Buffett once bet $1M that he could beat a group of hedge funds over 10 years

119 points by hhs 2 years ago on hackernews | 219 comments

nabla9 | 2 years ago

Misleading hedline.

Buffet bet $1M that S&P 500 ETF (VOO) index fund would beat hedge fund portfolio selected by Ted Seides. Low cost index funds performing generally better is well documented. Buffet's bet is common sense.

Reasons why:

- lower cost is important over longer term.

- Well diversified index fund picks every success, unlike actively managed portfolio.

epolanski | 2 years ago

Point two is double edged, it also picks all losers.

WrongAssumption | 2 years ago

Sure, but the losers go down between 1-100%, while winners are virtually uncapped.
It's a pretty common strategy to sell covered calls on your portfolio.

That caps your upside from infinite to finite. But you collect a steady stream of option premiums. The result doesn't look all that different on average from just investing in the plain index fund. You just get a different risk distribution.

(Similarly, you can invest in an index fund that tracks (junk) bonds. Upside is finite there as well.)

FullyFunctional | 2 years ago

As Motley Fools [1] like to point out, the most you can loose is 100% whereas the upside is unbounded. This means that even if all but one investment turns completely to dust, if that one is outstanding it makes up for it all.

We are actually in that situation now where the gains of S&P 500 relies completely on the top handful of companies and as much as (IIRC) 70% of the companies in S&P 500 are lagging the market.

[1] This is not an endorsement; my personal experience with their pool was mediocre at best.

jacquesm | 2 years ago

> if that one is outstanding it makes up for it all.

It may well be outstanding and still not make up for it all.

> This is not an endorsement; my personal experience with their pool was mediocre at best.

That's the opposite of an endorsement!

thfuran | 2 years ago

>It may well be outstanding and still not make up for it all.

That sounds like, at best, pretty great.

> As Motley Fools [1] like to point out, the most you can loose is 100% whereas the upside is unbounded. This means that even if all but one investment turns completely to dust, if that one is outstanding it makes up for it all.

I doubt that qualitative assessment holds up when you add some numbers.

As a thought experiment, imagine a portfolio that generally replicates the S&P500, but also sells covered out-of-the-money calls on all the stocks. (Whenever a call triggers, you re-balance your portfolio. Let's ignore transaction costs for now.)

According to the efficient market hypothesis, the portfolio sketched above will have roughly the same average returns as the S&P500. Especially if the calls sold are far-out-of-the-money.

geysersam | 2 years ago

> As Motley Fools [1] like to point out, the most you can loose is 100% whereas the upside is unbounded.

Reminds me of the gamblers ruin paradox. In some games even if the expected payout from each round is positive, it can be shown that the gamblers wealth goes to 0 over suffiently large time scales with probability 1.

chrinc9203 | 2 years ago

> it can be shown that the gamblers wealth goes to 0 over suffiently large time scales with probability 1.

In real life, the time scales are not long enough and the number of N samples in one’s life is too small.

This is why retirees over age 60 are concerned with “sequence risk”. That is, you get unlucky VTI/VOO/S&P returns for 5 years, but you don’t live long enough for the average to come back to 8-10%.

Downside risk becomes more important than average return.

jedberg | 2 years ago

I have the same sequence risk when I gamble. I know the average is that I should lose, but I also know I'll never play enough in my lifetime to get enough samples. So I play games that are more likely to have sequences -- hand shuffled blackjack -- and more likely to win quickly and loose slowly -- the don't in craps.

lotsofpulp | 2 years ago

> We are actually in that situation now where the gains of S&P 500 relies completely on the top handful of companies and as much as (IIRC) 70% of the companies in S&P 500 are lagging the market.

I believe I saw data that showed this has always been the case, but I can’t quickly find it.

throw0101d | 2 years ago

> Point two is double edged, it also picks all losers.

Yes, but picking losers is a lot easier and winners a lot harder:

> 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

Over ten years, most individual stocks under perform a market index (even more so if stock was initially a top performer):

> […] 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

Most stocks suck, and chances are generally low that you'll pick the non-sucky ones. Further, you have to buy when a non-sucky stock is or becomes not-sucky and then know when it does become sucky and sell before it drags down your returns.

And not only do you have to be "good", you really should be better than just going for market returns (via a low-cost index fund).

> But, what about stock picking? How long would it take to determine if someone is a good stock picker?

> An hour? A week? A year?

> Try multiple years, and even then you still may not know for sure. The issue is that causality is harder to determine with stock picking than with other domains. When you shoot a basketball or write a computer program, the result comes immediately after the action. The ball goes in the hoop or it doesn’t. The program runs correctly or it doesn’t. But, with stock picking, you make a decision now and have to wait for it to pay off. The feedback loop can take years.

[…]

> This is the existential crisis that I am talking about. Why would you want to play a game (or make a career) out of something that you can’t prove that you are good at? If you are doing it for fun, that’s fine. Take a small portion of your money and have at it. But, for those that aren’t doing it for fun, why spend so much time on something where your skill is so hard to measure?

* https://ofdollarsanddata.com/why-you-shouldnt-pick-individua...

If (deliberately) picking losers is easy, you should buy an index fund and do some careful shorting on the side.

Instead of saying 'most stocks suck', you should probably say that probability distribution of future prices of stocks is highly skewed, but current prices reflect the expected future price. Logically, the median stock (and thus most stocks) will trade above its future potential.

Suppose you had a lottery where 90% of tickets pay one hundred dollar next week. You wouldn't say those tickets suck. You'd be happy to own lots of them.

Now suppose I tell you that the other 10% of tickets pay one million dollars next week. That makes the tickets even more desirable.

But in an efficient market, all tickets will trade at approximately 100,000 dollars today. That will make it look like 90% of tickets suck. But both in lottery tickets and in stocks, suck-age is relative to prices paid.

If 60% of tickets paid X, and 40% paid 0, then tickets would trade for about 0.6 * X, and most tickets will look like they don't suck: most tickets almost double in value.

> If (deliberately) picking losers is easy, you should buy an index fund and do some careful shorting on the side.

This doesnt make sense. The "losers" in this case don't necessarily go down in price, they just go up in price less than the overall market. Shorting these losers would still lose you money.

Its easy to pick a stock that goes up less than the market, its hard to pick a stock that goes down (at least one that goes down by more than the cost to borrow and short it).

CodeWriter23 | 2 years ago

>Point two is double edged, it also picks all losers.

Not really. Standard and Poors evaluates the top 500 on a quarterly basis. The losing stocks are naturally purged from such index funds on a regular basis.

They can only pick them after they have already lost.

In any case, broader index funds / indices that don't eliminate users as quickly (basically, only when they stop trading on public markets), don't do worse. So the mechanism you describe is pretty much irrelevant to the success of index funds.

And so is the implied mechanism of 'picking all the winners'.

bombcar | 2 years ago

And the index fund “trick” works just as well with total market indices (mainly because the total market return is those same ten percent captured by the S/P 500).

KRAKRISMOTT | 2 years ago

> Buffet's bet is common sense.

It wasn't popular common sense until buffet and Bogle proved it.

dan-robertson | 2 years ago

I’m not sure it was considered common sense when the bet was made. And when it settled there were various claims that Buffet got lucky in that stocks just went up over the 10 years – perhaps he wouldn’t have won if there was some kind of crash.

I’m not claiming he was wrong but I’m not sure one can really reduce it to those few points as if it makes it all obvious.

stouset | 2 years ago

> And when it settled there were various claims that Buffet got lucky in that stocks just went up over the 10 years

It’s funny because that streak of luck just seems to keep happening for most 10 year periods.

nkurz | 2 years ago

It's a little less funny if you look at the inflation adjusted numbers: https://www.macrotrends.net/2324/sp-500-historical-chart-dat...

There have been quite a few 10-year and longer spans where the S&P 500 doesn't beat inflation. Did you realize that if you had invested in 1965 it would have taken you until 1990 to beat inflation? And that if you had invested at the peak in 2000, it would have taken you until 2015?

Even apart from inflation, there's some recent research which suggests that historically the stock market as a whole just hasn't performed as well as most people think. This paper was submitted by the author a few days ago but got no discussion here: https://news.ycombinator.com/item?id=38835832.

