I'll throw the British South Sea Trading Company into the ring for that title of most overvalued company ever.
It had the king himself on the board. The company value represented a decent fraction of the national gdp at the time. All without actually never producing anything of actual value. It was just bribes and speculation all the way through. It's wild.
The Wikipedia user based his drawing off of this sculpture from 1711, so it seems the american continent was actually already quite well mapped back then.
The armorials of the South Sea Company, according to a grant of arms dated 31 October 1711, were: Azure, a globe whereon are represented the Straits of Magellan and Cape Horn all proper and in sinister chief point two herrings haurient in saltire argent crowned or, in a canton the united arms of Great Britain. Crest: A ship of three masts in full sail. Supporters, dexter: The emblematic figure of Britannia, with the shield, lance etc all proper; sinister: A fisherman completely clothed, with cap boots fishing net etc and in his hand a string of fish, all proper.[61]
The artifact you link shows a map of the Americas in which California is an island and either Tierra Del Fuego is huge or the bottom of Argentina is an island and the northwest of the continent trails off into nothing, and Florida is sort of a stubby nub (other maps from this period show a more accurate Florida, so this might be a small-size-of-the-object problem).
They had a decent view onto the east coast of the Americas, but after that things got quite inaccurate. It's like... I don't know what anyone's expectations are, but it certainly isn't the perfect world map that's shown in the main image of Wikipedia's article.
We should stop comparing incomparable things, like companies market cap (measured in dollars or pounds) to national GDPs (measured in dollars or pounds per year,) unless we want to reach outrageous but incorrect conclusions.
It's a dubious comparison but not because of "per year". The comparison is implicitly to one year's worth of GDP, which is a currency amount.
It's dubious because whereas a year's worth of GDP has some claim to actually being the value of something (with many caveats but it's engineered to behave like that as much as possible), market cap isn't. It's the amount all the shares would cost if someone bought them all in one go for the price some shares were most recently purchased for, which would never happen.
Hashicorp was bought for $35 per share at a time when it was trading a little above $25. Not saying crazy market caps aren't a sign of a bubble (not sure how you'd read that in my comment), just that market cap is not the value of the company.
Variation in price doesn't prove that the market cap is not (a good estimate of) the company value for highly liquid stocks.
Value is subjective. Stock prices measure peoples perception of the value. Your thesis that it is incorrect can only come from 2 places (I think)
1. Dumb money - the market cant see that XYZ is overvalued or undervalued. My rebutal there is nonetheless XYZ has been valued by a conpletely open continuous auction that people are not restricted to participate in.
2. The parts are less than their sum. This may be somewhat true... total control over a company may be more (or less) valuable than splitting. But I dont think it is order of magnitude. And if it is, it is because the value to you isnt the value to me (the value of RAM to a gamer < value of RAM to OpenAI).
Well, based on that last share purchase, we have incontrovertible proof that there was indeed one person in history who thought it was worth 3x GPD.
And the fact that in the entire BSSTC shareholder universe, there wasn't any noticeable volume for a sell, or a registered sell limit, at a lower value leading up to the last peak.
That must have been a rough trade, but someone got something out at the last moment.
1. No, we don't have proof that there was one person who thought it was worth 3x GDP. We have proof that there was one person who thought a 0.001% share of the company was worth 0.003% of GDP or whatever. They could think it was worth that much for plenty of reasons; maybe they thought the share price would grow for a bit more before collapsing so that they could make some profit, maybe they invested in order to just be an investor and have a say in investor meetings or however things worked back then. Maybe it was a status thing.
2. Why are we using the opinion of one random person to determine the value of a company
To add to this, this type of thing happens all the time in crypto.
A coin will release 1/1000000000th of it's eventual supply, have some trades at 10c and then claim the value of the entire supply as the headline value.
That makes no sense. House price/salary is used to compare payment periods or affordability. The context is important. Share value and GDP are totally different things and there is no direct relationship.
It is effective but wrong. It is the same as to compare a river flow to dam volume. There is a relation, but does not mean anything.
A better comparison is to compare somebody's net value to the total housing value of a city. For instance comparing Musk's net value to the total value of all the houses/apartments in New York. Or the the value of the gold in Fort Knox.
What’s the issue? If a market cap is three times an annual GDP, it means it would take three years of that economy’s entire production to purchase the company. It’s obviously a very fuzzy measure for various reasons, but it can give some idea of the market value of a company.
For people who are unclear what Palantir actually does
Palantir is a tech platform that consumes data from their clients in return for providing high level data-driven insights. They assign FDEs (or consultants) to really learn the details of a customers data. Foundry allows them to get single pane view of the data in an org and they actually have both the tech and engineering skills to do the dirty data cleaning jobs.
For an extravagant fee, you give them your data, they clean it for you, and then those same FDEs can tell you interesting things that you should have known, had you actually done proper data architecture in the first place.
Does it add value, yes. Is the value worth the fee? Like snowflake, they throw some very very good parties.
The reason Palantir is valued as highly is because it is a Peter Thiel company, and he also owns the Vice President of the United States.
The thesis is "The current US administration is making enormous profits from their powerful positions. Whether of not I agree with this, I want some of that money too".
I'm not sure why this isn't mentioned because it's the number one factor driving the price.
And then you have Thiel going crazy about biblical end times prophecies involving the anti-christ and I'm starting to wonder wtf they're doing at Palantir.
