Historically stocks that had a good run then tended to underperform:
> […] Since 1926, the median ten-year return on individual U.S. stocks relative to the broad equity market is –7.9%, underperforming by 0.82% per year. For stocks that have been among the top 20% performers over the previous five years, the median ten-year market-adjusted return falls to –17.8%, underperforming by 1.94% per year. Since the end of World War II, the median ten-year market-adjusted return of recent winners has been negative for 93% of the time. The case for diversifying concentrated positions in individual stocks, particularly in recent market winners, is even stronger than most investors realize.
It's interesting, but the usual disclaimers apply with factor research
- Does it replicate internationally?
- Is it explained by another phenomenon such as the beta anomaly or small cap premium. Implication: large caps with high beta are already known to underperform, so this this isn't a new regularity.
> Historically stocks that had a good run then tended to underperform
This is more of a mathematical axiom than a financial effect, because you're defining "underperform/overperform" with respect to an average that contains them.
>> Historically stocks that had a good run then tended to underperform
>because you're defining "underperform/overperform" with respect to an average that contains them.
Why is this true? For instance, if you're comparing the GDP growth of countries in the G7, why is it that one country (eg. US) can't consistently overperform year after year?
Shouldn't you look at the YoY change instead, to compare to stock returns ? Otherwise that's like comparing market cap, and then it is obvious that a big company tends to stay big.
>Shouldn't you look at the YoY change instead, to compare to stock returns ?
This might work for the G7 case[1], but not the US vs DRC case, where it's an obvious case of sloping up vs sloping down. Granted, the case is contrived, but the original claim was that it was an "mathematical axiom", so it should still hold.
[1] though even in the G7 sample, you can find counterexamples. If you switch to "relative growth" you can clearly see that italy has lagging since the mid 2000s, with no accompanying faster-than-average growth to make up for it. If the claim is that "Historically stocks that had a good run then tended to underperform", then surely the opposite must also hold?
> This is more of a mathematical axiom than a financial effect, because you're defining "underperform/overperform" with respect to an average that contains them.
Most stocks suck:
> We study long-run shareholder outcomes for over 64,000 global common stocks during the January 1990 to December 2020 period. We document that the majority, 55.2% of U.S. stocks and 57.4% of non-U.S. stocks, underperform one-month U.S. Treasury bills in terms of compound returns over the full sample. Focusing on aggregate shareholder outcomes, we find that the top-performing 2.4% of firms account for all of the $US 75.7 trillion in net global stock market wealth creation from 1990 to December 2020. Outside the US, 1.41% of firms account for the $US 30.7 trillion in net wealth creation.
> Four out of every seven common stocks that have appeared in the CRSP database since 1926 have lifetime buy-and-hold returns less than one-month Treasuries. When stated in terms of lifetime dollar wealth creation, the best-performing four percent of listed companies explain the net gain for the entire U.S. stock market since 1926, as other stocks collectively matched Treasury bills. These results highlight the important role of positive skewness in the distribution of individual stock returns, attributable both to skewness in monthly returns and to the effects of compounding. The results help to explain why poorly-diversified active strategies most often underperform market averages.
And this certainly can have a financial effect on your finances: having the "wrong" stocks in your portfolio (i.e., most of them) and not have the "correct" ones will mean a (e.g.) comfortable retirement or not.
> I look forward to your weather report too: "It's always sunny outside until one day it starts raining. Every time."
I once ran across the comment that if you simply predict tomorrow's weather will be the same as today's you'd be correct 80% of the time. Not sure how true that is (can't find the source).
Not at all, if someone tells me that "This stock is historically likely to regress to and beyond the mean," it's information I can use to evaluate my risk tolerance. Just because a piece of information doesn't let you time the market like a psychic doesn't make it worthless, it's just not what you were looking for.
Productive workers in society need to learn new things all the time. I can't think of any career that hasn't changed in my life. I recall a garbage man (sexism probably wasn't required even then, but I never recall females) hanging off the back of the truck while the driver drove to the next house - the driver today needs to know how to operate the arm on the truck that lifts my can. Fast food used to be cooked within 10 minutes of when it was thrown, now they obviously are keeping things warm for a lot longer.
> I mean these stocks have been performers for decades. If you posted this 10 years ago you'd look really wrong.
And Japan performed ridiculously well for over decades and then stagnated for decades after that, but it averaged out between the two periods:
> Ben Carlson: It's just a really long mean reversion. You got like 22% per year from 1970 to 1989 in Japan. Small caps in Japan did 30% per year for two decades.
> It's insane. The returns almost had to be poor after that. If you put them together, the boom with the bust, it's like almost 9% per year.
> It's kind of crazy. Over 50 years, the long-term worked. It's just that over that 20 or 30-year period, it didn't work so well.
Annualized 9% per year is pretty good: the S&P 500 has average 10% since 1957 (70 years). Is there anything preventing US equities from doing the same thing: great performance from 2010 until now, and then 10+ years of stagnation starting (theoretically) tomorrow. If you look at 2000s S&P 500 you got zero returns, and the only thing that would have saved a US domestic (only) investor was having a bond allocation:
This is why diversification is important. People talk about "US stocks" doing well, but have US industrials done better than non-US industrials? US finances or energy done better than non-US? Or are "US stocks" doing better simply because tech stocks specifically have done better? Perhaps a US allocation is really a tech sector play:
I am invested in some of the companies that are downstream of the capital expenditures of Big Tech (e.g., COHR), so I have nothing to complain about.
I am really struggling to see what's the investment thesis behind Google valuation increasing 2x in response to AI, though. Assuming no magical AGI singularity, by the end of the day, they're still selling the same services, but the services have gotten more expensive for them to provide. Everyone was already using Google Search, but now, provisioning AI summaries on top of requires more compute. Everyone was already using Google Docs and Meet, but now, AI features cost Google more. Etc, etc.
The only place where they stand to make money is selling AI compute to enterprises. But with the current supply-chain challenges, the margins there are probably getting thinner.