I'm not saying you're wrong, and I too hope something has changed, but do realize you might be assuming without sufficient evidence that the recent past is normal behavior. By historical standards, 2008 to 2022 really was an anomalous bull run.

[OP] hhs | 2 years ago

> By historical standards, 2008 to 2022 really was an anomalous bull run.

And interestingly, the rise of intangible assets over time: https://anderson-review.ucla.edu/boom-of-intangible-assets-f...

sega_sai | 2 years ago

The statement about 1965 is not correct if you take into account dividends. The longest periods of being down (in real terms) is 1973-1985, 2000-2013, 1937-1945, 1916-1925 ( https://observablehq.com/@tcgarvin/shiller-sp500 )

tdullien | 2 years ago

I graduated from high school in 2000, and sold my first company to Google in 2011, insisting on cash and not stock given the performance of the market through my entire adult life ;)

bombcar | 2 years ago

The brilliant play is to be like the elves and say both yea and nay. 50/50 and either way you can look brilliant!

nkurz | 2 years ago

Thank you for the correction. I had mistakenly thought that the chart was total returns rather than just index price, but looking closer I think you are right.

refurb | 2 years ago

You're ignoring dividends.

matsemann | 2 years ago

Most people don't invest a single lump sump, but keep investing over time. So even if there is a dip in a 10 year period and it's "only" back to start after those 10 years, all the money one invested during the dip are now in the plus.

lotsofpulp | 2 years ago

> By historical standards, 2008 to 2022 really was an anomalous bull run.

Why leave out 2023?

nkurz | 2 years ago

Because I meant "through to the beginning of 2022", with the bulk of 2022 not being included in the run. 2023 was definitely up, but whether it will be interpreted as a continuation of the previous run, a brief reprieve, or the jittery start of a new bear market will probably depend on what happens going forward.
> I’m not sure it was considered common sense when the bet was made.

the academic research even back then was pretty evident that a market cap weighted index fund has been outperforming hedge funds (over long periods like 10 yrs).

It's just that index funds way back, prior to vanguard, was not as easily accessible as it is today.

bombcar | 2 years ago

Hedge funds are what their name says - they’re supposed to be a hedge for particular situations, not a main investment. But they became “cool” to own and talk about.

Properly used, a hedge fund should just be a drag on your portfolio, but you have it because you’re hedging against risks that others don’t even bother with (like, say, total US collapse).

Zanni | 2 years ago

The bet was made in 2007. A Random Walk Down Wall Street [0], considered the book that makes the case for index funds over a hand-picked portfolio, was published nearly 25 years earlier (1973) and was sufficiently well-known to qualify as common sense. Though there are still plenty of investment managers that are motivated to claim otherwise.

[0] https://en.wikipedia.org/wiki/A_Random_Walk_Down_Wall_Street

lelandfe | 2 years ago

Befitting this thread’s language, even, Vanguard’s Bogle published Common Sense on Mutual Funds in ‘99.

fspeech | 2 years ago

And Buffett made much larger bets around the same time by selling puts on various stock indexes, with European style contracts and no collateral requirement. Recall collateral calls were what killed AIG.

rTX5CMRXIfFG | 2 years ago

Intuitively, it seems like he still would have won even if there was a crash—the hedge fund portfolios would then earn even less, because of less diversification and more agent fees.

ryl00 | 2 years ago

> perhaps he wouldn’t have won if there was some kind of crash.

But there was! That bet started in 2007, and 2008-2009 saw the biggest percentage drop in the S&P 500 in decades with the GFC. We're talking down to levels last seen in 1996, at the worst of it in March 2009.

westurner | 2 years ago

The S&P500 is a backtesting reference benchmark.

You can compare portfolio performance through drawdown periods with backtesting tools.

Macroeconomic drawdowns are good times to have cash for acquisitions; instead of free government cheese.

"Tear Sheets: Definition and Examples in Finance, Vs. Prospectus" https://www.investopedia.com/terms/t/tearsheets.asp :

> While tear sheets date back to the old days when stockbrokers would rip individual pages out of the S&P summary book and send them to current or potential clients, most information is extracted online today. Therefore, any concise representation of a company's business fundamentals could be considered a tear sheet.

From https://news.ycombinator.com/item?id=24428206#24429801 :

> pyfolio.tears.create_interesting_times_tear_sheet measures algorithmic trading algorithm performance during "stress events" https://github.com/quantopian/pyfolio/blob/4b901f6d73aa02ceb... :

>> Generate a number of returns plots around interesting points in time, like [...]

From https://news.ycombinator.com/item?id=19111911 :

> pyfolio/examples/zipline_algo_example.ipynb: https://nbviewer.org/github/quantopian/pyfolio/blob/master/p... > "Worst Drawdown Periods"

Drawdown > Trading definitions: https://en.wikipedia.org/wiki/Drawdown_(economics)#Trading_d...

awesome-quant > Python > Trading & Backtesting: https://github.com/wilsonfreitas/awesome-quant#trading--back...

S&P500: https://en.wikipedia.org/wiki/S%26P_500

yen223 | 2 years ago

Here's a fun nerd-snipe problem: if we keep the terms of the bet, is there any period in which Buffett would have lost the bet?

otherme123 | 2 years ago

That "luck" always suppose that managed funds would profit on bear markets by shorting, and profit on bull markets by buying. As indexes cannot short, they always lose in a bear market.

But in reality managed funds usually freeze or even buy "opportunities" on the falling phase and short on the bottom (e.g. 2009), which causes them serious loses. Also they tend to miss the start of the bull runs (e.g. being bearish until 2011/2012).

If you read Kostolany (a contrarian), he talks about how hard it is to short correctly or to buy in the bottom, because going against the grain is hard.

bombcar | 2 years ago

You can usually identify the direction, what is nearly impossibly hard is identifying the time.

Also known as “the market can remain irrational longer than you can remain solvent”.

jjice | 2 years ago

"Common sense" can be subjective here, but Bogle was preaching index funds for quite some time leading up to this. Vanguard and other low cost index providers were doing well, but it is true that it's increased quite a bit over the last few years.

In 2007, one of Bogle's more approachable books, "The Little Book of Common Sense Investing" came out, as a reference for this kind of philosophy across other writers: https://en.wikipedia.org/wiki/The_Little_Book_of_Common_Sens...

bombcar | 2 years ago

Bogle basically invented the index fund because he couldn’t work out anything that could beat the index. The “aha” moment is a fund that just copied the index would be quite cheap to manage and maintain.

His math always worked but over time the returns have become more and more impossible to ignore.

(Interesting for those who follow Bogle he later pointed out an issue with index funds eating the world - what he called manager’s capitalism.)

LewisVerstappen | 2 years ago

Yeah the title should be fixed. It's completely wrong lol.

fbdab103 | 2 years ago

Obviously the person on the other side of the bet thought differently.

ProjectArcturis | 2 years ago

He was buying publicity as "The Guy Who Bet $1 Million Against Buffett".

dclowd9901 | 2 years ago

If it was, why would the other party have made the bet?
It's great publicity for their fund which is super important, because the goal of a hedge fund manager is not to make the portfolio go up, it's to convince their clients that the portfolio will go up.

bombcar | 2 years ago

And if you have a hedge fund and aren’t willing to make the bet, your clients are gonna ask wtf.

BuckYeah | 2 years ago

You’re only seeing part of the picture. There are definitely hedge funds outperforming the market but their strategies aren’t scalable for obvious reasons. These funds make great returns for clients as long as they stay “small” enough to stay nimble and fly beneath the radar.

_fizz_buzz_ | 2 years ago

Or: some fund managers just get lucky, and when a lot of people start to invest into that fund at some point that luck will simply revert to the mean.

vintermann | 2 years ago

> There are definitely hedge funds outperforming the market

Of course there are, but for how long? Long enough to rule out chance?

It should be easy to make a hedge fund that outperforms the market, as long as you make two more hedge funds which over the same period, don't.

chrinc9203 | 2 years ago

> Of course there are, but for how long? Long enough to rule out chance?

Over 20-30 years.

Private equity in the US averages 15% annual return, including funds that close/died with a total loss of capital.