It is pretty clearly labelled with "(Sponsor)". Couldn't be less obtrusive, either. I think on balance it's at least neutral compared with the other nonsense available to both the publisher and advertiser these days.
It seems like the financial outlet made a maths error and then was amazed by the absurdity of the incorrect result they got.
Specifically their whole outrage is based on having accidentally calculated what would need to happen for Palantir's revenues to grow 1500x as opposed to 15x. The corrected statistic would be the Palantir requires 11.4% revenue growth year over year not 35%.
90% of TSLA income is from EV sales. Everything else is just Musk's grandiose and increasingly absurd predictions which have a long history of falling short of reality.
Meanwhile the story of Jensen and Musk continued onward with custom chips to support FSDv1, in which J's personal delivery of the DGX1 to OpenAI served as the catalyst to the relationship, iirc.
The problem for nvidia is ... where do you go from here and still have spectacular performance improvements?
Nvidia got extremely lucky again and again and again, and what specifically did it is that right in time non-Nvidia researchers learned to train on smaller floating point bit lengths, which Nvidia raced to support. And great, well, done! A list of ironies though ... for example it's Google Deepmind that made the Turing generation of cards viable for nvidia. However, the new floating point formats train has arrived at it's last station, the NXFP4 station. There is no FP3 and no FP2 to go to. Is there a new train to get on? I'm not aware of one.
Nvidia's argument is "Blackwell easily doubles Ada performance!" ... but that is deceptive. The actual improvement is that Blackwell NXFP4 (4-bits) is more than double Ada FP8 (8-bit) performance in ops. That's the train that's arrived at its last station. Go back further and the same is true, just with larger and larger FP formats, starting at FP32 (single precision). Aside from a small FP64 detour, and a few "oopses" in some of the format they chose turning out useless or unstable, all quickly abandoned that's the story of nvidia in ML.
Comparing, for example, FP32 you don't see big improvements: e.g. 4090: 83 FP32 TFLOPS, 5090: 104 FP32 TFLOPS. Given the power requirements involved that's actually a regression. If you're stuck at 8 bits, nvidia's story breaks down and Ada cards beat Blackwell cards in performance per watt: 4090: 5.44 Watt/FP32 TFLOP, 5090: 5.5 Watt/FP32 TFLOP. Or, FP8, same story: 4090 is 0.681 Watt/FP8 TFLOP, 5090 is 0.686 Watt/FP8 TFLOP. Now effectively the new memory still buys some improvement but not much.
Will the next generation after Blackwell, with the same floating format as the previous generation be a 10% improvement and subject to further diminishing returns and stuck there until ... well, until we find something better than silicon? I should point out 10% is generous, because for FP8, Blackwell is actually not an improvement at all over Ada, on a per-watt basis for equivalent floating point lengths.
Plus Blackwell is ahead of the competition ... but only 1 generation. If nvidia doesn't get on a new train, the next generation of AMD cards will match the current nvidia generation. Then the next TPU generation will match nvidia.
Yes. Most companies play these financial games to some extent.
I lumped government subsidies in with EV sales since they are related. Trump wiped these out.
Robotaxi and robots are the fantasy category. They are not currently income producers and may not be for years to come. His robot demos have been widely panned as fake.
For the sake of democracy and freedom of human expression, I hope that their software is a bloated, bureaucratic mess and can’t power the efficient dystopia which Alex Karp wants.
To me it would depend a lot on how US/EU relations continue to grow/break. If digital sovereignty ends up putting Palantir in a position where they won't be able to seel their intelligence tools to European secret police organisations, then it's probably overvalued. If relationships work out fine, then they'll basically be the Microsoft of mass survailance, which may make them overvalued but surely not the most overvalued company that ever existed?
Once you’re embedded in govt, the ‘OS’ for government, then you can raise charges in order to keep the hoards at bay. Hoards that are angry at cuts to public services made in order to finance Palantir profits…
What 4/5yr term govt can turn off or replace their OS?
I do not think the EU is critical. Palantir's potential market is global, so one region is not that important to them. They also sell to the private sector and this will increase, and EU countries will have different policies so losing a few EU governments business would not be that big a deal.
Right now 73% or their sales are to the US, and the next biggest source of revenue is the UK, far behind at 10%. The whole of the rest of the world, including the EU, adds up to 17%.
Is their market really global though? If it wasn't for political reasons I suspect everyone would be buying Chinese mass surveilance tools. Traditional US allies won't do that, at least not yet, but why would South America trust US over China?
Would you trust China? it has an incomparably worse record than the US.
A lot of the biggest potential markets are in Asia, and almost all Asia countries fear China, and to many of them it is their main potential military threat.
IMO the best interests of either of the two countries of which i am a citizen would be to be less dependant on the US, however the US is nothing like the threat China presents.
Edit: to add, they are competing with whatever equivalent product China has to sell anyway.
For the Mexico example, China isn't the neighboring USA whose president out of nowhere started talking about taking over it's only other neighbor, or government overthrow in Venezuela. Or the recent attempts to sway elections in other central American countries, pardoning of convicted drug dealing ex-Presidents. Attacks on Columbia's government. Sweetheart deals with central American dictators to illegally house immigration deporties in sub-human conditions with zero due process.
It isn't. It's just that they're doing it where others aren't. They're a child of Total Information Awareness that has never realized or just doesn't care about the societal poison pill inherent to it.