`I am really struggling to see what's the investment thesis behind Google valuation increasing 2x in response to AI`
Google is basically Nvdia (TPUs), Tesla (Waymo Self-Driving), Hyperscaler, Netflix (YouTube) and a massive VC (Anthropic, Databricks, SpaceX, etc.) all rolled into one.
Their valuation isn't really a 2x'ing so much as a reversion from halving.
It's frequently said that vonglomerates usually carry through as their valuation, the multiplier of the lowest holding. If that's the case, Alphabet is worth more broken up. Similarly that makes their current valuation (and nay any valuation) theoretically too low.
>Assuming no magical AGI singularity, by the end of the day, they're still selling the same services, but the services have gotten more expensive for them to provide.
Well, they're hoping to sell new services on expensive $200/month subscriptions.
The hope is that agents have more value than traditional software, because they do the work for you instead of just enabling you to do the work.
Google has a compelling story for many AI scenarios: it has lots of outs. It's the only frontier lab for which that's true. A massive bubble bursting wouldn't be existential for Google; it would be quite painful, but survivable, and even offers some potential upside (picking up assets and researchers from the wrecked, mangled corpses of other frontier labs on the cheap).
Google is a good bet because if AI continues to boom, they're in a good position (frontier lab, vertically integrated). If the bubble bursts and the frontier labs fail, they might do even better.
The market sets prices, and they are set based on multiple things. One of those is fundamentals. Consider the value of assets, whether tangible or intellectual property, human resources, binding contracts, etc. that add up to reasonable revenue forecasts and so forth.
And the other aspect of prices is based on conjecture, speculation, meme-joiners, believing hype, and in some cases, fraud.
The secret sauce is always going to be the one who can figure out, between the two factors going into price, what's right, and when.
BUT... just saying that all stock market pricing is based on unreliable factors? That's not a useful, actionable statement. You can certainly stay out of investing in that market, but is that going to be your best course of action?
And a market cap that exceeded 4.5 trillion dollars. I think we can all agree that Google is a legit business. But they were doing fine in 2023 too and had solid y/y revenue growth for much of their history. Their market cap tripled since 2023 and doubled since mid-2025. So, it's clearly more than "they continue to be a profitable company that grows around 15% y/y in real terms".
That's trailing and mighty impressive. However when you decided to quote Q1 earnings. $62bn of Q1 earnings ends up at only $10bn. The LTM numbers are less relevant.
The discussion here is whether Google can make money in the AI world. They obviously can. Whether they turn around and spend that money on capex to try and increase future earnings is irrelevant.
I like that the invisible hand of market is slapping the Mag-7 for capex which is the only way to discipline them. Investors are waking up to say: hey, you are spending all your profits on data-centers, where is the return for me ? But, it surprises me that there are vast pools of capital which we collectively call the "market" that makes these calculations, or maybe a simpler causal explanation is the missing stock repurchase bid. At some point, one of the hyperscalers (msft ?) will break from the pack and announce reductions to capex and increase stock repurchases to stem the decline.
You can simultaneously believe that the hyperscalers are becoming capital-intensive long-term in a way that’s bad for their profits and that as a result will be raising the COGS of Apple’s business in a way that also hurts profits.
My perception is that the market lumps companies together into sectors whether they have similar business models or not. When one is punished, other associated firms tend to be, too. You see this in any industry (not as a guaranteed rule, but in general).
Could outperformance of largest cap companies be partially explained by dividends paid by smaller caps? Reasoning behind this:
* Index investing raises in popularity, with index funds that automatically reinvest dividends being often preferred due to their tax efficiency.
* Large caps prefer to repurchase stocks, stock repurchases contribute fully towards a given company share price increase.
* Smaller caps still pay dividends, these dividends are then reinvested by index funds and the reinvestment is weighted by capitalization, so large caps share price benefits more from repurchases done with dividend cash paid by smaller caps. When dividend is paid, share price of a company that paid it is reduced, which further widens the performance gap between large and smaller caps.
>* Large caps prefer to repurchase stocks, stock repurchases contribute fully towards a given company share price increase.
That doesn't work because indices are typically market cap weighted. A stock buyback might increase nominal stock prices, but not the market cap, otherwise it'd be a free money machine.
While I agree with other comments indicating that the headline is drawing on a small period of data, other data in the deck is pretty compelling.
Page 25 "The number of data centers in the US" gives an interesting insight as to the magnitude of the data center boom. 60% more data centers are being planned or are under construction. This might actually be underselling it in dollar amount, as I believe the average data center size under construction is larger than the average data center already constructed.
Page 27 "Cyclically adjusted P/E ratio near all-time highs" is certainly concerning and points to a near term correction.
Right, but they're smart enough to know how to manipulate the market short term with such "research" and then take advantage of it, all without breaking the law.
I'm astoundingly unimpressed by the quality of this slide deck. There's no analysis except that crammed into slide titles, like it's designed to bombard a room full of analysts with so many graphs that they shut off their critical thinking. Could a freshman business student not make this? Could a freshman business student with an LLM not make something more convincing than this?
I agree with the headline, but this really feels like analysis-slop. It's only remarkable in terms of who is publicizing it.
I assumed it was accompanied by an oral presentation or another report. I didn't look around enough to see if there is a transcript or anything though.
What TSLA lacks in cashflow, they more than make-up in TAM. Make-up as in create out of thin air. We'll have self driving cars in a few months, robots, semi-trucks, etc., etc.
Elon is missing the bigger TAM here. Once they've created a time machine, they can not only sell all these things backwards and forwards in time, they can also start making industrialization come way earlier, which will lead to exponentially turbo charged growth. Plus importing future designs into the now.
It makes sense to focus all effort into building a time machine.
On slide 6, they list the Mag7 stocks (not defined in a foreword), but on slide 8 they list the free cash flow (FCF) of a somewhat different set of companies. Why not stick to the Mag7 FCF only? It muddies the waters.