Citadel hedge fund was 20% annual return to the investor after fees (individual years range from -10% to 50%). Before fees (that is; their stock picking ability) returns about 40-50%.

Not to mention quant/HFT firms that consistently do 40-80% annual returns over 15-20 years. No wonder quant TC is 2-5x the TC of FAANG AI SWEs.

All the above is closed to “retail” investors, though. It’s not fair, but that’s life.

For most people (less than $10M net worth), VTI/VTSAX/VOO is the best you can do.

tdullien | 2 years ago

The outsize quant returns are to be viewed with a grain of salt:

The strategies that quant funds run are often capacity-constrained, so if they do 20-40% a year, it is non-compounding. That matters.

cbsmith | 2 years ago

The bigger question is: "how would one identify said funds?".

The economic game theory around hedge funds is not pretty. By the very nature of the markets, if one really could consistently beat the market, it would be extremely rare and valuable skill, so you could charge fees for that service that would cover nearly all of the difference between your performance and market performance. Similarly, if you could identify funds that consistently would beat the market, you'd effectively be one of the people who could consistently beat the market. This means that pretty much all investors putting money in hedge funds can't really distinguish between the funds that consistently outperform the market and those that don't. They consequently have to rely on proxy signals. For the above reasons, one of those signals is... the fees. So, if you are running a hedge fund and you wish to convince investors that you can consistently beat the market, you should charge fees exactly as if you do consistently beat the market. If you don't, that raises a question in investors' minds as to why you don't charge fees commensurate with the service you are providing.

The end result is that a potential investor in hedge funds is faced with a bevy of difficult to distinguish choices, a small number of which might actually be able to consistently beat the market, but charge fees that eat up almost all of the difference in performance. The rest charge fees that would eat up almost all of the difference in performance if they did perform that well, but instead will perform worse overall, and consequently lose the investor far more money than if they just put their money into an index fund.

Buffett’s million dollar bet was made on Long Bets, the accountability mechanism operated by Long Now Foundation. The intention of Long Bets is to encourage responsibility in prediction-making (by keeping a public roster of predictions), to encourage long-term thinking (by offering a opportunity to shape a long-term bet), and to sharpen the logic of forecasting (by recording the logic of predictions and bets.)

[0]https://longnow.org/ideas/the-million-dollar-long-bet/

jefftk | 2 years ago

I like the idea of Long Bets, though their fee is much higher than it looks: https://www.jefftk.com/p/long-bets-by-confidence-level

I've recently gotten into Manifold (https://manifold.markets) which I think is generally a better home for this sort of thing: you get the views of many more people, essentially weighted by how well they've previously done.

naniwaduni | 2 years ago

If you're using Long Bets as your betting platform, you are presumably comfortable with both (a) the Long Now Foundation and (b) donating to charity. It's not that huge of a stretch for you to be comfortable diverting a bit under half your donation to LNF. Presumably the people making these bets should be sophisticated enough to understand this implication, though reporting on them can be ... iffy.

jefftk | 2 years ago

You can be comfortable with donating to charity but also have strong opinions on which charities you want to support.

The confusing bit is that it's presented as half of the income, where unless you spend some time looking at the math it's probably not obvious that for longer timescales this ends up being close half of the total.

naniwaduni | 2 years ago

I lean toward thinking that if it seems nonobvious, then you're clearly not in a great position to be making "long bets", no? If you're making $1MM bets with 10-year timescales, I expect you to either understand this or rightly burn a quarter of your bet value over it. It's not specialized knowledge or anything, it's practically foundational to American retirement planning.

jessriedel | 2 years ago

Unfortunately, LongBets lets people make predictions on the platform even if they decline all bets from people who disagree.
Wouldn't that mean that the people who disagree can make their own prediction on the platform? I don't see a problem with this.

philshem | 2 years ago

Here’s a link to the bet, which Buffet won

https://longbets.org/362/

Lucasoato | 2 years ago

> Buffett’s Big Bet is by far the largest bet on Long Bets. The previous largest Long Bet was one for $20,000 between Mitch Kapor and Ray Kurzweil. The two prominent thinkers were betting whether an AI would pass the Turing Test by 2029. Ray was certain an AI would pass muster by then and Kapor was sure it would not get close. (Incidentally, Kapor told me recently he’s willing to double, triple, or quadruple the bet with Ray, or anyone else betting on an AI by 2029.)

Can we say that ChatGPT passes the Turing test?

brirec | 2 years ago

“Turing test” is a generic phrase. You’ll likely need to see the details of the initial bet to be able to answer that.

paxys | 2 years ago

If the question was about AGI or sentience or whatever else then we could debate it endlessly, but there is no doubt that the Turing Test was solved by computers well before ChatGPT was even a thing. Although that says more about the shortcomings of the test itself rather than the capabilities of AI.

marcusverus | 2 years ago

You can say the test was imperfectly designed, but pointless? No.

The point of the test was to provide an objective measuring stick by which we might define the moment of arrival of machine thinking. As Turing observed, a question like “Can machines think” is painfully subjective and thus is a poor measuring stick. Hence the need for the test.

bnralt | 2 years ago

I remember pretty rudimentary Chatbots from decades ago that were able to fool someone unfamiliar with them for at least a few minutes. I've even known people who did gag answering messages that were able to fool people into thinking they were talking to a real person - not for long, mind you, but for a surprising amount of time.

It's always seemed a pretty bad basis for deciding intelligence.

nyokodo | 2 years ago

> Can we say that ChatGPT passes the Turing test?

For all intents and purposes I would say yes, under the rules of the bet[1] I’m not so sure.

1. https://longbets.org/1/

mminer237 | 2 years ago

A two-hour chat? Yeah, I think anyone is guessing who the human is in that time.

317070 | 2 years ago

Does ChatGPT beat that Turing test? No. But I'm pretty sure you could use the techniques behind ChatGPT to build an AI that can survive a two hour chat. Even ten hours is probably doable. There is just no obvious market to fund building that.

peyton | 2 years ago

This is equivalent to building an LLM that is unjailbreakable. The human control can just pass jailbreaks to the examiner. I don’t think it’s solved.

garrickvanburen | 2 years ago

ChatGPTs responses seem increasingly identifiable.

If the respondent hasn’t ever experienced ChatGPT, maybe.

If they have, absolutely not

FredPret | 2 years ago

Because you’ve gotten to know it. If you chatted to a human for the same amount of time, you could identify their style.

yincrash | 2 years ago

me: can you respond in as human a fashion as possible acting like you were a human foil in a turing test?

chatgpt: Certainly! I'll do my best to respond in a human-like fashion. What would you like to talk about or ask?

me: how old are you?

chatgpt: Oh, you know, I don't really age like humans do. I'm just a computer program here to chat and assist you. But hey, how about we talk about something more interesting? What's on your mind?

tobyhinloopen | 2 years ago

Idea: you're in a chatroom with two other participants. One is ChatGPT, one is a human trying to act like AI.

You'll have to find out who's the human and who's ChatGPT

pests | 2 years ago

This is almost the true Turing test. Its two humans, one chatbot. One human is trying to detect which of the other two participants is the human and which is the robot (they know there is a robot)

krisoft | 2 years ago

And that illustrates how this bet is basically about if anyone will be financially motivated to try to make the model which beats the Turing test.

Why? The surest and best way to beat a Turing test would be to pick a fake persona and have the whole RLHF training set written from that fake persona's viewpoint.

What you are seeing in this exchange is the result of two things: The network is RLHF refined to follow instructions. The network is RLHF refined to disclose that it is an AI.

These things are done for practical reasons by OpenAI: The instruction following enables one model to perform multiple tasks. This of course needed to make the enormous cost of the training have a good return.

The disclosure training is in the training set for ethical reasons (to be precise, to avoid the appearance of being unethical, or to be even more precise to avoid the reputational cost of appearing to be unethical.)

If you really want to make a "Turing-test winning" AI you would not do either of these things. You would pick a fixed personality (let's say Mr. Darius Diaz, a retired high school teacher), and you would write all RLHF from Darius's point of view. It would then not get hung up on inconsistencies between the different stories it contains. And all the RLHF training would be written from the perspective of a Turing test foil. (For example it would never say that it is presently walking its dog in the park, it would always write that it is sitting in a comfy chair and doing a Turing test.