The magic sauce in terms of bubbleability is that the product is all secret and investors can't form any natural opinion beyond what's published in company performance publications. Because they aren't allowed to know anything beyond that. It's just blind trust, "governments have bought before, they might continue buying, perhaps buy more". Reminds me a bit of the psychological mechanisms that make a con work, might be something to it.
There's a few things. Let's start with the core of what they say is their value. They have forward deployed engineers - this a totally new, previously unknown innovation - who go to a company, understand their needs and build data processing tools to give them insights. Then, they generalize these tools so that they can essentially sell them as SAAS software, giving them SAAS-type economics.
What other people say is their secret sauce, is they do consulting work for the government (a forward deployed engineer is just a consultant) and they make incredible margins because their senior management and early investors have connections to the government which gets them exclusive access to incredibly juicy contracts. As these contracts paid off they leant heavily into the social media meme stock trend so their CEO spends time talking like a psychopath and doing various non-economic things like spending huge amounts of money running adverts about how they're going to use AI to unleash Americas workers (America's workers aren't able to buy Palantir software or services, but they can buy it's stock).
OK. The first part sounds a bit like an innovation.
I was kind of expecting someone to say it had EG really sophisticated ETL tools that can normalise loads of different data or can query across disparate data sources or something.
The first part sounds like a basic ERP implementation. Only instead of leaning on your in-house domain experts who have years of experience/relevant knowledge/know the relevant caveats, you pay consulting rates to train up new domain experts who don't understand/know the caveats and who will charge you consulting rates to gain access to the results of the training you overpaid for.
'But they cleaned the data up'. That data was also cleaned up during all the last major system updates. And during the implementation of those systems. And the implementation of the systems before that.
P/E isn't the only way to value a company. A company can be unprofitable, and therefore not have a P/E, and still be worth something. A quickly growing startup that is losing money on paper but has free positive cash flow is one example.
Well, Palantir is the only really large company in the P/E top section. You could argue smaller companies have more upside and could thus justify their valuation easier
Today the fast growing US stocks are those that discount the further concentration of wealth to the current private-government elite. They believe that in the coming years they will flourish at the expense of the middle/lower class.
>Palantir would need to grow its revenues roughly 15-fold (yes, 1,500%) over the next quarter century, implying sustained annual revenue growth in the 35% range over this time frame.
That's actually a pretty reasonnable ask for a tech company, if you believe Palantir can grow to the level of FAANG.
Sustained growth of 35% p/y over a _quarter of a century_ is not a 'reasonable' ask, tech company or no tech company. Virtually no company has ever done that.
One of the concerns with the company though is are they really “software.” Many cite experiences with the where they deploy armies of people doing things to clean up data and other hands on labor for the “software” to work just “software” where you buy a license and off you go.
No. Thats the point. There nothing groundbreaking except their branding and lobby. Every other enterprise sofware company has data analytics platforms similar to Palantir - including the big microsoft / google. Palantir might even be best by a big margin but none of that matters. Nothing would justify this valuation. But what matters is that govs around the world will buy Palantir to show their alliance to US/Trump/Thiel. European far right politicians will happily push Palantir everywhere.
Like Tesla, Palantir is an ideological investment for many, and likely why it's so over valued. Retail investors buy these stocks because they're buying into a world view. Tesla's is obvious (and it's not about green investments). Palantir's is a little less so, but they have built up an image that's right wing (but mostly not culture war), military-coded, and layered on American exceptionalism. They sell to the CIA, they sell to ICE, they help private healthcare companies milk their customers. This is all part of a world view that many people are in favour of, and Palantir represents it well. Andruil will likely do well for similar reasons once they go public.
This is not really to fault them. I see the company as far more mature than Tesla. They've cultivated a brand that works well for them. This is also not intended to be a criticism of Palantir investors, there are many reasons to invest, and even if you subscribe to the above that is of course your right! Many people invest for ideological reasons, myself included, and that's fine.
> Many people invest for ideological reasons, myself included, and that's fine.
It's fine if you don't mind losing your shirt.
For us normal people, we invest to try and grow the money we've saved for the future (in my case, for my retirement), so investing in a company whose stock is insanely overvalued is a great way to blow up my pension pot.
I don't understand the rationale behind "investing for ideological reasons", can you explain it?
Investment for ideological reasons is basically "I like what this company is doing and want to see it succeed/want to make my returns on investment via it", and that investment helps it succeed, to at least some degree. You also get the opposite: "I don't like what this company is doing and don't want it to succeed/don't want to make money from this activity, so I won't invest in it". ESG investments were largely built up from this (much as they generally turned into box-ticking exercises as opposed to a useful distinction). Of course this is generally going to make less money than "I just want to make money/predict winners in the marketplace", but that's not everything people want to do with their money.
The difference being not investing in something for ideological reasons won’t jeopardise your assets; whereas YOLOing your 401k on Palantir because “AI surveillance capitalism is dope lol” is potentially risking significant portions of your future.
Can’t people just retweet Karp on X if they want to “support Palantir ideologically”?
Depends how you do it (I wouldn't put everything in one company no matter how much I liked them), but having money to invest does represent some power over the world so I understand people exercising it (though not why someone would choose Palantir).