True. This slide also struck me the most, but for the data it shows. Free cash flow for all the hyperscalers basically evaporated in the last couple of months, with Oracle in the minus. What does this mean?
Yes but the Mag 7 is the main focus, suddenly switching to a different set of companies is comparing apples to oranges. That slide belongs in an annex.
Seems like we're seeing margin get eaten up by companies upstream of the buildout, and the costs have not been fully passed downstream yet. Eg most consumers still get free/cheap AI.
It'll take a few years to see what happens at scale when prices go up and purchasing behavior changes. Hardware, services etc all downstream of the buildout and supply constraints.
Markets are obviously, rationally, not happy about the overspending.
But what gives me pause is that a some of the mag 7 (think Meta) could change their mind on AI build-out tomorrow, and 1-year from now have the same amazing free cash flow they always did.
I don't think most people quite grasp the sheer size of the debt obligations Meta and specially Oracle have gotten into. Meta has done some clever tricks to keep it off the main books, but it IS there. If no AGI, its gonna drag them down for a decade.
I tend to agree. I suspect if we look back a few years from now that there was some overspending but I still believe the risk of not investing and missing out is greater than the current historical trend of capex spend. I have not looked recently but has demand of compute started to slow down. It was just a few months ago when companies like Anthropic were throttling users because there was not enough of it.
What has been the best way to determine return on the AI specific capex for hyperscalers?
I would naively expect Microsoft’s to be the highest since they are probably mainly just selling access to their capex through cloud since they aren’t seriously pursuing frontier AI, I’d imagine Google to be in the middle (selling TPUs, general cloud GPUs, Gemini, revenue lift on ads from better AI) but also spending heavily on infra to compete with OAI/ Anthropic, and then Meta to be on the low end since they are likely getting serious revenue lift from AI but not monetizing their models by API.
yeah i was interested as well. -$2.5B from $11B last year. However the analysts see it recovering quickly. If you put all the Trainium(N) chip investments aside as being all AI risk, they still have some good pipeline with the satellite internet stuff (Delta, JetBlue, Vodaphone).
I think page 4 is a little disingenuous, I'm pretty sure you could pick lots of windows and show the mag 7 deviating negatively, but then later trends positive. I believe this whole thing will pop, but I'm not quite convinced it is just yet.
Page 8 for Oracle's free cash flow trending negative is really quite impactful. The Bloomberg AI bubble diagram [1] shows how this could really blow up. If Oracle falls they could take the whole market with them. We're just waiting on the mag 7 or any closely linked companies to fail to raise investment.
Page 16 really outlines how insane these evaluations are. I think most countries see it, hence aggressive selloffs of US bonds [2]. But everybody is just too insanely heavily exposed to it all now, it's going to wipe out everything. It's going to be a very awful time when heavily debt strapped countries can't issue debt anymore.
I think what we're going to see is some insane moves to keep these companies afloat longer in some desperate attempt to delay the pop, which will just make a bigger bubble. I could see Nvidia for example issuing bonds in excess of $100bn soon if the market has appetite for it [3].
as soon as you start calling a group of stock tickers the “magnificent 7” they’re destined to underperform, as that’s a feelings based assessment and many investors will continue to buy the feelings long past the value being fair.
“Do they make money? I don’t know but I know they’re magnificent!”
> The once high-flying "Magnificent Seven" are looking more like the Dreadful Seven.
> The why: Wall Street is growing increasingly impatient with Big Tech's astronomical capital expenditures on artificial intelligence, projected to balloon 70% to exceed $700 billion this year.
Completely missing the point of not adding Anthropic, OpenAI, SpaceX.
Mag7 is just an arbitrary list of companies that were coined at a specific time with no rigor.
If you must do rigorous analysis, then just like S&P 500 you need to add drop BigTech / High Growth companies continuously to this Mag7 (or Mag10) and then do analysis
This feels like telecom "massive demand, bad returns".
If a good enough model can be swapped in every few months, the value moves away from the model and toward cheap inference. That is great for users, but not always great for returns on huge capex.
The only mag 7 that had a shooter chance with models is Google rn so doesn’t seem like there being good enough models will be that negative for mag 7. Might even be better since they are the providers for the inference
Not mentioned in the comments but all those companies are taking significant debt and even issuing new shares.
One major part of the investment thesis in these companies were their constant stock buy backs. Now their gargantuan Capex that sees no end but acceleration is back at diluting investors.
I think the companies will keep doing fine, but the financial outlooks are no longer as rosy.
This is probably a hot take and I am by no means a financial expert and this is probably quite wrong. I personally think that attempting to value these companies using the same methodology as the history of all American companies is fundamentally wrong. Sure, some mom and pop small local regional business that overperformed is probably more likely to underperform. But when it comes to big tech companies, these companies are operating a data and capital flywheel that doesn't easily just slow down. I mean, you look at the history of these computer tech companies, especially software companies. They really haven't slowed down. Like, look at Microsoft. It's just been growing from the very beginning, pretty much.
This seems healthy to me. A market where returns are less dependent on seven mega-cap names is probably more stable, not less. If earnings growth broadens out and capital starts flowing back toward quality and free cash flow outside the obvious AI winners, that should reduce concentration risk and make the whole market less fragile.
You can tell these guys know nothing about LLMs or how they’re provided. I love how they show OpenRouter’s graph of token usage as if it speaks for usage across the board. DeepSeek looks like the king because people who use Anthropic and OpenAI use them on either a direct basis or AWS Bedrock …
And the bar chart for token costs, really? As if that’s information? Their sources are the API docs ffs. If they had at least modeled something to estimate token costs that would be interesting, but showing the public prices and calling it research is dumb.
Apollo Group has $1T assets under management, I believe them over most folks on HN. Their argument that the Mag 7 is burning up all of their free cash flow is factually accurate, as the returns are not materializing for the investments being made (ie underperformance).
