Hacker Newsnew | past | comments | ask | show | jobs | submit | groundlogic's commentslogin

Wow, I had no idea American was still flying these.

Back in 2013, SAS retired their last MD-80. The last flight was a special "museum flight" with the cabin filled with flight enthusiasts. They flow quite low-height over beautiful places in the Nordics during this flight.


"Unfortunately, our website is currently unavailable in most European countries. We are engaged on the issue and committed to looking at options that support our full range of digital offerings to the EU market. We continue to identify technical compliance solutions that will provide all readers with our award-winning journalism."

How about just disabling the third-party ads along with their invasive tracking until you've figured out a way to monetize the long tail of traffic from EU?


Why would anyone accept an unbounded liability for no benefit?


There wouldn't be any supposedly unbounded liability if they disabled the ads for that region they aren't really targetting anyway.

Also: GDPR hygiene works in the US as well. You've got some way to catch up though; you even allow a lot of really bad food substances that are banned in the EU (and most of the civilized world), just because... profits. It's all kind of a sad states of affairs for a supposedly first world country.


I kinda feel like organizing a program blocking journalists at these american newspapers that are blocking europeans.. from accessing european newspapers/sites. Just because. I'm sure there's a way of figuring out the public IPs for their NATs etc.


Or just use google ads for those regions.

Which probably pays the most anyway because I doubt the Chicago Tribune has a big EU direct marketing department.


"Copy-paste programmers across America are distrought, don't know what to do; suddenly try learning things on their own out of sheer desperation".


Not an economist, but it's obvious to anyone used to thinking in terms of systems that index funds can't work after a certain amount of the money poured into the system is managed by index funds.

What's the limit - 30% 40%, 50%, 60%? What's the current level in terms of managed capital? (Edit: https://www.cnbc.com/2019/03/19/passive-investing-now-contro... says 45% for US stock-based funds, half a year ago, so maybe like 48% now)

I wonder if the endgame is that index funds won't be allowed to trade on the the stock exchanges? I can't imagine how that would be enforced.

"Mr/Mrs/Ms Fund manager, you've been trading too close to the index, we'll be forced to terminate your access to the markets"?


You only need the marginal investor to be informed, so it's not clear that you couldn't have a much higher percentage of passive investment (say 90%) and only a small amount of active investors who are providing price discovery.

The bigger problem is that most passive investments are not really passive - for example, choosing to invest in a "passive" S&P 500 ETF over a "passive" Russell 2000 ETF is an "active" investment choice (preferring large cap over small cap) so valuation errors and bubbles could develop in segments of the market, even if constituents with an index are all fairly valued relative to one another. These valuation errors could sustain for a long period of time, because it takes much more money to correct a valuation error in a huge market segment than it takes to correct a valuation error in an individual stock.


This lists 3575 stocks among its holdings, which includes both large and small cap stocks:

https://investor.vanguard.com/mutual-funds/profile/overview/...


Sure, which is why I said that _most_ passive investments are not really passive. Vanguard Total Stock Market is pretty passive, as long as you consider your investment universe to be "US stocks".

But even in this case the fund only holds stocks (no bonds or real estate) and only US stocks at that (no international or emerging market exposure).


Mind explaining the concept of the marginal investor in this context?


It should be self-regulating, though. The higher the portion of the market that is passively investing, the easier it should be to beat their returns by actively investing so the more incentive there will be to actively invest.


Once indexing gets to be a certain size, you run into the "markets irrational longer than you can stay solvent" issue at a much higher level. Active management "correction" doesn't really work if active managers are a much smaller portion of the market or no longer around at all.


You're only taking into account making money via stock trading. There is always the option of buying a company to actually run it for a profit which is not reliant on the market's belief in the company's value.


Holding doesn't change the price: buying moves it up and selling moves it down. An index fund holding 50% of all shares on the market but not trading them would have no influence at all on prices.


Does this exist, though? If you presume some % of America is putting their paycheck into indexes via Vanguard, Betterment, and some other % is selling due to being retired or whatnot, then this isn't an equilibrium.