GME is not the same thing: while I think there are some true believers in the company itself (and this is part of what drove the initial interest in it), for the most part interest in modern meme stocks is driven by the polar opposite: a kind of financial nihilism that believes hype and gambling on popularity is the only thing left in the public stock market, and they're basically just trying to ride the wave of a crowdsourced pump-and-dump and fleece others via selling to greater fools (and/or creating a cult around a mythical short-squeeze that is reaching ever more fanciful heights).
What I find the strangest is Palantir's market cap and long term valuation show that this is not just retail at the margins - it's significant amounts of money buying up the stock despite its sky high valuation. I don't know if it can be sufficiently explained by "individual actors investing ideologically" or not.
Just another day of the stock market being irrational though, I guess!
> Many people invest for ideological reasons, myself included, and that's fine.
Sure, but I'm under the assumption that many people, yourself included, are conflating "buying some company stocks" with "investing".
To first order approximation, it does not benefit a company that you trade stocks on the market once they have been emitted, any more than it benefits GM if you opt to buy a second hand chevrolet.
If they keep their shit together and ride the surveillance wave then I don't see what should keep them from increasing 15fold by 2050. Their technology is not at all dependent on AGI, which is why I don't understand why they are always thrown in together with OpenAI et al.
it may well be terrible software, but Palantir are effectively an arm of the new Trumpian state, so guaranteed access (and ownership?) to the largest public sector projects of the largest economy in the world. It's got a good chance of becoming the Samsung of the US, so any value is not an overvalue
Ridiculous statement. If you're actually investing your money in a company then you need to run the numbers, it's the most basic kind of due diligence you should do.
I'll throw the British South Sea Trading Company into the ring for that title of most overvalued company ever.
It had the king himself on the board. The company value represented a decent fraction of the national gdp at the time. All without actually never producing anything of actual value. It was just bribes and speculation all the way through. It's wild.
https://en.wikipedia.org/wiki/South_Sea_Company
Extra History did a more easily digestible series, which was how I learned about it in the first place. https://youtu.be/k1kndKWJKB8
Unrelated question: did we have a good enough idea of the world map in 1711 for the coat of arms shown on Wikipedia to be accurate?
No, and the article shows a not-even-attempting-accuracy period version of the coat of arms a little below: https://en.wikipedia.org/wiki/File:SouthSeaCompany_TradeLabe...
The Wikipedia user based his drawing off of this sculpture from 1711, so it seems the american continent was actually already quite well mapped back then.
https://www.rmg.co.uk/collections/objects/rmgc-object-36062
The Armorials section in Wikipedia:
The armorials of the South Sea Company, according to a grant of arms dated 31 October 1711, were: Azure, a globe whereon are represented the Straits of Magellan and Cape Horn all proper and in sinister chief point two herrings haurient in saltire argent crowned or, in a canton the united arms of Great Britain. Crest: A ship of three masts in full sail. Supporters, dexter: The emblematic figure of Britannia, with the shield, lance etc all proper; sinister: A fisherman completely clothed, with cap boots fishing net etc and in his hand a string of fish, all proper.[61]
I mean.
The artifact you link shows a map of the Americas in which California is an island and either Tierra Del Fuego is huge or the bottom of Argentina is an island and the northwest of the continent trails off into nothing, and Florida is sort of a stubby nub (other maps from this period show a more accurate Florida, so this might be a small-size-of-the-object problem).
They had a decent view onto the east coast of the Americas, but after that things got quite inaccurate. It's like... I don't know what anyone's expectations are, but it certainly isn't the perfect world map that's shown in the main image of Wikipedia's article.
Gell Mann amnesia effect!
WTF is this? Another thing wikipedia invents out of thin air.
> The company value represented a decent fraction of the national gdp at the time.
The company's market-cap was almost 3 times national GDP
We should stop comparing incomparable things, like companies market cap (measured in dollars or pounds) to national GDPs (measured in dollars or pounds per year,) unless we want to reach outrageous but incorrect conclusions.
It's a dubious comparison but not because of "per year". The comparison is implicitly to one year's worth of GDP, which is a currency amount.
It's dubious because whereas a year's worth of GDP has some claim to actually being the value of something (with many caveats but it's engineered to behave like that as much as possible), market cap isn't. It's the amount all the shares would cost if someone bought them all in one go for the price some shares were most recently purchased for, which would never happen.
Er... happens all the time sir. Happened to me (Hashicorp shares)
Market cap being rediculous is the hallmark of a bubble.
Hashicorp was bought for $35 per share at a time when it was trading a little above $25. Not saying crazy market caps aren't a sign of a bubble (not sure how you'd read that in my comment), just that market cap is not the value of the company.
Variation in price doesn't prove that the market cap is not (a good estimate of) the company value for highly liquid stocks.
Value is subjective. Stock prices measure peoples perception of the value. Your thesis that it is incorrect can only come from 2 places (I think)
1. Dumb money - the market cant see that XYZ is overvalued or undervalued. My rebutal there is nonetheless XYZ has been valued by a conpletely open continuous auction that people are not restricted to participate in.
2. The parts are less than their sum. This may be somewhat true... total control over a company may be more (or less) valuable than splitting. But I dont think it is order of magnitude. And if it is, it is because the value to you isnt the value to me (the value of RAM to a gamer < value of RAM to OpenAI).
Well, based on that last share purchase, we have incontrovertible proof that there was indeed one person in history who thought it was worth 3x GPD.