One is not spending anything out of the ordinary up until this point in comparison to their competition. But the Apple Silicon design group might start spending some money on something practical, something that might begin the typical Apple long-range process of replacing two or three existing companies within their current supply chain, but that’s nothing new over the last 25 years.
It is a bit unfair to lump Apple in with everyone else in the Mag7, as their strategy is arguably reasonable and prudent for the current position of the hype circle. Apple Silicon was a spectacular decision and investment. Everyone else has lost their damn mind.
Not based on the evidence of how they allocate capital. Meta wasted $80B on the Metaverse, for example. Apollo allocates capital as their day job. Mag7 allocates by vibes.
I have an opinion based on the data, yours may differ. Favorite this thread for when the music stops. It has before (1999-2000, 2007-2008), it will again.
I'm objecting to your categorization of the job of running one of the Mag7s as something other than "capital allocation". The CEO of Meta or Amazon or any of these other companies allocates capital all day long. They do it differently than an asset manager like Apollo for sure, and sometimes make worse decisions. But it's in the same category.
Vanguard holds and manages index funds and are, broadly speaking, not active managers. Apollo is a sophisticated, active capital market participant|manager.
I mean, I'd assume this deck was trying to sell something. But the relevant comparison is to stonk-bros on HN LARPing as hedge fund managers, not Vanguard.
Which, to be fair, Vanguard has earned a good deal of trust from me on the passive investment side of the equation, although I don't think it's meaningful to make it linearly proportional to the total size of their managed assets? I'm not even sure how I'd operationalize that in reality.
Having actual skin in the game on getting an answer right is generally a sign of credibility, though.
That said, without context I'm not really drawing any conclusions from this.
> Apollo Group has $1T assets under management, I believe them over most folks on HN
If only that's a measure of if they know how to time the stock market. The real measure is how much they earn not if their sales or marketing department manages to get more customers.
they are building new narritaves with made up BS name like mag 7 this, that. it used to be faang BS but they updated so lower IQ folks gets into new narratives. Marketing dep. and fin. engineering working together. Another older one was ZIRP that, ZIRP this and you see lower and avg. IQ folks talk about it along with lots of bots. ZIRP bots stopped so MAG whatever bots can take over.
Waiting for next game, bets are open, any new names? O A S I S? upcoming IPOs and new names needed so they can exit scam
Historically stocks that had a good run then tended to underperform:
> […] Since 1926, the median ten-year return on individual U.S. stocks relative to the broad equity market is –7.9%, underperforming by 0.82% per year. For stocks that have been among the top 20% performers over the previous five years, the median ten-year market-adjusted return falls to –17.8%, underperforming by 1.94% per year. Since the end of World War II, the median ten-year market-adjusted return of recent winners has been negative for 93% of the time. The case for diversifying concentrated positions in individual stocks, particularly in recent market winners, is even stronger than most investors realize.
* https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4541122
It's interesting, but the usual disclaimers apply with factor research
- Does it replicate internationally?
- Is it explained by another phenomenon such as the beta anomaly or small cap premium. Implication: large caps with high beta are already known to underperform, so this this isn't a new regularity.
> Historically stocks that had a good run then tended to underperform
This is more of a mathematical axiom than a financial effect, because you're defining "underperform/overperform" with respect to an average that contains them.
>> Historically stocks that had a good run then tended to underperform
>because you're defining "underperform/overperform" with respect to an average that contains them.
Why is this true? For instance, if you're comparing the GDP growth of countries in the G7, why is it that one country (eg. US) can't consistently overperform year after year?
https://ourworldindata.org/grapher/gdp-per-capita-worldbank?...
Or if you want make it even more clear, you can construct a index consisting of two countries: a normal country (eg. US) and a basketcase (eg. DRC):
https://ourworldindata.org/grapher/gdp-per-capita-worldbank?...
Shouldn't you look at the YoY change instead, to compare to stock returns ? Otherwise that's like comparing market cap, and then it is obvious that a big company tends to stay big.
>Shouldn't you look at the YoY change instead, to compare to stock returns ?
This might work for the G7 case[1], but not the US vs DRC case, where it's an obvious case of sloping up vs sloping down. Granted, the case is contrived, but the original claim was that it was an "mathematical axiom", so it should still hold.
[1] though even in the G7 sample, you can find counterexamples. If you switch to "relative growth" you can clearly see that italy has lagging since the mid 2000s, with no accompanying faster-than-average growth to make up for it. If the claim is that "Historically stocks that had a good run then tended to underperform", then surely the opposite must also hold?
> This is more of a mathematical axiom than a financial effect, because you're defining "underperform/overperform" with respect to an average that contains them.
Most stocks suck:
> We study long-run shareholder outcomes for over 64,000 global common stocks during the January 1990 to December 2020 period. We document that the majority, 55.2% of U.S. stocks and 57.4% of non-U.S. stocks, underperform one-month U.S. Treasury bills in terms of compound returns over the full sample. Focusing on aggregate shareholder outcomes, we find that the top-performing 2.4% of firms account for all of the $US 75.7 trillion in net global stock market wealth creation from 1990 to December 2020. Outside the US, 1.41% of firms account for the $US 30.7 trillion in net wealth creation.
* https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3710251
> Four out of every seven common stocks that have appeared in the CRSP database since 1926 have lifetime buy-and-hold returns less than one-month Treasuries. When stated in terms of lifetime dollar wealth creation, the best-performing four percent of listed companies explain the net gain for the entire U.S. stock market since 1926, as other stocks collectively matched Treasury bills. These results highlight the important role of positive skewness in the distribution of individual stock returns, attributable both to skewness in monthly returns and to the effects of compounding. The results help to explain why poorly-diversified active strategies most often underperform market averages.
* https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2900447
And this certainly can have a financial effect on your finances: having the "wrong" stocks in your portfolio (i.e., most of them) and not have the "correct" ones will mean a (e.g.) comfortable retirement or not.
This is wrong. There is no lookahead bias in the factor.