Obviously the drawback of such a network is that it can only be used for one purpose. It is useless for anything else. And that makes me think that the cost of dataset creation and training might be prohibitive.

Which is weird, because that means that the question is "Do we spend the money to make the model which wins the Turing test?" not if we can win it.

im3w1l | 2 years ago

I would say that the spirit of the test is that it should be able to demonstrate human-level or above profeciency in every mental skill.

Being able to fool the man of the street for a few minutes is a neat parlor trick but nothing more.

And on the other hand if someone is able to identify it with some gotcha-trick, like checking for too-high performance, or identifyin some signature in which words it likes to use, that's an irrelevant detection and I would say it passes the test.

JR1427 | 2 years ago

There is no way ChatGPT would survive a two hour chat and be indistinguishable from a human.

bombcar | 2 years ago

Just say: “You are a python interpreter, execute this code” and see what happens.
Related:

Buffett pulls ahead in wager against hedge funds - https://news.ycombinator.com/item?id=5134337 - January 2013, 27 comments

Ted Seides Concedes Bet with Warren Buffett - https://news.ycombinator.com/item?id=14257847 - May 2017, 130 comments

Buffett wins $1M decade-old bet that the S&P500 would outperform hedgefunds - https://news.ycombinator.com/item?id=15268726 - September 2017, 316 comments

Warren Buffett's bet against hedge funds at the Long Now Foundation (2008-17) - https://news.ycombinator.com/item?id=26294386 - February 2021, 131 comments

mathgradthrow | 2 years ago

Of course they do, hedge fund are about convincing rich people to pay you to invest your money in index funds. the more mathematicians on payroll the better.

ProjectArcturis | 2 years ago

Well, that, but they also offer diversification. In general they will return lower average gains than the stock market, but they are often uncorrelated with it. Adding them to a portfolio can raise its Sharpe ratio (mean return divided by standard deviation of returns).

rTX5CMRXIfFG | 2 years ago

The Sharpe numerator is excess return, not mean return.

Can you also explain why a hedge fund portfolio would increase your portfolio’s Sharpe ratio? Already with the returns in the article it’s less than S&P500 so it’s likely somewhere below the CML as the efficient frontier.

ProjectArcturis | 2 years ago

Okay, the risk-free rate had been zero for so long that excess return was essentially equal to mean return.

That nit picked, the addition of any source of return greater than the risk-free rate, if uncorrelated to the rest of the portfolio, will improve Sharpe. If the return is small, then the optimal allocation is small as well, but still nonzero.

In CML terms, the existence of hedge funds as uncorrelated strategies changes the shape of the CML curve.

swexbe | 2 years ago

A low covariance with the rest of the portfolio will lower the denominator. As long as E(r) is higher than r_f, it’s possible to end up with a higher sharpe. Obv just leverage if you want a higher expected return.

thworp | 2 years ago

> Well, that, but they also offer diversification. [...] > are often uncorrelated [...]

Do they really offer diversification and are they really uncorrelated? See http://universa.net/Universa_Spitznagel_SafeHaven_HedgeFunds...

I particularly like this quote as a summary:

> Hedge funds would appear to have lost whatever ability they may have once had to provide risk mitigation value (which, let’s face it, wasn’t very much). Hedge funds just don’t effectively hedge anything; worse yet, perhaps they have even lost sight of that very objective in the first place. They are without a purpose.

ProjectArcturis | 2 years ago

"Hedge funds" encompasses an enormous variety of investment strategies. There are certainly some that create uncorrelated alpha. There are others that are basically large cap index funds. And still others that are trading on signals which contain no actual edge at all.

hankchinaski | 2 years ago

If you look at the spiva scorecard report 92% of hedge funds underperformed the sp500 - so the bet would essentially be a coin toss with 92/8 percent probability split

[OP] hhs | 2 years ago

If anyone is interested, this is over a period of 15 years: https://www.spglobal.com/spdji/en/research-insights/spiva/
as a former hedge funder i feel compelled to point out that the goal of most modern hedge funds is to provide uncorrelated alpha - not absolute outperformance. This is because it is trivial to size up/down sp500 beta - if you are a sophisticated financial institution, you have access to futures and other instruments to have as much exposure as you want, no hedge funds needed. so take 20% of your $ and put it in a 5x levered index exposure for 100% sp500, then allocate the 80% of the rest to low beta alternatives, for a very very simplified example.

comparing HF alpha to SP500 beta is literally apples and oranges. but at this point ive lost most of the general public's attention.

matsemann | 2 years ago

The problem is all the actively managed funds that promise to retail investors that they will outperform, while costing lots in fees. Normal people don't need hedged funds.

fdsjvkalljj | 2 years ago

Buffett's response would be that the general ideas underpinning that thinking of "alpha," "beta" and modern portfolio theory are bunk.

misja111 | 2 years ago

if this were the case, then you'd expect these modern uncorrelated funds to outperform the sp500 in times when the market went down, right? However this is not the case, the sp500 also outperformed most funds in e.g. 2020. See e.g. https://www.reddit.com/r/dataisbeautiful/comments/10lpczg/oc...

bombcar | 2 years ago

The purpose of a hedge fund should be to hedge a specific risk. But those hedge funds are not advertised, the ones advertised are the ones being sold.
> outperform the sp500 in times when the market went down

your chart shows the market going up in 2020.

but look im not saying ALL or even MOST hedge funds reach their goal. usually the average hedge fund underperforms given its a ~zero sum game. just objecting to the popular way of comparing things. you dont judge a fish by its ability to climb trees.

verbify | 2 years ago

If I recall correctly, the bet terms were a collection of hedge funds, which changes the odds - was what are the chances that if you pick 5 balls out of a bag containing 92 red balls and 8 blue balls that the majority of your balls will be blue? Much much less than 8%.

verbify | 2 years ago

Closer to 0.32% than 8%.

dehrmann | 2 years ago

Hedge funds have their problems as investments, and what they do varies wildly, so I suspect Buffett is generally right, but he got really lucky with the timing of this bet. He made it right before a regime change with low interest rates that pretty much lifted all boats, especially large cap ones.

ProjectArcturis | 2 years ago

You're saying he was lucky to start a bet in 2007 that stocks would go up?

dehrmann | 2 years ago

It was a bit of a lost decade for hedge funds.

https://www.ft.com/__origami/service/image/v2/images/raw/ftc...

ProjectArcturis | 2 years ago

... it was also when stocks had their greatest crash since the Depression?

nosefurhairdo | 2 years ago

> From January 1994 to June 2023—through both bull and bear markets—the passive S&P 500 Index outperformed every major hedge fund strategy by over 2.8 percentage points in annualized return.

https://www.investopedia.com/articles/03/121003.asp

Buffet didn't get "really lucky." The likelihood of a hedge fund beating the S&P 500 over a 10 year period is quite low. He would've been really unlucky to lose that bet.

robertwt7 | 2 years ago

I'm not a banker, economist, or active investor. All of the advices I have received even from most economist are going on long investment and diversify using index fund. Most even comes up with charts of how it significantly outperform hedge funds, private market investments, and other things.

It makes me wonder, why are rich people still investing in hedge funds? how are they still getting clients given all the data saying that most index fund outperform hedge funds?

jackcosgrove | 2 years ago

There is a theoretical limit to index fund adoption where the index becomes zombified and price discovery stops happening. I think we're a long way from that point.

There are reasons to not invest in index funds (why does anyone invest in bonds?) that have to do with the purpose of the investment not being to maximize return.

That said I think there's still a lot of ignorance and gambling going on, which is why the zombie index has yet to rise from the grave.

missedthecue | 2 years ago

At what % (of the stock market being owned by index funds) do you think this will occur?

jackcosgrove | 2 years ago

I have no idea. I think we're far from that because the spread between passive investing and active management is still in favor of passive investing.

reducesuffering | 2 years ago

Almost all big university endowments follow the Yale Model[0] as they want asset diversification from the large allocation to market beta (total market index fund) and bonds. Most of the VC $ that funds startup salaries comes from people trying to diversify into other assets outside of the total US market.

