-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable
- Thread starter AudubonB
- Start date
For autonomy and robotaxi to truly be competitive with Tesla, they need full integration. Nvdia selling their hardware platform to everyone else is going to result a mishaps of different autonomy experiences which will be the end of Nvdias autonomy effort.
You are not understanding right. If you own a call that is in the money and you exercise it, the purchase date of the shares is the day you exercise. If you wait a year after exercising before selling the shares, you will then have a long term capital gain (or loss).
Your cost basis is the cost of the shares plus the cost of the option.
Exercising the option is not a taxable event. You don't owe taxes until you sell the shares.
If anyone sees anything wrong here, let me know. I am assuming a taxable account in the US. I am not an accountant.
They don't inherit the purchase date, but the cost of the call is added to the call strike for a new cost basis on the shares.
You should ask your accountant and your brokerage so that you are confident in the answer. But the answer is yes in the United States. You've purchased shares and have not sold anything hence nothing to tax. The premium you paid for the options is added to your cost basis and you'll sit on an unrealized gain until the day you sell. For example a single option to buy 100 shares at $100 that cost you $100 ($1 per share premium) would, after being exercised, show a purchase price of $101 per share for 100 shares. Nothing to be taxed.
I did exactly this with 25 dec options I had for a $300 strike. Was a short term play last year when we traded sideways and instead of paying $300k+ in taxes I used margin to exercise the shares. Took on $750k in margin that costs me $21k in buying power annually. I can add zero cash theoretically into my account and as long as my account value grows higher than $21k annually I wouldn't notice any change. My tax documents also reflect no giant short term gains that will be taxed for 2020.
Nvidia keep moving the goal post. Look at the orin slide. How much hardware is needed for Level 5? Looks like it's 400 Tops today, but 2000 Tops in 2 years. How come this fsd computer is downgraded to L3 in 2022? You would think with two additional years of training my self driving ability would be better not worst? Or is Nvidia just bullshitting throwing arbitrary nonsense out there to sell hardware?
Mo City
Active Member
If Tesla cracks autonomy in the next couple of years, they will be both the Apple and the Android of FSD. Tesla can offer instantaneous scaling to millions of vehicles with one OTA vs a comparative snail's pace for other OEMs.
No disrespect to Nvidia.
While I'm sure the Nvidia hardware is excellent, the rate at which it achieves FSD is the rate at which the first Nvidia customer achieves FSD. That rate depends in part of who big the customer's fleet is, and how much data they can gather.
Tesla will not own 100% of the EV market or FSD for any lasting period of time, but Tesla can achieve an out sized growth in market share because it will be the first to achieve a working, affordable, efficient and profitable solution at scale.
The key point competitors are now missing is that game has moved on and scale ASAP is very important.
This is the reason why I don't doubt that Tesla can scale to produce 20 million EVs per year by 2030 and sell every one,.
Tesla will almost certainly be the first to have 1 million regulatory approved Robo-taxis on the road..
So my view remains Tesla can grow into any valuation between now and 2030. After 2030 they are going to have to significantly diversify to maintain the growth trajectory... clean energy and transport will be a mature market at scale.
There are other disruptive opportunities out there, in fact at present I'm finding it hard to name a single industry that can't be disrupted in the next 10-20 years. If you read the Tony Seba report I linked in the appropriate forum, you know that includes live stock farming and food production in general.
Last edited:
I noticed a lot of questioning of why anyone would disagree with this. I was the disagree and I did so not because there wasn't a rumor of this but because it was roundly debunked shortly thereafter. Don't like seeing false information, especially overly optimistic speculation on Tesla spread.
I don't recall the debunking information. Do you have a link to it?
Knightshade
Well-Known Member
Except he was already looking at selling shares to cover the bill.
If you just sell shares, you lose shares. 100% for sure.
If you instead sell a covered call you not only might not lose shares (but still get money), if you do lose them you get more money for them (both the premium, and the higher strike price, as compared to simply selling the shares outright.
One would need a better idea of specific numbers.... how many shares he was considering selling, how many (in blocks of 100) he's got, etc... to get a better idea how good a plan it is (or how far OTM one could sell them... Selling $400 OTM weeklies is an hilariously tiny risk for example but you'd need to sell a decent # to cover much of a tax bill over a few months)
But I'll mention another aspect of the idea depending again on his account position.
Is he using margin?
If he's got a pretty significant amount of that available he can buy 100 shares (or 200, or whatever depending on what represents a fairly safe % of available margin) and sell covered calls very near the money on those.
Premiums should be quite good- and if THOSE shares get called away- great... you make a few extra bucks on top of the premiums too.