I mean these stocks have been performers for decades. If you posted this 10 years ago you'd look really wrong.
Yes, but when their run ends they tend to underperform.
Every time.
If a stock market observation has no predictive power, then it's worthless.
I look forward to your weather report too: "It's always sunny outside until one day it starts raining. Every time."
> I look forward to your weather report too: "It's always sunny outside until one day it starts raining. Every time."
I once ran across the comment that if you simply predict tomorrow's weather will be the same as today's you'd be correct 80% of the time. Not sure how true that is (can't find the source).
Allegedly momentum investing does pretty well:
* https://en.wikipedia.org/wiki/Momentum_investing
(I'm more of an index guy myself.)
ye, the stock market isn't magic, it's just a collection of what people think. and people can be very wrong in a big way.
Buddy I think you missed the joke.
Off topic, but I love your username.
hey he hacked my computer his user has a home folder inside of /dev
Not at all, if someone tells me that "This stock is historically likely to regress to and beyond the mean," it's information I can use to evaluate my risk tolerance. Just because a piece of information doesn't let you time the market like a psychic doesn't make it worthless, it's just not what you were looking for.
Many people who replied to you seem to have missed your joke. I appreciated it.
It is my curse.
Years ago, My daughter's science teacher said that school should teach a love of learning.
I replied 'I thought the point of school was to make productive worker units in society'
And while he explained to me why I was wrong I was thinking to myself 'great, now he thinks I'm a terrible person'
It seems I deadpan too effectively.
Productive workers in society need to learn new things all the time. I can't think of any career that hasn't changed in my life. I recall a garbage man (sexism probably wasn't required even then, but I never recall females) hanging off the back of the truck while the driver drove to the next house - the driver today needs to know how to operate the arm on the truck that lifts my can. Fast food used to be cooked within 10 minutes of when it was thrown, now they obviously are keeping things warm for a lot longer.
Poe's law apply to real life too.
Once you lose, you have lost. Ok, but how does that help us predict when something will lose?
Yes, it is predictive. But only retroactively.
Sometimes people bring me things that are broken 'cause I like to fix stuff. They always say "it was just working!"
Tautologies 'R us
well I laughed
“When the stocks don’t go up they don’t match the market which generally goes up”
> I mean these stocks have been performers for decades. If you posted this 10 years ago you'd look really wrong.
And Japan performed ridiculously well for over decades and then stagnated for decades after that, but it averaged out between the two periods:
> Ben Carlson: It's just a really long mean reversion. You got like 22% per year from 1970 to 1989 in Japan. Small caps in Japan did 30% per year for two decades.
> It's insane. The returns almost had to be poor after that. If you put them together, the boom with the bust, it's like almost 9% per year.
> It's kind of crazy. Over 50 years, the long-term worked. It's just that over that 20 or 30-year period, it didn't work so well.
* https://rationalreminder.ca/podcast/412 (~4m20s)
Annualized 9% per year is pretty good: the S&P 500 has average 10% since 1957 (70 years). Is there anything preventing US equities from doing the same thing: great performance from 2010 until now, and then 10+ years of stagnation starting (theoretically) tomorrow. If you look at 2000s S&P 500 you got zero returns, and the only thing that would have saved a US domestic (only) investor was having a bond allocation:
* https://www.forbes.com/sites/advisor/2010/09/13/its-not-real...
This is why diversification is important. People talk about "US stocks" doing well, but have US industrials done better than non-US industrials? US finances or energy done better than non-US? Or are "US stocks" doing better simply because tech stocks specifically have done better? Perhaps a US allocation is really a tech sector play:
* https://ofdollarsanddata.com/should-your-portfolio-be-100-us...
I am invested in some of the companies that are downstream of the capital expenditures of Big Tech (e.g., COHR), so I have nothing to complain about.
I am really struggling to see what's the investment thesis behind Google valuation increasing 2x in response to AI, though. Assuming no magical AGI singularity, by the end of the day, they're still selling the same services, but the services have gotten more expensive for them to provide. Everyone was already using Google Search, but now, provisioning AI summaries on top of requires more compute. Everyone was already using Google Docs and Meet, but now, AI features cost Google more. Etc, etc.
The only place where they stand to make money is selling AI compute to enterprises. But with the current supply-chain challenges, the margins there are probably getting thinner.
`I am really struggling to see what's the investment thesis behind Google valuation increasing 2x in response to AI`
Google is basically Nvdia (TPUs), Tesla (Waymo Self-Driving), Hyperscaler, Netflix (YouTube) and a massive VC (Anthropic, Databricks, SpaceX, etc.) all rolled into one.
Their valuation isn't really a 2x'ing so much as a reversion from halving.
It's frequently said that vonglomerates usually carry through as their valuation, the multiplier of the lowest holding. If that's the case, Alphabet is worth more broken up. Similarly that makes their current valuation (and nay any valuation) theoretically too low.
>Assuming no magical AGI singularity, by the end of the day, they're still selling the same services, but the services have gotten more expensive for them to provide.
Well, they're hoping to sell new services on expensive $200/month subscriptions.
The hope is that agents have more value than traditional software, because they do the work for you instead of just enabling you to do the work.
The investment thesis is a torrent of cash arriving to index funds and retirement target date funds has to go somewhere.
Google has a compelling story for many AI scenarios: it has lots of outs. It's the only frontier lab for which that's true. A massive bubble bursting wouldn't be existential for Google; it would be quite painful, but survivable, and even offers some potential upside (picking up assets and researchers from the wrecked, mangled corpses of other frontier labs on the cheap).
Google is a good bet because if AI continues to boom, they're in a good position (frontier lab, vertically integrated). If the bubble bursts and the frontier labs fail, they might do even better.
I know several non tech companies that use Gemini and NotebookLM heavily (banks, insurance, consulting).
Google is a money printing machine, and their Q1 revenue and profit were up significantly vs last year.