And the fact that in the entire BSSTC shareholder universe, there wasn't any noticeable volume for a sell, or a registered sell limit, at a lower value leading up to the last peak.
That must have been a rough trade, but someone got something out at the last moment.
A couple of points:
1. No, we don't have proof that there was one person who thought it was worth 3x GDP. We have proof that there was one person who thought a 0.001% share of the company was worth 0.003% of GDP or whatever. They could think it was worth that much for plenty of reasons; maybe they thought the share price would grow for a bit more before collapsing so that they could make some profit, maybe they invested in order to just be an investor and have a say in investor meetings or however things worked back then. Maybe it was a status thing.
2. Why are we using the opinion of one random person to determine the value of a company
To add to this, this type of thing happens all the time in crypto.
A coin will release 1/1000000000th of it's eventual supply, have some trades at 10c and then claim the value of the entire supply as the headline value.
It's obviously dumb.
> “I have found out what economics is; it is the science of confusing stocks with flows”
- Michael Kalecki
Comparing stocks to Flo's is perfectly meaningful. It's often done when comparing, for instance, the price of a house to one's yearly salary.
That makes no sense. House price/salary is used to compare payment periods or affordability. The context is important. Share value and GDP are totally different things and there is no direct relationship.
What simple method would you use instead? It is effective to have such comparisons for quick communication of the relative size of things.
It is effective but wrong. It is the same as to compare a river flow to dam volume. There is a relation, but does not mean anything.
A better comparison is to compare somebody's net value to the total housing value of a city. For instance comparing Musk's net value to the total value of all the houses/apartments in New York. Or the the value of the gold in Fort Knox.
How do people value companies (i.e. decide what the market cap should be)? Most often by comparing it to a corresponding per year number or two.
What’s the issue? If a market cap is three times an annual GDP, it means it would take three years of that economy’s entire production to purchase the company. It’s obviously a very fuzzy measure for various reasons, but it can give some idea of the market value of a company.
Oh wow, thanks. I misremembered, thought it was like 1/3 and didn't have time to look it up again.
For people who are unclear what Palantir actually does
Palantir is a tech platform that consumes data from their clients in return for providing high level data-driven insights. They assign FDEs (or consultants) to really learn the details of a customers data. Foundry allows them to get single pane view of the data in an org and they actually have both the tech and engineering skills to do the dirty data cleaning jobs.
For an extravagant fee, you give them your data, they clean it for you, and then those same FDEs can tell you interesting things that you should have known, had you actually done proper data architecture in the first place.
Does it add value, yes. Is the value worth the fee? Like snowflake, they throw some very very good parties.
The reason Palantir is valued as highly is because it is a Peter Thiel company, and he also owns the Vice President of the United States.
The thesis is "The current US administration is making enormous profits from their powerful positions. Whether of not I agree with this, I want some of that money too".
I'm not sure why this isn't mentioned because it's the number one factor driving the price.
its not mentioned because its boring political pablum
One illuminating factor as to why: watch their CEO Alex Carp overdose on cocaine while on stage in public.
https://bsky.app/profile/ronfilipkowski.bsky.social/post/3m7...
And then you have Thiel going crazy about biblical end times prophecies involving the anti-christ and I'm starting to wonder wtf they're doing at Palantir.
The other founder is in the news for calling for public hangings to be brought back.
My shitty country bought it anyway.
It seems like he is just trying to make a point with his moving around about being "annoying" - which is what he is talking about.
He appears to be making that point for the entirety of the interview: https://www.youtube.com/watch?v=8i-ys9faa74
I'm pretty sure his surname is a deliberate typo.
you haven't seen the real ontology yet
https://x.com/justintrimble/status/1997733647223603259?s=20
https://xcancel.com/justintrimble/status/1997733647223603259
This is the first time I've seen an ad presented as a bullet point in a list of otherwise-salient bullet points. Nefarious.
Yeah, all credibility was immediately lost when I read that. I just closed the tab and put it away as ad piece.
Try using Safari’s Reader view on this ad-riddled page, and you’re greeted with all-text content that’s… 100% ad!
It is pretty clearly labelled with "(Sponsor)". Couldn't be less obtrusive, either. I think on balance it's at least neutral compared with the other nonsense available to both the publisher and advertiser these days.
Yeah that's insane.
Unreadable...
Putting this here for visibility, but go upvote the people who actually did the sleuthwork: https://news.ycombinator.com/item?id=46190013
It seems like the financial outlet made a maths error and then was amazed by the absurdity of the incorrect result they got.
Specifically their whole outrage is based on having accidentally calculated what would need to happen for Palantir's revenues to grow 1500x as opposed to 15x. The corrected statistic would be the Palantir requires 11.4% revenue growth year over year not 35%.
TSLA has to be in the mix with a PE over 300.
I don't know. People judge Tesla and other companies on their links with the Trump cult and trade tariffs.
And to be honest car manufacturers have always had very deep links with politics.
Most car manufacturers have a PE around 10.
TSLA is over 300. And sales of EVs are actually falling under Trump. And tariffs are not helping.
Nothing about the valuation of TSLA or PLTR make any sense.
I think it would blow their minds that Tesla isn't just an EV manufacturer.
90% of TSLA income is from EV sales. Everything else is just Musk's grandiose and increasingly absurd predictions which have a long history of falling short of reality.