There's maybe some room for redeeming index value and "caching" that demand from within Vanguard, etc, but I tend to doubt this action wouldn't hit the market at all.


It can lead to pricing inefficiencies with shares inside vs outside the index. And those should be kept in check by non-indexing value investors.


Did you mean to reply to my post? I don't see the relevance.


An index passively holding 99% of the market would not interfere with price discovery because the remaining 1% would go about its business as if nothing was different. Indexes can't sustain irrational prices because they have no impact on prices.


A single entity holding 99% of anything will absolutely have an impact on liquidity which absolutely impacts price discovery.


> An index passively holding 99% of the market would not interfere with price discovery because the remaining 1% would go about its business as if nothing was different.

That may be true at the point in time where it's already at 99%, but consider the impact on prices as funds were poured into it over time on the way to 99%..


If the funds put in to the index were taken out of mutual funds, there might not be any impact on the prices at all.


There’s an equilibrium to be reached, for sure. The market just hasn’t discovered what it is yet.


I am not an economist either, but it seems that there is a danger in assuming there will be an equilibrium. I think it is entirely possible that there won't be one.

If there is an equilibrium, it may take a significant price shock to discover where it is. I.e. that equilibrium could be years behind us, and if there is a crash we might never recover the value that our current market assumes is there.


> index funds can't work after a certain amount of the money poured into the system is managed by index funds

That's not true. They'll still function just fine.

What will likely change is that they will begin to underperform other strategies, including different types of indexing and active investing.

At that point the market will self-correct and simple indexing will fall out of favor.


Index funds have become successful since they've performed well compared to active investment funds. Why would the active investors suddenly get better at guessing the future?


Active investors are already really good at guessing the future. Index funds work because they follow the decisions made by active investors without needing to pay said investors.

As more of the market moved to indec funds, a smaller amount will be controlled by active investors, which will make the entite market dumber (I would say less efficient, but that would include the cost of managing the fund). As the market gets dumber, an active investor can make more profit without needing to be any better then he currently is.

At some point (in theory), the marginal profit one can make with an active fund will equal the added cost.


I think all available evidence points to active investors NOT being good at guessing the future, when compared to the general consensus. The second half of your argument seems like an interesting opinion, but I'm curious where the evidence is for it.


What do you mean by "general consensus"?

If you mean "current market values", then those are being set by the aggragate opinions of active traders. If there are less active traders, there is less brainpower being devoted to finding this consensus, so it would be suprising if the consensus did not get less accurate.


Isn't the price set by everyone, not just the active traders? It's set each time a sale happens, but the price itself is also related to how many people are holding the stock long term.

I don't really consider it possible for pricing to be "accurate". It is what it is, but accurate implies there's a correct valuation, which I don't think there is.


For financial products, there is a correct price but it cannot be known for certain until far in the future. Stocks in particular represent a claim on the future dividends of the company, whether issued during operation or at the dissolution of the company. If both those future dividends and future inflation were known, you could accurately calculate the present value of that cash flow and get the correct price.


I'm over my toes here, being a programmer and not a finance person, but isn't this viewpoint controversial? IE, does everyone (relatively well informed) agree there are correct prices?


Somewhat.

There is arguably an objectivly true answer for exactly what payments a given stock will make, and so there is an objectivly true answer for what the present value of the stock is (also dependent on other aspects of the future market). This is somewhat of a philosiphical question, largly boiling down to determinism; and is largly moot because no one claims to be able to predict the future well enough for this to work.

Instead, the working position that most take (at least implicitly) is that there is an objectivly correct probability curve of what the future payouts will be, and therefore an objectivly correct probability curve of present values. How to determine what this curve is is a matter of great debate. Further, there is a sufficient lack of objective methodology, that many of the factors that people use in this calculation would be refered to as "opinion", but there is still an objective reality out there.

However, this only gets us an (unknowable) objective probability curve of present values. In general, there is no objective way to turn this into a price. That is to say it is a matter of opinion how much a 50% chance of making $100 is [0].

In an ideal market, you would be able to sell a stock for its objective value at any time. However, "the market can stay irrational longer then you can stay solvent", so you may pay a premium for stocks that you expect to not be undervalued when want to sell them.