Same for META
The market runs on memes, hype and fraud. Fundamentals haven’t mattered for a long time.
You're half right.
The market sets prices, and they are set based on multiple things. One of those is fundamentals. Consider the value of assets, whether tangible or intellectual property, human resources, binding contracts, etc. that add up to reasonable revenue forecasts and so forth.
And the other aspect of prices is based on conjecture, speculation, meme-joiners, believing hype, and in some cases, fraud.
The secret sauce is always going to be the one who can figure out, between the two factors going into price, what's right, and when.
BUT... just saying that all stock market pricing is based on unreliable factors? That's not a useful, actionable statement. You can certainly stay out of investing in that market, but is that going to be your best course of action?
I guess Google’s Q1 earnings of $62 billion were just hype.
And a market cap that exceeded 4.5 trillion dollars. I think we can all agree that Google is a legit business. But they were doing fine in 2023 too and had solid y/y revenue growth for much of their history. Their market cap tripled since 2023 and doubled since mid-2025. So, it's clearly more than "they continue to be a profitable company that grows around 15% y/y in real terms".
Yes, earnings are essentially hype/BS.
Focus on cash/cashflow.
$64B trailing twelve months free cash flow. (https://s206.q4cdn.com/479360582/files/doc_financials/2026/q...)
That's trailing and mighty impressive. However when you decided to quote Q1 earnings. $62bn of Q1 earnings ends up at only $10bn. The LTM numbers are less relevant.
AI-related capex is one hell of a drug.
The discussion here is whether Google can make money in the AI world. They obviously can. Whether they turn around and spend that money on capex to try and increase future earnings is irrelevant.
I like that the invisible hand of market is slapping the Mag-7 for capex which is the only way to discipline them. Investors are waking up to say: hey, you are spending all your profits on data-centers, where is the return for me ? But, it surprises me that there are vast pools of capital which we collectively call the "market" that makes these calculations, or maybe a simpler causal explanation is the missing stock repurchase bid. At some point, one of the hyperscalers (msft ?) will break from the pack and announce reductions to capex and increase stock repurchases to stem the decline.
Why is Apple included though?
You can simultaneously believe that the hyperscalers are becoming capital-intensive long-term in a way that’s bad for their profits and that as a result will be raising the COGS of Apple’s business in a way that also hurts profits.
My perception is that the market lumps companies together into sectors whether they have similar business models or not. When one is punished, other associated firms tend to be, too. You see this in any industry (not as a guaranteed rule, but in general).
Share repurchases also aren't great I guess?
Markets generally love buybacks.
The market is both happy that Apple didn’t spend all its cash on AI build-out, but also at the same time angry that Apple is “missing AI”.
Not to mention the grumblings that Apple has peaked.
because $100bn annual revenue, high FCF, high margin. Its a monster.
Could outperformance of largest cap companies be partially explained by dividends paid by smaller caps? Reasoning behind this:
* Index investing raises in popularity, with index funds that automatically reinvest dividends being often preferred due to their tax efficiency.
* Large caps prefer to repurchase stocks, stock repurchases contribute fully towards a given company share price increase.
* Smaller caps still pay dividends, these dividends are then reinvested by index funds and the reinvestment is weighted by capitalization, so large caps share price benefits more from repurchases done with dividend cash paid by smaller caps. When dividend is paid, share price of a company that paid it is reduced, which further widens the performance gap between large and smaller caps.
>* Large caps prefer to repurchase stocks, stock repurchases contribute fully towards a given company share price increase.
That doesn't work because indices are typically market cap weighted. A stock buyback might increase nominal stock prices, but not the market cap, otherwise it'd be a free money machine.
While I agree with other comments indicating that the headline is drawing on a small period of data, other data in the deck is pretty compelling.
Page 25 "The number of data centers in the US" gives an interesting insight as to the magnitude of the data center boom. 60% more data centers are being planned or are under construction. This might actually be underselling it in dollar amount, as I believe the average data center size under construction is larger than the average data center already constructed.
Page 27 "Cyclically adjusted P/E ratio near all-time highs" is certainly concerning and points to a near term correction.
Apollo should be smart enough to know that you can't draw any conclusions from 1 month of market data (especially when there was a big, relevant IPO).
It's more like a year of underperformance, sometimes longer
Right, but they're smart enough to know how to manipulate the market short term with such "research" and then take advantage of it, all without breaking the law.
I'm astoundingly unimpressed by the quality of this slide deck. There's no analysis except that crammed into slide titles, like it's designed to bombard a room full of analysts with so many graphs that they shut off their critical thinking. Could a freshman business student not make this? Could a freshman business student with an LLM not make something more convincing than this?
I agree with the headline, but this really feels like analysis-slop. It's only remarkable in terms of who is publicizing it.
I assumed it was accompanied by an oral presentation or another report. I didn't look around enough to see if there is a transcript or anything though.
This was my reaction, as well. It's almost entirely technical analysis mumbo jumbo.
Actually it's almost 2 months /s
Nothing wrong, necessarily, but Tesla doesn’t belong to that seven they never have.
What TSLA lacks in cashflow, they more than make-up in TAM. Make-up as in create out of thin air. We'll have self driving cars in a few months, robots, semi-trucks, etc., etc.
Elon is missing the bigger TAM here. Once they've created a time machine, they can not only sell all these things backwards and forwards in time, they can also start making industrialization come way earlier, which will lead to exponentially turbo charged growth. Plus importing future designs into the now.
It makes sense to focus all effort into building a time machine.
Always a few months away. The nuclear fusion of products portfolios.
Don't forget that no auto manufacturer in the world, and certainly not in China, is poised to capture any of that TAM.
Maybe they'll buy a failing AI company and update their grift portfolio.
6 months ago when Mag7 was overperforming everyone was worried about it being too high a fraction of the S&P500.
Exactly. As usual people are bullish at the top and bearish at the bottom.
Is this the bottom?
