believe that there was a small bit of cannabilizing by the 3 as well. The 3 for the most part is a cheaper product and arguably has better tech, the supercharger event only magnified this to the public. The powertrain refresh is good and SHOULD alleviate some of that, as well as the free ludicrous offer demand lever for old S&X owners.

The 10k to 7k is what about an 8 Billion dollar annualized trim off the top?

 

[ItsNotAboutTheMoney]

I'm sceptical the camera resolution and frame rate will suffice to tell the difference between a drunk bike rider about to weave and one that will keep course. Or a rear rack passenger about to hop off. A weave from someone just getting going along the sidewalk. Also with young children, I can tell from their finer posture nuances whether they're about to leap right in front of my bike. I see what their eyes are focused on, interaction with people, animals and objects. Give me 100 such situation and I will beat 2025 FSD at it. What could help FSD is the response time but they'd likely be too late already.
I warn people about to step in front of my bike (voice, bell) all the time. In a car I'd have been slowing down before the point I'd like to alert the pedestrian. How can FSD ever be 10x better than me at seeing such accidents coming? An infant under a wheels is just game over.

As long as the autonomous car recognizes a danger it can react quickly, significantly reducing speed from the already-legal speed at which it was traveling, and perhaps avoid the collision entirely.

People tend not deliberately to step out directly in front of cars, and if they do, and the driver has clear evidence of it, the driver is not held responsible.

There _might_ be some kind of public outcry when an autonomous car crashes into a young child and the young child dies, but I suspect that clear video evidence of an unattended child giving the vehicle little-to-no time to react would rapidly redirect public anger at the parent.

 

[anthonyj]

Honestly Musk should raise capital now. I really don’t see the stock going over $300 until their cash on hand is a lot higher. Nothing he says or tweets does much to the SP. He lost his magic pump touch. Doubt many institutions believe his guidance at this point. The money can be used for growth, they need to build the Semi, Roadster, and Y next year. Stock would get upgrades and shorts would cover

I think Musk, Ellison, and Saudi’s can raise $3b by themselves

 

[Nocturnal]

Yeah, the topic of which way a pedestrian was facing, if they are looking down at a cell phone, etc was mentioned. Essentially all those visual cues that we inherently pick up, the NN can also pick up in time. (Observe all children and their projected vs actual trajectory, base future predictions on lessons learned.)

Not to underestimate the need for this or how much I'd like to see fewer children killed...but, FSD just has to be better than humans on average. I'm not confident that the average person is paying attention often enough to predict that a child will run out in front of their car. Then take away speeding, which is sadly so common in residential areas. I drive very fast myself, just not in places like that.

 

[DurandalAI]

At the risk of sounding insensitive, FSD needs to accomplish two things:
1. Ensure that the risk of harm to a human either inside the FSD Tesla or outside it is 2-3x lower than a human driver.
2. Ensure that any remaining collisions that occur are NOT the fault of the FSD computer. (If someone runs a red light, hits the FSD Tesla, it's not the FSD computer's fault. In time, the likelihood of the FSD computer preventing such a collision despite it not being at fault will increase, which will improve task #1 of reducing the likelihood of injury.)

There are a ton of edge cases, for sure, say people deciding to lay down in the middle of the street (I've seen it before), people falling out of vehicles, etc. Most of these are general obstacle avoidance. Other edge cases involve road construction or damage. I think level 4 plus remote "instruction" to the cars will be quite possible within the next year to two years.

 

[Dan Detweiler]

There will always be more edge cases. The way the car sees the world changes. It's like a whack a mole. And people's behavior change too. It will be a constant struggle. It is extremely hard to anticipate a timeline given how complex the problem is.
.

So here's my rub with what you're saying. I don't argue the fact that the driving environment will continue to evolve and change. The human eye has been able to adapt and change over the last 100 years or so. Cruising speeds have increased with safer road surfaces and better drive trains. We have adapted. Granted, you could make an argument that we don't all do it very well but none the less, we manage ok. If we can do it then there is no reason a neural net can't do it, and do it exponentially faster. I think we are living in incredibly exciting times and I think things are going to happen faster than any of us could dream possibe.

Dan