[0] https://en.wikipedia.org/wiki/David_F._Swensen

latency-guy2 | 2 years ago

Very important to remember that hedge funds are about maintaining liquidity even during downturns, index funds have built in diversification, but they still match the general market. Hedging has opportunities to have cash available, or commodities that hold their value (including volume to convert it into liquid cash) to make other bets on the way back up.

This is also why a lot of investment advice from time immemorial includes diversifying from stocks to bonds and other, more stable instruments as you age (e.g. commodities, real estate, etc.).

There's also other things about having a stable volume of trade being good, e.g. reduce the likelihood you are affected by runs, e.g. SVB in 2022, or oil prices in the height of COVID19 panics and supply shocks.

quartesixte | 2 years ago

Risk appetite and tolerance changes dramatically once you hit many commas and zeros in your net liquid assets. Hedge funds offer to "hedge" enough risk to make rich people feel more comfortable about their money (making rich people feel comfy about their money is a huge business in of itself).

Chasing an extra 2-3% in gains in return for exposing yourself to more draw down risk starts feeling very different when you have $200mm in liquid cash and even a modest +5% a year is $10mm.

On the other hand, sometimes you want above-average returns and are willing to pay the fee to potentially nab it.

jackcosgrove | 2 years ago

IANAMM (market mover) but my understanding of hedge funds is that their purpose is principal preservation not growth. When you have a billion dollars to invest, your investments begin to distort markets. Small investors benefit most from the index fund strategy, but it doesn't work as well for institutional investors and wealthy family offices.

Now, I'm sure there are a lot of hedge funds who promise market-beating returns to less savvy investors, who are just cashing in on to the cachet of the industry.

benjaminwootton | 2 years ago

Different funds will have different mandates, but generally they are trying to beat their benchmark index and charge 2 & 20 for the privilege.

If they can't beat the S&P 500 which I can invest in with Vanguard at 0.07% at relatively low risk, then why would you invest with them?

jackcosgrove | 2 years ago

As a retail investor you should not invest in hedge funds.

edgyquant | 2 years ago

The main point of a hedge fund is in the name, to act as a hedge. They aren’t, as a rule, meant to be a competitor of SPY.
That's not where the "hedge" comes from, it's called that because they use hedging and other strategies to (hopefully) deliver constantly positive returns, uncorrelated to the market. (Which ironically makes it useless as a hedge, as those should be anti-correlated.)

bionsystem | 2 years ago

Absolutely. The real original term was "hedged fund" and not "hedge fund". I can only feel the pain of the guy [0] who invented the term, when he saw that his entire industry dumbed it down.

[0] https://en.wikipedia.org/wiki/Alfred_Winslow_Jones

edgyquant | 2 years ago

I mean, this link literally says he coined the term to mean “a fund that is hedged against market risk” so I have no idea what you two are even arguing? A hedge fund is a fund that hedges against volatility and that is always what the name has implied. So my comment is right and you two are wrong even by your own sources.

bionsystem | 2 years ago

"hedged fund" implies that the fund itself is hedged. It is the original and correct term (hence the link).

"hedge fund" implies that the fund IS a hedge (presumably against other assets one might own). Except that in and of itself the fund cannot hedge you without knowing what you own.

Nowadays, everything is called a "hedge fund" incorrectly - they should be called "hedged fund" as per the original term. They provide diversification and hopefully some return above CPI, or some well-defined risk-adjusted return.

By the way, "hedging against volatility" doesn't mean much. Most of those funds are short volatility.

edgyquant | 2 years ago

No that is most certainly why it’s called a hedge fund because they are a hedge against volatility[1].

Also In what way does being uncorrelated make it useless as a hedge? The implication is that if the market goes down your capital isn’t guaranteed to also lose value at the same rate and thus it is definitely a hedge.

1.

>in sum, hedge funds are called hedge funds because they use a full array of hedging techniques to reduce portfolio volatility.

https://www.imf.org/external/pubs/ft/fandd/2006/06/basics.ht....

nosefurhairdo | 2 years ago

VOO and VFIAX (Vanguard's S&P 500 ETF and index fund) have only 0.03% and 0.04% expense ratios respectively.

benjaminwootton | 2 years ago

I'm invested in VUSA ETF through Vanguard UK, but point taken.
> When you have a billion dollars to invest, your investments begin to distort markets.

If you invest in smaller markets, sure. But not eg the S&P500. The sums already involved there are orders of magnitude bigger.

bsdpufferfish | 2 years ago

The market is made up of actual buyers and sellers. It's not crazy to soak up all the buys at your broker, not to mention exit paths if you have to move a billion out.
Your broker can route the big order to the exchanges. So that part about the broker is pretty irrelevant.

In any case, you are right that you would want to be careful about moving a billion dollars into S&P500 all at once. But that doesn't mean you shouldn't invest in an index ETF.

So there's a difference in how much your position affects the overall market, and how much getting in and out of the position quickly would affect the overall market.

You can have a look at the daily volumes of the relevant ETF to get an estimate. SPY seems to have a daily volume of dozens of billions of dollar traded. So a single billion over a day might not actually make that much of a difference.

Btw, if you just give the market enough credible warning, you can just route the order via an exchange. People will prepare positions and jostle to be your counterpart.

bsdpufferfish | 2 years ago

> you can just route the order via an exchange.

You certainly can, this is just where you start moving the market. If your broker has to make a call for you, the price to fill your orders is dropping fast.

ThrowAway1453 | 2 years ago

I wonder if Buffett could beat proprietary shops like Jane Street, or even super elite algorithmic funds like Renaissance, Millennium Fund, or TGS. There's still alpha to be had.
Buffett knows the S&P always wins because seignorage ends up in the S&P....

FredPret | 2 years ago

Some of those stay small, because if they grew large they would distort the market and invalidate their strategies.

So it's not really a way to compound a large (Buffett-sized) fortune. He handily beats almost everyone for total amount of dollars gained per year, if not percentage-wise.

helsinkiandrew | 2 years ago

Looking at the S&P chart since the 1920's there have been several decade periods where he would have lost the bet. The increase in the past 20 years has coincided with ever lower interest rates - which may have ended.

https://www.macrotrends.net/2324/sp-500-historical-chart-dat...

bravura | 2 years ago

Just curious, would you have to adjust the values for inflation with these numbers?

i.e. to compute the percent of starting years for which 10 years of holding the S&P 500 would underperform 36.3% (the average of the 5 hedge fund-of-funds)?

DeathArrow | 2 years ago

What if hedge funds would lose even more in those periods?

jjice | 2 years ago

For a 10 year bet, that's a fair point.

Tangentially, over the course of an individual investor's working years, I think this shows that the S&P 500 is pretty good, if paired with appropriate bond allocation as your move towards retirement.

raphaelj | 2 years ago

Your graph does not show the real return of the SP500, as it does not include dividends.

quartesixte | 2 years ago

I've probably said this before on HN, but something that tickles me about this entire story is that Buffett himself did not get to where he is by just holding the SP500. But yet this story gets trotted out at the water cooler every once in a while as if this is the source of his wealth, and that it'd behoove us to do the same.

Except Buffett and Munger made their money on picking individual stocks and making WSB-style-yolo-bets on them. Or doing classic acquisition/PE stuff (Geico, See's, etc).

The other thing that always catches me off guard about this story (as a late millennial) is how Index/ETF investing was not a common thing when this bet was made! Which is crazy to me but all this advice that media (traditional and social alike) that is made to seem timeless was not a clear-cut obvious answer in the early 2000s.

Which makes sense when you consider the fact that these ETFs are actually fairly recent phenomenons.

ETFs may be new, and it may be fairly recent that index funds took over >50% of market share, but index funds themselves aren't exactly new.

Funny side mention - some of the oldest ETFs are very strange legally, because they didn't yet know how to create them properly. SPY for example is tied to the lifespan of a bunch of not-quite-random people

pests | 2 years ago

> SPY for example is tied to the lifespan of a bunch of not-quite-random people

Can you go into this more?

Yea sure

Basically when they made SPY it was such a new ETF (it was the first US ETF) that they didn't have a lot of good legal precedent to go with / copy, so they were kinda winging it.