You shouldn't take financial advice from me :)
On slide 6, they list the Mag7 stocks (not defined in a foreword), but on slide 8 they list the free cash flow (FCF) of a somewhat different set of companies. Why not stick to the Mag7 FCF only? It muddies the waters.
Apple is not burning its cash on buildout, which would make their graph less interesting.
True. This slide also struck me the most, but for the data it shows. Free cash flow for all the hyperscalers basically evaporated in the last couple of months, with Oracle in the minus. What does this mean?
the FCF slide is for hyperscalers - which excludes TSLA, NVDA
Yes but the Mag 7 is the main focus, suddenly switching to a different set of companies is comparing apples to oranges. That slide belongs in an annex.
That slide should probably be the main focus, with the rest supplementary.
That is the slide that is likely to pop the economy.
I was here when you called it!
Is that Oracle down there increasingly in the negative? & Amazon's free cash flow reaching zero?
So many chart crimes, but aside from that.
Seems like we're seeing margin get eaten up by companies upstream of the buildout, and the costs have not been fully passed downstream yet. Eg most consumers still get free/cheap AI.
It'll take a few years to see what happens at scale when prices go up and purchasing behavior changes. Hardware, services etc all downstream of the buildout and supply constraints.
Markets are obviously, rationally, not happy about the overspending.
But what gives me pause is that a some of the mag 7 (think Meta) could change their mind on AI build-out tomorrow, and 1-year from now have the same amazing free cash flow they always did.
I don't think most people quite grasp the sheer size of the debt obligations Meta and specially Oracle have gotten into. Meta has done some clever tricks to keep it off the main books, but it IS there. If no AGI, its gonna drag them down for a decade.
I tend to agree. I suspect if we look back a few years from now that there was some overspending but I still believe the risk of not investing and missing out is greater than the current historical trend of capex spend. I have not looked recently but has demand of compute started to slow down. It was just a few months ago when companies like Anthropic were throttling users because there was not enough of it.
What has been the best way to determine return on the AI specific capex for hyperscalers?
I would naively expect Microsoft’s to be the highest since they are probably mainly just selling access to their capex through cloud since they aren’t seriously pursuing frontier AI, I’d imagine Google to be in the middle (selling TPUs, general cloud GPUs, Gemini, revenue lift on ads from better AI) but also spending heavily on infra to compete with OAI/ Anthropic, and then Meta to be on the low end since they are likely getting serious revenue lift from AI but not monetizing their models by API.
That free cash flow drop at AMZN is surprising.
yeah i was interested as well. -$2.5B from $11B last year. However the analysts see it recovering quickly. If you put all the Trainium(N) chip investments aside as being all AI risk, they still have some good pipeline with the satellite internet stuff (Delta, JetBlue, Vodaphone).
OMG - ridiculous to evaluate it based on the last 3 mo.
I think page 4 is a little disingenuous, I'm pretty sure you could pick lots of windows and show the mag 7 deviating negatively, but then later trends positive. I believe this whole thing will pop, but I'm not quite convinced it is just yet.
Page 8 for Oracle's free cash flow trending negative is really quite impactful. The Bloomberg AI bubble diagram [1] shows how this could really blow up. If Oracle falls they could take the whole market with them. We're just waiting on the mag 7 or any closely linked companies to fail to raise investment.
Page 16 really outlines how insane these evaluations are. I think most countries see it, hence aggressive selloffs of US bonds [2]. But everybody is just too insanely heavily exposed to it all now, it's going to wipe out everything. It's going to be a very awful time when heavily debt strapped countries can't issue debt anymore.
I think what we're going to see is some insane moves to keep these companies afloat longer in some desperate attempt to delay the pop, which will just make a bigger bubble. I could see Nvidia for example issuing bonds in excess of $100bn soon if the market has appetite for it [3].
[1] https://archive.is/0bYLS
[2] https://sg.finance.yahoo.com/news/china-japan-uae-india-sell...
[3] https://uk.finance.yahoo.com/news/nvidia-raises-over-21-5bn-...
as soon as you start calling a group of stock tickers the “magnificent 7” they’re destined to underperform, as that’s a feelings based assessment and many investors will continue to buy the feelings long past the value being fair.
“Do they make money? I don’t know but I know they’re magnificent!”
Some of this might be mechanical since so much money went into the chip companies.
See also:
https://finance.yahoo.com/markets/article/magnificent-7-stoc...
> The once high-flying "Magnificent Seven" are looking more like the Dreadful Seven.
> The why: Wall Street is growing increasingly impatient with Big Tech's astronomical capital expenditures on artificial intelligence, projected to balloon 70% to exceed $700 billion this year.
Completely missing the point of not adding Anthropic, OpenAI, SpaceX.
Mag7 is just an arbitrary list of companies that were coined at a specific time with no rigor.
If you must do rigorous analysis, then just like S&P 500 you need to add drop BigTech / High Growth companies continuously to this Mag7 (or Mag10) and then do analysis
I only care what the market is at >25 years from now
Yeahhh but...
OK, I guess it _would_ be pretty hard to have an _unrecoverable_ crash...
Hyperscaler free cash flow chart is interesting, but why does it omit Apple?
This is like a straightforward Red Queens Race story, isn't it?
Planned Capex for 2026:
Amazon $200B
MS $190B
Alphabet $175B-$185B
Meta $115B-135B
Apple is nowhere to be found.
$14B (!!!)
https://finance.yahoo.com/news/apple-lazy-ai-strategy-could-...
This feels like telecom "massive demand, bad returns".
If a good enough model can be swapped in every few months, the value moves away from the model and toward cheap inference. That is great for users, but not always great for returns on huge capex.
The only mag 7 that had a shooter chance with models is Google rn so doesn’t seem like there being good enough models will be that negative for mag 7. Might even be better since they are the providers for the inference
The dumps on memory/hardware stocks when these companies scale back purchasing is going to make Trump's shitcoins look sovereign bonds
-signed a bitter somebody that had to buy a new SD card for my camera last week.
haha Oracle bleeding money. If AI ended up killing Oracle that would've been so great.