For example, selling a Jan 15 $900 strike right now is showing me a $3000 premium.
If it exercises he collects that, plus $2000 profit on the sale of the stock (assuming he can buy the margin shares at 880, but same basic idea whatever the current vs slightly OTM amount is)
5 grand for paying maybe a few dozen bucks in margin interest for the week- multiple 5 grands if he can comfortably do that multiple times.
The nearest to a downside here (again assuming he's only using a pretty safe % of available margin) is if the stock drops significantly, he's "stuck" with the shares.
But he still made the 3k premium (per margin CC he can sell)....and he can keep selling at-least-what-he-paid covered calls against them that'd deliver profit above the margin interest by a wide amount unless the stock just flat collapses....
and you don't expect it'll do that, right?
That... has not been my experience with selling covered calls. Especially against shares I bought with the brokers money.
Hasn't been the experience of quite a lot of others as well- there's a number of other threads further discussion is probably better off in though.
The poor judgement seems to come in when people sell them for too-aggressive strikes against shares they just CAN NOT stand to lose.
This guy began with he might have to sell shares to cover a bill.
This idea presents him with an alternative to that- where he might not need sell them at all, or that he'd get more money out of doing so if he does.
Oh- one other thing I'd mention to the person with the tax bill- assuming you're in that 90% situation and can wait till April 15th to pay.
Look at opening a rewards credit card (or more than one depending how big that tax bill is) to pay it.
You'll pay a 1.something percent fee to do it, but can usually get somewhere on the order of 10-20% cash back compared to required spend on the signup bonuses.
(you can potentially get a lot better than 10-15% back in terms of travel rewards- but you need the cash it seems, and not a ton of travel going on right now for most anyway)
Last edited:
No I did search the news and forums but couldn't find anything specific to support my recollection. The rumor broke Nov 19 and Rob Maurer even discussed it in a video that day but as I remember it turned out there was no evidence that the stock Berkshire was investing in (but had to keep hidden) was Tesla and reviews of comments from Buffet made it highly unlikely that it was Tesla.
Ah, Ok. So while it isn't likely, you are saying that there is still a chance that it is Tesla that Berkshire bought and didn't want to disclose...
In a little more than a month we will hopefully know...
pz1975
Active Member
I don’t think the average person understands Tesla very well at all - I didn’t say that. I believe more people would buy a Tesla now more than ever and the number of people who would consider it when looking for a car is going up faster than ever, and this is being helped by the positive news about the rocketing TSLA share price and Elon being the richest person in the world.
There is one fundamental flaw with the nVidia approach. Don't get me wrong, I am long NVDA and a shareholder, but the problem is that FSD computer is an application-specific design. Everything in there serves a purpose and at maximum efficiency possible. And as it's intended ONLY for Tesla, they can cut a lot of fat, and focus on what matters
nVidia designs chips with everyone in mind. Thus, it includes components that would maximize clients instead of optimal performance/effciency.
And I was very underwhelmed by NIO day. Same like many other Chinese firms. Talk about hardware prowess, and how they "best" their competitor. But when you think deeper... NONE of their meaningful technology, be it camera/sensor suite, autodriving processing unit... etc. They are all 3rd party solutions.
This is what legacy automakers have been doing. Taking OEM/ODM partners solutions, put them together, add their own touches, and call it a day.
This is why legacy automakers are lagging behind TSLA and until they start developing meaningful technology on their own, they will always be dependent on others. It's something that works for ICE cars. But again, EV is closer, architecture-wise, to an iPhone than an ICE car. So, if legacy manufacturers, or EV players like NIO don't stop thinking in the way ICE car work, they will never catch up.
Why did Apple decided to switch to their own M1 architecture in their newest computers, even though the cost of such move is basically giving up all the application libraries that their developers have done for the last decade? Yes, there's emulation, but it only goes so far. For years, as an Apple Mac user, I've been complaining that Mac is just no longer what it used to be. It lacks performance comparing their PC counterparts and if it wasn't I'm too heavily invested in the MacOS platform, I'd probably have switched back to PC.
Looking it back retrospectively, I got it now. Apple's computers rose to a new height thanks to shifting to x86 architecture, but as Apple becomes more and more dependable on Intel's schedule on shipping new processors, there's only so much Apple can do to their computers. That's why Apple first decided to start developing its own chip solution for their phones, and now they no longer need to be dependent on Intel as they are able to offer solutions that are not only comparable to Intel, but leap ahead.
Now coming back to TSLA, using that very same idea of Apple's decision... I believe the reason TSLA went on to design their own FSD computer is that they see FSD as a very important aspect of their future. And they don't want to depend on anyone when it comes to one of the most aspects of their business.