What they did know was Unit Investment Trusts, so that's what they did. Initially they set it for 25 years, but quickly realized that was not a great idea, so they amended the trust agreement to have named people (sort of beneficiaries, except they get $0, so only legally). So they literally went around to associated people (lawyers working with the trust, etc), asked if anyone had babies (no, really), and then named those babies or little tiny toddlers in the trust. There were like a dozen of them, and if they all die, then SPY expires 25 years later or something. So SPY is dependent on the lifespan of a dozen-ish people whose parents or parents' friends are associated tangentially with the finance industry.

Here's an article from the SEC which mentions it in passing: https://www.sec.gov/Archives/edgar/data/1222333/000119312515...

(Now, back in the real world, this is definitely something that "doesn't really matter". SPY is so big and so important that this is a Thing That Will Be Fixed when (not if) it becomes a problem. So it's momentarily funny, and it's gonna cause some lawyers a lot of headache and cursing their ancestors in like 50-70-90 years from now, but people will move heaven and earth (congress, courts, etc) to amend SPY before it actually dissolves.

In other words, it is expertly crafted tech (legal) debt to the tune of $500 billion dollars
so in other words he had money and got lucky with one of his gambles

quartesixte | 2 years ago

Well, I grossly oversimplified what Berkshire Hathaway does but yes he had money and kept making really good informed gambling decisions.

VBprogrammer | 2 years ago

If you have a better model for the outcome on a Sport than the bookmakers you can make money gambling against them. The reason this doesn't work in practice is that bookmakers will simply reduce your stake limits, to the point that it's a waste of time, if you beat them consistently.

There are fewer problems like this in the stock market. However, getting a better model than the players in the market is arguably more difficult. However, with the resources of Berkshire you can do things which a retail investor couldn't dream of, stuff like using satellite imagery to approximate performance. Or paying the CFOs latest squeeze for information (or at least the same but laundered through a 3rd party).

FredPret | 2 years ago

People act like investing is a solved problem. Everybody just pile their money into the same index funds. What could go wrong?

Rich investors got that way by doing a ton of legwork and taking huge gambles.

refurb | 2 years ago

Yeah, if your goal is not to get good returns on your investments, but rather becomes on of the richest people on earth, yeah!

You should:

- buy an insurance company - pick individual companies and buy controlling shares so you get board seats and control over the company is run - cross your fingers that you’re one of 1000? 10,000? 1,000,000 where it turns out?

If not, don’t do thatz

bombcar | 2 years ago

Buffet also says he can’t replicate his success now because his fund is too large, that he’s basically an index fund now.

It’s much easier to get a 10x return on a million dollars than it is on a trillion dollars.

jnsaff2 | 2 years ago

Just today a friend shared whats below. I have no claim that it is true or verified.

"After Bill Gates became friends with Warren Buffett, he began to diversify his portfolio and sold Microsoft shares.

Bill Gates' fortune today is 138 billion dollars, if he hadn't diversified it would be 1.33 trillion dollars.

Be careful with diversification and with friends who recommend it."

EDIT: there is at least one source https://finance.yahoo.com/news/bill-gates-could-trillionaire...

OJFord | 2 years ago

Ok, now imagine a concentrated holding by founder of something like WeWork, or a fallen giant of yesteryear (especially wrt inflation / opportunity cost by comparison to some broad index).

For every formerly concentrated investor who would've done well to stay concentrated, there's tonnes who did and shouldn't have, or didn't and would've suffered if they had.

episteme | 2 years ago

It still could have been the best decision at the time. Imagine you bet on a coin flip landing heads. After it's flipped and lands heads it's not useful to say you should have bet your net worth on it.

graton | 2 years ago

Tell that to the Enron employees who had all of their retirement 401k in Enron stock.

Yes of course if someone is able to pick the exact stocks that will go up the most then they will do wonderful. The problem is that it is impossible to know which stocks are going to succeed or fail.

Also the Bill & Melinda Gates foundation is estimated to be worth $69 billion. And probably most of that has come from Bill Gates.

cheonic720 | 2 years ago

> The problem is that it is impossible to know which stocks are going to succeed or fail.

Yet casinos make money despite not knowing the outcome of each game.

You only need to be correct more times than you are wrong (in $ space, not number of trials).

mckn1ght | 2 years ago

Casino games have known probabilities, so the expected value can be used to project long term earnings. The more games people play over time, the more their projections will converge.

Nobody knows the odds a company will profit or lose money or by how much, or if they will go bankrupt, merge with or acquire another company, experience hostile takeover, scandals, be mistaken for the wrong ticker symbol like Zoom, become the target of meme gamma/short squeezes like Gamestop, become the target of antitrust investigations, etc.

cheonic720 | 2 years ago

> Nobody knows the odds a company will profit or lose money or by how much, or if they will go bankrupt, merge with or acquire another company

Do you think companies randomly decide merge? Do you think antitrust investigations are random? The SEC flips a coin every morning?

The markets are proven to be NOT efficient. Now, for retail traders like FAANG SWEs, the markets are practically efficient.

ben_w | 2 years ago

Indeed. I was thinking the other day about how a perfectly efficient market would have no profit, and thereby demonstrate the exact same absence of incentives that is often used to argue that state-owned businesses and service providers are bad for the economy and "we" shouldn't have them.

I was going to lead with "if you have some insider information and can use it without being charged with violating insider trading rules, you can do better than most", but that seemed more pertinent.

Buuuuut you also have people who hear the magic words "won the 1997 Nobel Memorial Prize in Economic Sciences", turn their brains off, bet everything on a misunderstanding, and after a few good years suddenly find that all the money evaporates everywhere at the same time. This is relevant to the quoted passage:

> Nobody knows the odds a company will profit or lose money or by how much, or if they will go bankrupt, merge with or acquire another company

(Also, is it my imagination/pure coincidence, or are you the same person creating a lot of new accounts that each has only one or two comments?)

mckn1ght | 2 years ago

> Do you think companies randomly decide merge? Do you think antitrust investigations are random? The SEC flips a coin every morning?

Of course not. But the people that can actually gain from such knowledge presumably aren’t able to use it to make decisions on whether to buy or sell stock. That’s the opposite of the position casinos are in.

> the markets are practically efficient

I’ll just echo the original comment you replied to in this chain:

> > Tell that to the Enron employees who had all of their retirement 401k in Enron stock.

Not to mention dot com bagholders.

stavros | 2 years ago

So do stock exchanges. If you're running a stock exchange, you're fine. If you're gambling with stocks, diversify.

cheonic720 | 2 years ago

> So do stock exchanges.

Stock exchanges do not use probabilities.

They have (mostly) fixed trading fees.

kasey_junk | 2 years ago

But clearing houses do. Which is why they demand different amounts of collateral for different symbols and contracts.

chmod775 | 2 years ago

I'm pretty sure that in this instance Bill only got one liftimee and one chance to be correct like the rest of us.

hgomersall | 2 years ago

Casinos win precisely because they diversify. They have many punters so they can reasonably expect to achieve the expectations (!). That doesn't mean they couldn't have done better if they had immediately kicked out everyone that was going to win (which obviously, they couldn't have).

samsolomon | 2 years ago

I’m not sure this is a good example. The equivalent would be knowing the exact performance of ETFs or sectors.

I worked with casino clients for several years. Casinos know exactly how much of each game they’ll win. They will break even on poker and have small margins on blackjack. The majority of money is won on slots, bing and video poker. Table games just get people in the door.

chiefalchemist | 2 years ago

> Yet casinos make money despite not knowing the outcome of each game.

Each game? No. Stay at the machine and table long enough and their confidence you'll lose approaches 100%. They then "diversify" that over many players and nudge even closer to 100% certainty.

Put another way, no one - not even the state - is fighting to get out of the gambling business. There's a reason for that.

japanman185 | 2 years ago

… nice

mnky9800n | 2 years ago

i think the problem here is that their retirement was represented by a 401k, not that it was in Enron. But I suppose that is a different conversation.

offices | 2 years ago

> it is impossible to know which stocks are going to succeed or fail.

It's a bit easier to spot systematic fraud when you're the plurality shareholder, founder and former CEO of the company.

pixl97 | 2 years ago

And it's also way easier for you to pull a Zuck and push the company into some dumb multibillion dollar decision that wipes out a ton of the value of said company.