For example:
https://electrek.co/2025/07/23/elon-musk-with-straight-face-...
> Everything else is just Musk's grandiose..
Meanwhile the story of Jensen and Musk continued onward with custom chips to support FSDv1, in which J's personal delivery of the DGX1 to OpenAI served as the catalyst to the relationship, iirc.
The problem for nvidia is ... where do you go from here and still have spectacular performance improvements?
Nvidia got extremely lucky again and again and again, and what specifically did it is that right in time non-Nvidia researchers learned to train on smaller floating point bit lengths, which Nvidia raced to support. And great, well, done! A list of ironies though ... for example it's Google Deepmind that made the Turing generation of cards viable for nvidia. However, the new floating point formats train has arrived at it's last station, the NXFP4 station. There is no FP3 and no FP2 to go to. Is there a new train to get on? I'm not aware of one.
Nvidia's argument is "Blackwell easily doubles Ada performance!" ... but that is deceptive. The actual improvement is that Blackwell NXFP4 (4-bits) is more than double Ada FP8 (8-bit) performance in ops. That's the train that's arrived at its last station. Go back further and the same is true, just with larger and larger FP formats, starting at FP32 (single precision). Aside from a small FP64 detour, and a few "oopses" in some of the format they chose turning out useless or unstable, all quickly abandoned that's the story of nvidia in ML.
Comparing, for example, FP32 you don't see big improvements: e.g. 4090: 83 FP32 TFLOPS, 5090: 104 FP32 TFLOPS. Given the power requirements involved that's actually a regression. If you're stuck at 8 bits, nvidia's story breaks down and Ada cards beat Blackwell cards in performance per watt: 4090: 5.44 Watt/FP32 TFLOP, 5090: 5.5 Watt/FP32 TFLOP. Or, FP8, same story: 4090 is 0.681 Watt/FP8 TFLOP, 5090 is 0.686 Watt/FP8 TFLOP. Now effectively the new memory still buys some improvement but not much.
Will the next generation after Blackwell, with the same floating format as the previous generation be a 10% improvement and subject to further diminishing returns and stuck there until ... well, until we find something better than silicon? I should point out 10% is generous, because for FP8, Blackwell is actually not an improvement at all over Ada, on a per-watt basis for equivalent floating point lengths.
Plus Blackwell is ahead of the competition ... but only 1 generation. If nvidia doesn't get on a new train, the next generation of AMD cards will match the current nvidia generation. Then the next TPU generation will match nvidia.
Isn't a bunch of it just bitcoin accounting tricks that were 9% of Tesla's profit in 2024?
https://fortune.com/2025/01/30/tesla-profits-bitcoin-crypto-...
Oh yeah, and government subsidies (making up 38% of profits in 2024)
https://news.sky.com/story/elons-playing-a-very-dangerous-ga...
Yes. Most companies play these financial games to some extent.
I lumped government subsidies in with EV sales since they are related. Trump wiped these out.
Robotaxi and robots are the fantasy category. They are not currently income producers and may not be for years to come. His robot demos have been widely panned as fake.
https://mashable.com/article/fake-tech-demos-tesla-musk
Ah right, I forgot about the robots that definitely aren’t remote controlled.
How’s that Full Self Driving coming along? Does it still disable itself right before a crash so the driver has liability?
For the sake of democracy and freedom of human expression, I hope that their software is a bloated, bureaucratic mess and can’t power the efficient dystopia which Alex Karp wants.
To me it would depend a lot on how US/EU relations continue to grow/break. If digital sovereignty ends up putting Palantir in a position where they won't be able to seel their intelligence tools to European secret police organisations, then it's probably overvalued. If relationships work out fine, then they'll basically be the Microsoft of mass survailance, which may make them overvalued but surely not the most overvalued company that ever existed?
Once you’re embedded in govt, the ‘OS’ for government, then you can raise charges in order to keep the hoards at bay. Hoards that are angry at cuts to public services made in order to finance Palantir profits…
What 4/5yr term govt can turn off or replace their OS?
I do not think the EU is critical. Palantir's potential market is global, so one region is not that important to them. They also sell to the private sector and this will increase, and EU countries will have different policies so losing a few EU governments business would not be that big a deal.
Right now 73% or their sales are to the US, and the next biggest source of revenue is the UK, far behind at 10%. The whole of the rest of the world, including the EU, adds up to 17%.
Is their market really global though? If it wasn't for political reasons I suspect everyone would be buying Chinese mass surveilance tools. Traditional US allies won't do that, at least not yet, but why would South America trust US over China?
In that case Palantir has to compete with China.
And the US has a pretty bad neo colonialism and CIA rep. If I was the president of Mexico I would not trust anything from up North right now.
Would you trust China? it has an incomparably worse record than the US.
A lot of the biggest potential markets are in Asia, and almost all Asia countries fear China, and to many of them it is their main potential military threat.
IMO the best interests of either of the two countries of which i am a citizen would be to be less dependant on the US, however the US is nothing like the threat China presents.
Edit: to add, they are competing with whatever equivalent product China has to sell anyway.
For the Mexico example, China isn't the neighboring USA whose president out of nowhere started talking about taking over it's only other neighbor, or government overthrow in Venezuela. Or the recent attempts to sway elections in other central American countries, pardoning of convicted drug dealing ex-Presidents. Attacks on Columbia's government. Sweetheart deals with central American dictators to illegally house immigration deporties in sub-human conditions with zero due process.