Not mentioned in the comments but all those companies are taking significant debt and even issuing new shares.
One major part of the investment thesis in these companies were their constant stock buy backs. Now their gargantuan Capex that sees no end but acceleration is back at diluting investors.
I think the companies will keep doing fine, but the financial outlooks are no longer as rosy.
US Big Tech is betting the house on AI, news at 11.
MU is up like 1,700% since 2024, so yes they are, but "under" is relative.
I guess "nothing lasts forever".
Of course. The rest 'n vest people at FAANG companies will tell you its lasts forever 6 years ago. Now they are scared for their jobs.
This is why we are seeing a correction at those companies, perks and free food going away with constant layoffs and all time low morale.
Now the party is at Nvidia. But I will tell you that that will not last forever either.
Everything reverts to the mean in the long term.
Maybe even the state of unprecedented relative peace we had enjoyed.
It is less that the Mag 7 is starting to underperform and more that a market correction is likely and coming very soon.
This is probably a hot take and I am by no means a financial expert and this is probably quite wrong. I personally think that attempting to value these companies using the same methodology as the history of all American companies is fundamentally wrong. Sure, some mom and pop small local regional business that overperformed is probably more likely to underperform. But when it comes to big tech companies, these companies are operating a data and capital flywheel that doesn't easily just slow down. I mean, you look at the history of these computer tech companies, especially software companies. They really haven't slowed down. Like, look at Microsoft. It's just been growing from the very beginning, pretty much.
This seems healthy to me. A market where returns are less dependent on seven mega-cap names is probably more stable, not less. If earnings growth broadens out and capital starts flowing back toward quality and free cash flow outside the obvious AI winners, that should reduce concentration risk and make the whole market less fragile.
You can tell these guys know nothing about LLMs or how they’re provided. I love how they show OpenRouter’s graph of token usage as if it speaks for usage across the board. DeepSeek looks like the king because people who use Anthropic and OpenAI use them on either a direct basis or AWS Bedrock …
And the bar chart for token costs, really? As if that’s information? Their sources are the API docs ffs. If they had at least modeled something to estimate token costs that would be interesting, but showing the public prices and calling it research is dumb.
"I love how they show OpenRouter’s graph of token usage as if it speaks for usage across the board."
I'm not sure what was said during what looks like a deck of a presentation? I'm hoping it wasn't this, because that's an obvious misfire.
Apollo Group has $1T assets under management, I believe them over most folks on HN. Their argument that the Mag 7 is burning up all of their free cash flow is factually accurate, as the returns are not materializing for the investments being made (ie underperformance).
The Mag 7 spending around $700B on capex this year, expected around a trillion next year.
That's more than their combined FCF, and they're borrowing to bridge the gap.
One is not spending anything out of the ordinary up until this point in comparison to their competition. But the Apple Silicon design group might start spending some money on something practical, something that might begin the typical Apple long-range process of replacing two or three existing companies within their current supply chain, but that’s nothing new over the last 25 years.
It is a bit unfair to lump Apple in with everyone else in the Mag7, as their strategy is arguably reasonable and prudent for the current position of the hype circle. Apple Silicon was a spectacular decision and investment. Everyone else has lost their damn mind.
If we're doing money=smart, the mag 7 control almost $4T in assets and cash, so wouldn't you trust them even more?
Not based on the evidence of how they allocate capital. Meta wasted $80B on the Metaverse, for example. Apollo allocates capital as their day job. Mag7 allocates by vibes.
https://finance.yahoo.com/sectors/technology/articles/mark-z...
https://www.nytimes.com/2026/03/19/technology/mark-zuckerber... | https://archive.today/iEGAj
> Apollo allocates capital as their day job. Mag7 allocates by vibes
Lol what in the world? Is running a trillion dollar corporation anything other than allocating capital?
You can argue that Meta made a poor capital allocation decision with VR and perhaps continues to do so with AI.
I have an opinion based on the data, yours may differ. Favorite this thread for when the music stops. It has before (1999-2000, 2007-2008), it will again.
I'm objecting to your categorization of the job of running one of the Mag7s as something other than "capital allocation". The CEO of Meta or Amazon or any of these other companies allocates capital all day long. They do it differently than an asset manager like Apollo for sure, and sometimes make worse decisions. But it's in the same category.
Vanguard has 12T under management so you believe any PDF they put out 12 times more?
Vanguard holds and manages index funds and are, broadly speaking, not active managers. Apollo is a sophisticated, active capital market participant|manager.
https://en.wikipedia.org/wiki/Apollo_Global_Management
I mean, I'd assume this deck was trying to sell something. But the relevant comparison is to stonk-bros on HN LARPing as hedge fund managers, not Vanguard.
Which, to be fair, Vanguard has earned a good deal of trust from me on the passive investment side of the equation, although I don't think it's meaningful to make it linearly proportional to the total size of their managed assets? I'm not even sure how I'd operationalize that in reality.
Having actual skin in the game on getting an answer right is generally a sign of credibility, though.
That said, without context I'm not really drawing any conclusions from this.
> Apollo Group has $1T assets under management, I believe them over most folks on HN
Ah, the old _argumentum ad giletum Patagoniae_
in pecunia veritas , an old HN favorite.
> Apollo Group has $1T assets under management, I believe them over most folks on HN
If only that's a measure of if they know how to time the stock market. The real measure is how much they earn not if their sales or marketing department manages to get more customers.
they are building new narritaves with made up BS name like mag 7 this, that. it used to be faang BS but they updated so lower IQ folks gets into new narratives. Marketing dep. and fin. engineering working together. Another older one was ZIRP that, ZIRP this and you see lower and avg. IQ folks talk about it along with lots of bots. ZIRP bots stopped so MAG whatever bots can take over.
Waiting for next game, bets are open, any new names? O A S I S? upcoming IPOs and new names needed so they can exit scam