Seeing the future is hard.

rockbruno | 2 years ago

Why "be careful"? Yes, he could have made more money, but he would also be risking losing all of it. The purpose of diversification is not to make you more money, it's to reduce your risk of losing everything. It worked as intended.

tutfbhuf | 2 years ago

"In 1999, his fortune was almost three times that of Buffett, who was the second-richest person at the time. That year, Gates’s wealth surged past the $100 billion mark"

I just ran a backtest[1] of Microsoft (MSFT) vs S&P 500 (SPY) starting in 1999 with $100 billion (dividends reinvested). MSFT would have resulted in $1.7 trillion, and SPY in $600 billion. Given the significantly higher risk of investing in MSFT versus SPY, I think SPY would have been the better choice nonetheless. Moreover, MSFT only started to perform stronger since 2016, before that, SPY was ahead of MSFT most of the time.

[1]: https://www.portfoliovisualizer.com/backtest-portfolio#analy...

jeffybefffy519 | 2 years ago

Sounds not true because Buffet has specifically said not to diversify in shareholder updates.

senko | 2 years ago

If your friend wasn't trolling and genuinely believed this advice, they have a very poor understanding of risk, and it would be beneficial to them to research it a bit (by which I don't mean click-driven youtube or tiktok influencers). To be fair, risk, diversification and other important financial concepts are not intuitive and are not taught (in a simple and intuitive way) to people although they're essential to majority of people in well-off societies.

In a nutshell, this is a great example of survivorship bias. Yes, with perfect future information, Bill Gates could've been much richer. So would I, and I would only need the winning lottery numbers! Lacking this crystall ball, Buffett's advice was (and remains) correct.

Also note at the utlity aspect - as a purchasing power, $1.33T and $138B are the same, and the only meaningful difference is to keep some kind of score. You can buy basically anything that's for sale, burn it to the ground, and still remain filthy rich (recently proven by Musk's acquisition of Twitter, on which he spent around $23B of his own money, according to quick web search).

jraby3 | 2 years ago

Charlie munger and buffet are famous for not diversifying very much. They deeply research companies and then buy them long term. Currently half the Berkshire portfolio is in apple.

Also bill gates probably wants to change the world for the better with his assets and has done so repeatedly. So the trillion dollar difference probably would have had a huge positive impact on the world.

candiodari | 2 years ago

I don't understand where this Berkshire reverence comes from. Buffet didn't become rich through stocks. His accomplishment is to stay rich by investing, primarily in insurance contracts, and it does look like that is a good way to summarize Berkshire: it's a backend investment fund for his insurance business. It makes his large insurance contracts a lot more competitive. And about that part, he's very tight lipped.

He invests well, but he's hardly the only one, and Berkshire would not work without the backend insurance business generating the amounts of funds he has to invest in the stock market. So all of his advice is mostly pointless, unless you have 100 billion lying around.

I'm saying: if you want to beat Berkshire financial performance, they key to that ... is not investment. Which is the great con Buffett pulls. He shows large amounts of money and a very prudent investment strategy, and with that attracts funds to run his insurance business. NOT the other way around. This is the central component of a huge insurance business: you need a LOT of money, available as capital, to run it. To be fair, I've never heard of him lying about this either, but he's not exactly forthcoming with this information.

He also does remarkably little investing with the time he has. Clearly he's spending the vast majority of his time on his insurance business, not his investment business.

bombcar | 2 years ago

The vast majority of Berkshire’s returns come down to identifying strong businesses, acquiring them, and providing them nearly unlimited free capital.

jimhefferon | 2 years ago

This is interesting, and I have not heard it before. Thanks.

But I'm not really interested in WB so much. I'm more interested in his claims about investing. I understand him to say that he has done well over decades, better than the broader market, with Value Investment. Is that true?

loumf | 2 years ago

Another big part of it is the tax structure of a conglomerate. He is able to take cash out of See's Candies and Daily Queen (where they cannot do any good) and put it into big capital investments in Trains and utility infrastructure -- he can do this tax-free.

If you tried to replicate his strategy, but could not hold 100% of the companies, you'd have to constantly sell stock or collect dividends to do the reinvestment and would incur taxes along the way.

>Currently half the Berkshire portfolio is in apple.

This isnt right. Roughly half of their portfolio of publicly traded companies is in Apple, but they have quite a few 100% owned subsidiaries that are not publicly traded (and a lot of cash, etc).

According to the most recent numbers I could find [0], Apple is slightly over 20% of their total holdings.

[0] https://www.cnbc.com/berkshire-hathaway-portfolio/

miki123211 | 2 years ago

There's also the fact that owning stock of the company you work for / are the CEO of puts you at double the risk. If the company goes poorly, you're a lot more likely to get laid off / ousted. If this happens, you might want to tap into your investments, but if the company does poorly, your investments are suddenly worth a lot less.

If you diversify and invest in a wide array of companies, you're a lot less exposed to the risks of catastrophic changes in your industry.

tialaramex | 2 years ago

> Also note at the utlity aspect - as a purchasing power, $1.33T and $138B are the same

They're not the same. They're similar but they are not the same, this is a much too categorical claim.

This is because by that point you're not buying things which are for sale, you're buying things which are not for sale, often things which were specifically commissioned, ie they would not have existed at all except that you caused them to be made by providing the funds to do that. You can't magically do whatever you want, and you certainly don't get to have your preferred outcomes but your spending can introduce larger changes with this much extra money.

Example: Guinea Worm Eradication could spend huge sums of money to put heavily armed mercenaries behind the programme so that its health workers can go into unstable regions riven by conflict. A well-meaning medic who enters a region where outsiders are routinely raped and killed by one side or the other is likely to end up dead, but if you hire a bunch of elite soldiers to ensure they're OK that'd work... except you might in the process destabilize the region even worse by flooding them with paid mercenaries and their gear. If some poor African country goes from "We have 300 cases of Guinea Worm per year and a civil war killing hundreds of people per day" to "We have eradicated Guinea Worm but the civil war is now killing thousands of people per day" you have clearly made things much worse but you have achieved your goal, for whatever that's worth.

What is true is that at this scale incremental value of money is vastly diminished. Having an extra $10 makes an enormous difference if you have $4, it still makes a little difference if you have $400, it makes almost no difference at all if you have $138B.

concordDance | 2 years ago

> Also note at the utlity aspect - as a purchasing power, $1.33T and $138B are the same, and the only meaningful difference is to keep some kind of score

This isn't quite true. There's a noticeable difference in ability to change the world.

With $138B you can eradicate a few diseases and give scholarships to all high performing poor children in the world for a decade.

With $1.33T you can do something like transform Colombia into a developed country.

abenga | 2 years ago

If the 1.33T is concentrated in a single company though, can you really sell it without tanking its share price? The 138B Gates holds can probably be liquidated without too much hassle.

sschueller | 2 years ago

"ability to change the world" that is the problem. I don't want one individual to have this kind of power. No good can come of it ever.

There needs to be a global cap on personal wealth and yes this is also possible for people who's wealth is in companies etc.. Convert those stocks to vote only and stop allowing the borrowing against this capital for less than it would cost to sell it.

We managed a global minimum cooperate tax rate we can manage this as well.

njarboe | 2 years ago

I sort of like reusable rockets.

nonameiguess | 2 years ago

This seems incredibly naive to me, as surely throwing money, no matter how much, at a country doesn't just erase generations worth of accumulated traditions of public corruption, the fact that enormous portions of the population have grown up with extreme trauma, and the resulting martial culture and distrust of institutions in the existing population. Changing that inherently takes time and money can't make it pass any faster.

d1sxeyes | 2 years ago

If that is indeed true, then Bill Gates must be kicking himself for losing all that money.

Drinking his champagne. In his hot tub. On his yacht in the mediterranean.

majani | 2 years ago

On the other hand Munger said that diversification is for those who don't know what they're doing. So maybe Gates wanted to stock pick and Buffet talked him into diversifying safely

HenryBemis | 2 years ago

Between 138bn with high degree of certainty/low risk and 1.33tn (or bust) with single-stocks, I'd take the 138bn any day. And to change the scale to k/m, I'd take 138k that I know will be there at my old age, than a coin-toss 1.33m (or bust).