Can someone actually explain what the magic sauce is in what Palantir does? I understand it pulls data together?
Magic sauce is absolute lack of moral values.
But I mean in terms of technical capability? How is it different than chucking everything in Postgres or a Big Data lake or something?
It isn't. It's just that they're doing it where others aren't. They're a child of Total Information Awareness that has never realized or just doesn't care about the societal poison pill inherent to it.
lol and you know that based on what information ?
based on what their CEO is saying...he repeats like a broken record; "we do what other companies won't because of ethics and morals"
The magic sauce in terms of bubbleability is that the product is all secret and investors can't form any natural opinion beyond what's published in company performance publications. Because they aren't allowed to know anything beyond that. It's just blind trust, "governments have bought before, they might continue buying, perhaps buy more". Reminds me a bit of the psychological mechanisms that make a con work, might be something to it.
There's a few things. Let's start with the core of what they say is their value. They have forward deployed engineers - this a totally new, previously unknown innovation - who go to a company, understand their needs and build data processing tools to give them insights. Then, they generalize these tools so that they can essentially sell them as SAAS software, giving them SAAS-type economics.
What other people say is their secret sauce, is they do consulting work for the government (a forward deployed engineer is just a consultant) and they make incredible margins because their senior management and early investors have connections to the government which gets them exclusive access to incredibly juicy contracts. As these contracts paid off they leant heavily into the social media meme stock trend so their CEO spends time talking like a psychopath and doing various non-economic things like spending huge amounts of money running adverts about how they're going to use AI to unleash Americas workers (America's workers aren't able to buy Palantir software or services, but they can buy it's stock).
OK. The first part sounds a bit like an innovation.
I was kind of expecting someone to say it had EG really sophisticated ETL tools that can normalise loads of different data or can query across disparate data sources or something.
The first part sounds like a basic ERP implementation. Only instead of leaning on your in-house domain experts who have years of experience/relevant knowledge/know the relevant caveats, you pay consulting rates to train up new domain experts who don't understand/know the caveats and who will charge you consulting rates to gain access to the results of the training you overpaid for.
'But they cleaned the data up'. That data was also cleaned up during all the last major system updates. And during the implementation of those systems. And the implementation of the systems before that.
I mean they do advertise themselves as an ERP system so that part tracks.
> What other people say is their secret sauce, is they do consulting work for the government
Also for the IDF
How is this "forward deployed engineers" model different from that of basically any tech consulting firm?
It isn't.
Papypal mafia, thiel, musk, trump, money, oligarchy, conflicts of interests, military–industrial complex, 3 letters agencies, mass surveillance, &c.
It's kind of a conspiracy theory bingo at that point, it's probably somehow even worse than we can imagine
They think there is money on the table from maximizing anti-ethical behavior. This misinterprets capitalism's current state as something with ethics.
Michael Burry shares his own thoughts on the subject, toward the end
https://pca.st/episode/17966c7b-a56e-46f7-ab2d-03d86dbad650
The market can stay irrational longer than you can stay solvent.
Michael Burry famously did quite well betting against that advice.
Over 10 years he did even worse than Cathie Wood.
He's ahead of Buffett but WB has massive AUM and MB should be running circles around him.
So yeah... he had one movie worthy trade, but that doesn't make it a sound strategy.
It is like there is this online bot that scans for any mention of an asset being overpriced and then it automatically post this Keynes quote.
Palantir is only the 15th most overvalued:
https://www.macrotrends.net/stocks/stock-screener
P/E isn't the only way to value a company. A company can be unprofitable, and therefore not have a P/E, and still be worth something. A quickly growing startup that is losing money on paper but has free positive cash flow is one example.
Well, Palantir is the only really large company in the P/E top section. You could argue smaller companies have more upside and could thus justify their valuation easier
Today the fast growing US stocks are those that discount the further concentration of wealth to the current private-government elite. They believe that in the coming years they will flourish at the expense of the middle/lower class.
>Palantir would need to grow its revenues roughly 15-fold (yes, 1,500%) over the next quarter century, implying sustained annual revenue growth in the 35% range over this time frame.
That's actually a pretty reasonnable ask for a tech company, if you believe Palantir can grow to the level of FAANG.
Sustained growth of 35% p/y over a _quarter of a century_ is not a 'reasonable' ask, tech company or no tech company. Virtually no company has ever done that.
15x over 25 years is a reasonable ask for probably half of tech companies because you don't need 35%/y do accomplish it, you only need 11.4%.
Their math is just wrong. If they need to grow 15x over 25 years, they only need a modest 11.4% growth per year.
15^(1/35) = 1.11
1500^(1/35) = 1.34
Did the author maybe put a few extra zeros in their calculator when figuring out the annual revenue growth which would equal 15x over a 25 year span?
Dumb analysis. Assumes that profit margin stays the same as revenue grows, which is particularly untrue for software.
One of the concerns with the company though is are they really “software.” Many cite experiences with the where they deploy armies of people doing things to clean up data and other hands on labor for the “software” to work just “software” where you buy a license and off you go.
Is this a one time onboarding thing or is it ongoing human involvement?
Palantir has terrific sales and lobbyists. I'd never bet on them, nor against them.
if you had the money to spare (and don't mind losing it), would you still not bet on them then?