Apple never had the best hardware spec on their phones comparing to the rest of the industry. But because they operate such a closely controlled hardware-software environment, they are able to optimize their phones in levels other phonemakers can only dream. Same idea with TSLA as they keep a tight control on their hardware and software design. And it's a luxury to have when you are able to attract the best talents.
Ah, Ok. So while it isn't likely, you are saying that there is still a chance that it is Tesla that Berkshire bought and didn't want to disclose...
In a little more than a month we will hopefully know...
I was a BYD investor before Berkshire took their 10% stake in 2008 so I know there were those including Munger at Berkshire who were enthusiastic about EVs and the alternative energy revolution. But Buffet himself I don't think was really on board and together with BYD's relative disappointment and his low opinion of Elon I would say there's a chance but...
These processor architecture wars go back and forth all the time. For example, Sun didn't get enough performance from 68xxx based chips, designed a faster SPARC chip that was great for a number of years, then got killed when Intel caught up with faster, cheaper, higher sales volume processors.
Eventually (maybe in 5 -10 years), nVidia will have so many customers for FSD chips that they will have the volume and resources to build something FSD specific that is just as good or even better than Tesla's chip.
I get that. There are ways to get around the rules the law of physics apply on electrons: higher voltage, different battery composition, different battery design/size... I can go on.
But there is only so much you can get around on the part of law of physics that apply to swapping a battery pack this big, which to most part, is purely a mechanical thing. You can only automate so much. They still take time.
Think about how phone chargers has evolved. We went from a pitty 0.5A of USB1.0 that takes half a day to charge a phone to today where they just changed the charger, changed cable, and changed the entire architecture internally that now we have phones that when paired with the gears, can go from 0-80% in about the same time it takes me to brew a pot of coffee. And they aren't stopping. There's a new charging tech every single year.
And when you consider everything in a larger picture, given the cost of the station, the cost of the batteries in there, the real estate, the people operating them... etc. I just fail to see how such a solution is able to scale when you have millions of vehicles on road.
How many batteries and stations do you need to keep around? The idea of scale means that as you build more, the cost of it drops. From what I can see, the average operating cost per station looks about the same to me when it's 1 station vs. 1000 station. It doesn't get much cheaper.
Worse, when the supercharger sits empty with no users, the cost is just the real estate cost+equipment depreciation cost.
When NIO charging station sits empty with no users, the cost is the real estate cost+equipment depreciation cost+battery pack depreciation cost+operator wage (for now at least)+logistic costs of battery packs as they become unusable for whatever reason, or you intend to use data collection to swap the packs around between 1 station (with less usage on that period) to another (with higher usage on that same period)
NIO's battery swap needs everything power-supply-wise that a Supercharger site needs if they want to make sure they can have those packs charged by the time next car comes. That's why I said I cancel them out because I can't see it much less than a 4 spot supercharger (the space NIO swapping station needs).
Look at that video. I counted easily 50+ Teslas waiting to charge. This is a place middle of nowhere between LA and SF. And it's quite often like that during long weekends/holidays when people make that run between LA/SF. You can easily see why NIO's swap station wouldn't work for these places. Because other than these days, there simply wouldn't have any demand for stations. So, say if you have 2 or 3 stations there, they'd just sit empty for the most part of the year.
This is what my research concluded on why battery swap is not sustainable in the long run.
It's a great marketing gimmick that says "hey look ICE drivers, with the battery swap, we can also go from 0 to 100 in 5min".
But for most here who have been a Tesla owner long enough, we all know it's more like phone... we only need to charge it so it has enough juice to last to our next charging point/home.
Rough guess
Cost of Samsungs 12inch 14nm Wafer is $2000(it was 2500 in 2017 so I'm assuming it cost less today)
You can make ~200 FSD chips per wafer(assuming 85% yield)
Cost per FSD Chip is 10 dollars.
PCB is about 50 bucks
Total cost to Tesla is between 60-100 bucks per FSD computer.
Nvidia charges 2500/Xavier. That's their L2 stuff with 30TOPs. Their L5 stuff doesn't have a public price, but it's suppose to be 320TOPs. Knowing Nvidia, they will charge at least 5k to 10k for something like this.
So that's the Tesla FSD advantage. 100 bucks at most compared to who the heck knows what Nvidia charges. Plus they probably license out their software for additional milking.
You need to take charge of your own destiny or else you're just another automaker.
Last edited:
Not that it's important, but I was just replying to another poster who was puzzled as to why someone else was curious about the Berkshire report. I wasn't suggesting that I believed the rumor. No biggie, just wanted to explain myself.
Similar threads
- Locked
- Locked
- Locked