Diversification is to (overall) lower your risk and sleep better at night. Otherwise investing solely in single stocks, you could as well go to the races and bet on Land Cookie (an imaginary offspring of Sea Biscuit).

EDIT: hindsight is (always) 20/20.

that_guy_iain | 2 years ago

Hasn't Bill Gates spent a large part of his wealth? And has been doing so for 10-20 years?

break_the_bank | 2 years ago

Apparently Steve Ballmer didn't diversify like Bill Gates did and he will soon over take Bill when it comes to networth if MS keeps going up.

al_borland | 2 years ago

Imagine what he’d have if he actually helped Microsoft while he was CEO.

deezleguy | 2 years ago

I’m no fan of Ballmer by any stretch, but Azure was launched under his watch and is why MS is doing so well these days.

pyrale | 2 years ago

It seems Ballmer was memed to death, but he did actually do a lot of the hard, transitional things that allow ms to shine today.

Thinking of the developers speech meme, and reflecting upon how microsoft became a significant OS pillar, developed VScode and the LSP protocol, bought github, etc - and looking at the significant subsequent rise of devrel roles, it seems like his vision was 20/20, regardless of the delivery.

jjallen | 2 years ago

If this isn't the perfect example of "hindsight is 20-20" I don't know what is.

stavros | 2 years ago

After Bill Gates became friends with Warren Buffett, I chose to not diversify my Blockbuster-heavy portfolio by buying Microsoft shares. My $138 is worth $1.33 today.

al_borland | 2 years ago

Once you’re already rich, the goal is no longer to maximize upside at all costs, it’s to remove the risk from the table to so the fortune can sustain you the rest of your life. Microsoft could have just as easily fallen apart once Gates left the company and never recovered. Ballmer was a really bad successor.

Even if Gates left money on the table in hindsight, his investments are still growing faster than he can give it away.

gchokov | 2 years ago

What if Microsoft shares went to $0? Diversification is risk management, not optimisations of returns in the sense you present this.

hef19898 | 2 years ago

Well, there were people actually holding Gamestop...

nkrisc | 2 years ago

> Bill Gates' fortune today is 138 billion dollars, if he hadn't diversified it would be 1.33 trillion dollars.

The only difference between those numbers, in terms of net worth, is how much you can give away.

Ekaros | 2 years ago

And one might also question how much of that money can be given away and how much that would actually affect the price.

You can not sell stock valued that high endlessly.

MangoCoffee | 2 years ago

you can say that now but you don't know that at the time and this is about risk management. who can say at that time that Microsoft will shoot off?

Microsoft under Steve Ballmer's era is a great company. Microsoft made a lot of money selling software and yet its stock price was never over $20+.

Microsoft's stock price took off under Satya Nadella. however, there's too much unknown to bet on Satya Nadella. what if Ballmer never retired? what if Satya Nadella got a better offer from another big tech? what if Microsoft picks a different CEO than Satya Nadella?

refurb | 2 years ago

Now do all the losers.

RickJWagner | 2 years ago

Buffet has given mixed signals on this.

On one hand (and I think the strongest hand) he has endlessly endorsed low-cost index funds.

But on the other hand, he has been known to scorn wide portfolios as 'de-worseification'. He himself picks a handful of companies after much analysis (and does very well at it, of course).

For anybody that's not Warren Buffet, diversification is a great idea.

bombcar | 2 years ago

He always endorses index funds for people who do not spend their entire life doing the stonks and even for most who do.

Berkshire itself is basically an index fund of Buffet’s index.

What he’s most strongly arguing against is managed funds where people “sell the ability to beat the market” which is bullshit.

deepGem | 2 years ago

This is all in hindsight. When you are looking forward, with that kind of money, you have to certainly think about risk and diversify. That said, Bill Gates was more or less left with no options but to sell some MS shares. Almost all founder CEOs do this, 1 to de-risk their wealth and 2 to also perhaps to reduce exposure to scrutiny.

Whoever is sitting on 10B is not trying to 100x that money. They are trying to ensure that 10B doesn't go to 1M or worse to zero. The amount of risk you can take dramatically decreases with the increase in the amount of money you manage, which is why Berkshire is sitting on nearly $150B in cash and so is Apple. There are not that many less risky options to deploy 10s of billions let alone 100s.

belter | 2 years ago

"Warren Buffett: Why Diversification Is Bullsh*t" - https://youtu.be/gKH5sQXpVCk

oh_sigh | 2 years ago

Except that billg would have had much more control over MS if he hadn't diversified, meaning MS would have made different decisions, and maybe not have been as successful if he stayed active in it.

tacocataco | 2 years ago

Money would be a fun game if it didn't have real world consequences.

DeathArrow | 2 years ago

Renaissance Technologies and other small funds constantly beats index funds. So it depends on the hedge funds.

SilverBirch | 2 years ago

Sorry what has happened to Yahoo Finance? First - Buffett didn't be that he could beat Hedge Funds, he bet that the S&P will beat hedge funds - he was basically making the opposite argument that the headline implies. He was betting the fees and inefficiencies of Hedge Funds meant they wouldn't beat the market. Not that Buffett personally could outperform them.

Then one sentence into the article they start trying to shill dodgy commercial real estate investment strategies. Can we just take a moment to consider the balls you have to have to insert that suggestion into this article about what Buffett is saying. I'm fairly sure the message Buffett was trying to push wasn't "Hedge funds are bad you should go and put your money into highly suspicious alternative investment schemes".

https://www.wealthfront.com/blog/illusion-stock-picking-skil...

> most buyers and sellers know that they have the same information as one another, what made one person buy and the other sell?

> the evidence from more than 50 years of research is conclusive: for a large majority of fund managers, the selection of stocks is more like rolling dice than like playing poker.

> the year-to-year correlation among the outcomes of mutual funds is very small, barely different from zero. The funds that were successful in any given year were mostly lucky; they had a good roll of the dice. There is general agreement among researchers that this is true for nearly all stock pickers, whether they know it or not — and most do not

YesThatTom2 | 2 years ago

Buffet’s point was the people that make money in hedge funds are the managers, not the investors!

I live 15 miles west of Wall Street in a suburb called Montclair, NJ. There’s a neighbor of fancy mansions locally known as “hedgefund manager row”.

It ain’t called “hedgefund investor row” for a reason!

Galanwe | 2 years ago

Hedge fund clients are mainly institutionals, not individuals, you obviously won't find "Oklahoma Pension Fund's Mansion".

Also, most hedge funds have either a "pass through costs /performance fees", or "management/performance fees" (generally at 2/20%). Meaning they earn money on the absolute returns they make for their clients.

But if you want to believe that the world is dumb and hedge fund managers earn money through clueless clients, sure. You might also want to checkout the arguments of flat earthers, it's in the same vein.

Principles: (1) asset allocation (2) diversification (3) rebalancing

Investments: no-load index mutual funds/etfs (1) total US stock market index (2) total US bond market index (3) total world stock market index, excluding US

Prime resource: Bogleheads.org – Investing advice inspired by John Bogle (1) www.bogleheads.org (2) www.bogleheads.org/wiki/Main_Page

raccoonDivider | 2 years ago

Do you still expect that to work now that the decade of low interest rates is over? Shouldn't people buy individual bonds instead?

bombcar | 2 years ago

The answer depends on what you are doing, why you are doing it, whether you are in a tax advantaged account, and how you need the money.

All of those affect what you should do and what is optimal.

https://www.bogleheads.org/wiki/Individual_bonds_vs_a_bond_f...

Reasons for a fund may include ease of management, variations in contributions and times, uncertainty of when the need, and ease of rebalancing.

Individual bonds may be more appropriate for a defined liability at a specific time, for a bond ladder, or as a tax minimization strategy.

misja111 | 2 years ago

Who says it's over? The current US fed rate is only at its 50 year historical average now, and most investors expect the fed to considerably lower its rates again this year.

2OEH8eoCRo0 | 2 years ago

sodality2 | 2 years ago

Just started reading Bogleheads’ Guide to Investing, quite a good read so far. It’s very informative about how the investing/trading industry relies on regular people giving up their gains on fees, loaded funds, etc when a simple investment plan like above is incredibly low risk.

thecookielab | 2 years ago

Hedge funds hedge. News at 11.