I have very specifically invested in Defense funds that exclude Palantir both because I think it's overpriced and for ethical reasons.
Even if you have money you don't mind losing, there's plenty of options to choose from on financial markets and Palantir is just one of them.
So the question is still why Palantir and not X/Y/Z?
What an ad ridden mess. Even one of the summary bullets is an affiliate link.
Can somebody explain how the software actually adds value? How is it really that amazing? Or is it just a buzz word?
There was a great blog post by an early employee on here a while back that did a deep dive. It's a good read: https://nabeelqu.substack.com/p/reflections-on-palantir
No. Thats the point. There nothing groundbreaking except their branding and lobby. Every other enterprise sofware company has data analytics platforms similar to Palantir - including the big microsoft / google. Palantir might even be best by a big margin but none of that matters. Nothing would justify this valuation. But what matters is that govs around the world will buy Palantir to show their alliance to US/Trump/Thiel. European far right politicians will happily push Palantir everywhere.
Like Tesla, Palantir is an ideological investment for many, and likely why it's so over valued. Retail investors buy these stocks because they're buying into a world view. Tesla's is obvious (and it's not about green investments). Palantir's is a little less so, but they have built up an image that's right wing (but mostly not culture war), military-coded, and layered on American exceptionalism. They sell to the CIA, they sell to ICE, they help private healthcare companies milk their customers. This is all part of a world view that many people are in favour of, and Palantir represents it well. Andruil will likely do well for similar reasons once they go public.
This is not really to fault them. I see the company as far more mature than Tesla. They've cultivated a brand that works well for them. This is also not intended to be a criticism of Palantir investors, there are many reasons to invest, and even if you subscribe to the above that is of course your right! Many people invest for ideological reasons, myself included, and that's fine.
> Many people invest for ideological reasons, myself included, and that's fine.
It's fine if you don't mind losing your shirt.
For us normal people, we invest to try and grow the money we've saved for the future (in my case, for my retirement), so investing in a company whose stock is insanely overvalued is a great way to blow up my pension pot.
I don't understand the rationale behind "investing for ideological reasons", can you explain it?
Investment for ideological reasons is basically "I like what this company is doing and want to see it succeed/want to make my returns on investment via it", and that investment helps it succeed, to at least some degree. You also get the opposite: "I don't like what this company is doing and don't want it to succeed/don't want to make money from this activity, so I won't invest in it". ESG investments were largely built up from this (much as they generally turned into box-ticking exercises as opposed to a useful distinction). Of course this is generally going to make less money than "I just want to make money/predict winners in the marketplace", but that's not everything people want to do with their money.
The difference being not investing in something for ideological reasons won’t jeopardise your assets; whereas YOLOing your 401k on Palantir because “AI surveillance capitalism is dope lol” is potentially risking significant portions of your future.
Can’t people just retweet Karp on X if they want to “support Palantir ideologically”?
This is GME all over again.
Depends how you do it (I wouldn't put everything in one company no matter how much I liked them), but having money to invest does represent some power over the world so I understand people exercising it (though not why someone would choose Palantir).
GME is not the same thing: while I think there are some true believers in the company itself (and this is part of what drove the initial interest in it), for the most part interest in modern meme stocks is driven by the polar opposite: a kind of financial nihilism that believes hype and gambling on popularity is the only thing left in the public stock market, and they're basically just trying to ride the wave of a crowdsourced pump-and-dump and fleece others via selling to greater fools (and/or creating a cult around a mythical short-squeeze that is reaching ever more fanciful heights).
What I find the strangest is Palantir's market cap and long term valuation show that this is not just retail at the margins - it's significant amounts of money buying up the stock despite its sky high valuation. I don't know if it can be sufficiently explained by "individual actors investing ideologically" or not.
Just another day of the stock market being irrational though, I guess!
> Many people invest for ideological reasons, myself included, and that's fine.
Sure, but I'm under the assumption that many people, yourself included, are conflating "buying some company stocks" with "investing".
To first order approximation, it does not benefit a company that you trade stocks on the market once they have been emitted, any more than it benefits GM if you opt to buy a second hand chevrolet.
So the support is symbolic only.
If they keep their shit together and ride the surveillance wave then I don't see what should keep them from increasing 15fold by 2050. Their technology is not at all dependent on AGI, which is why I don't understand why they are always thrown in together with OpenAI et al.
The cope and envy in the comments here are among the worst I've read on Hacker News. The anti-tech hysteria has even made it here. Sad days.
they're gonna build the american social-discredit system for many filthy dollars
More than Tesla? good luck
Palantir is a shit show
it may well be terrible software, but Palantir are effectively an arm of the new Trumpian state, so guaranteed access (and ownership?) to the largest public sector projects of the largest economy in the world. It's got a good chance of becoming the Samsung of the US, so any value is not an overvalue
> any value is not an overvalue
Ridiculous statement. If you're actually investing your money in a company then you need to run the numbers, it's the most basic kind of due diligence you should do.
Let's reconsider what you think the word value means
> an arm of the new Trumpian state, so guaranteed access (and ownership?) to the largest public sector projects of the largest economy in the world
That puts them in a precarious situation when the political winds change
> That puts them in a precarious situation when the political winds change
I thought Alex Karp in a relatively recent interview said he didn't give a hoot?
So get in now, but get out soon before the Trumpian reign is over.
[dead]
As long as the AI hopium flies