Conversly, you may by a stock not because you think it is worth what you are paying, but instead because you expect to find a greater fool to pay you even more then you paid.

There are also cases where people value stocks not just because of their future payments, but because they actually care about the company (or, in the case of divestment, people dont buy them becausr of personal preferences).

In short, there is some matters of opinion in determining a correct price; but most of the disagreement comes from a factual disagreement about what the future looks like.

[0] this calculus changes when you have many such gambles with varying degrees of corralation (and many a financial problem have stemmed from underestimating this corralation)


Do you consider a probability distribution function as objectively true? If it's not obviously true beforehand, I'm not sure if it's only clear in hindsight, which has all kinds of psychological issues in interpretation.

IOW, is there any practical difference between "there is no objectively correct price" and "we'll never know what it is"? The price at any given time reflects the current consensus of the objective price, distorted through the current average psychological lens of the market?

If something is unknowable before it occurs, we will never know what the objectively true measure is until it's occurred. At which point it changes, since the market is dynamic. How could we ever know which point is the correct price?


There is a difference between "truth does not exist" and "we cannot know the truth".

In theory, we can take a now worthless stock and look back in time to determine the actual present value at a given point in the past.

In theory, we can imagine an outside observer running an arbitrarily large copies of our universe from a given point in time to determine the probability curve at said point in time (under whatever model of randomness you want to use). More plausibly, we can take a set of predicted probability curves and look back to see how accurate they were (did events predicted with uncorralated 50% probability happen half the time?).

Economics is hard, because it is very difficult to determine these facts, even in retrospect, but they still exist.


> There is a difference between "truth does not exist" and "we cannot know the truth".

If we can't know the truth, how do we know it exists (in this situation)? I'm not convinced there's an objective true value of a stock -- it seems like stock prices are the general consensus of a huge number of subjective inputs. And that will always be the case.

(Definitely appreciate the time and thought you've put into your responses, btw! Thank you.)


At any given point in time, it's possible to calculate how much value a stock has already given to its bearer via dividends, and adjust those values for the changing purchasing power over of currency over time.

Once a company dissolves, we have enough information to do this calculation over the entire lifetime of the stock issue. At that point, we can determine what real value that stock had at any point in the past for the bearer.

This explains how value investors interpret stock value, but most traders are speculators that expect to make their profit by selling the shares on to someone else. They will only buy a share of stock if they believe that a future investor will buy it off of them at a higher price. This future investor will either be a value investor that expects to get the dividend returns or another speculator that is making the same calculation. Thus, even if a share of stock will pass through many hands before it lands in the portfolio of a value investor, that value investor is the only real price anchor, and the entire chain of speculators are ultimately trying to sell to him/her.


Because when enough of the money is in an index fund, you can predict how a large part of the investors are going to invest (using the same algorithms they're using) and adjust based on that.


That's it exactly. If the index funds get so big that they basically are the market, then active investors will have to adjust their view of the market to be effectively just whatever the index funds do.

It may be possible that although they can't beat other active investors enough to justify their fees, they can beat a big dumb index fund enough to make their services worthwhile. That remains to be seen however.


Kinda makes sense. Still, it sounds like a very fragile system.

So instead if active fund managers "knowing the market better", we'll get active fund managers, "knowing the passive investor crowd better".

This is madness.


I mean, if the majority of the market is dumb passive index funds, those are one and the same :)


I'd assume there's some stability to knowing that X% of the market is passively invested in indexes. It could give active traders some extra edge, since it would make things more predictable? If everyone's an active trader, it's everyone competing against everyone else, where everyone is a live actor. On the other extreme, if you're the only active trader, and literally everyone else is in indexes, you'd be the only person actually moving the market in response to news, and could use that to your advantage.

Obviously we're not in either state, but presumably the closer we get to the latter state, the game gets a little easier for the active traders, not that that necessarily means they'll get a free "win"


>I'd assume there's some stability to knowing that X% of the market is passively invested in indexes.

Indexes holding shares have no influence on the price. The price only changes when traders trade.


To make money you don't need to "Guess the future" in an absolute sense, you just need to "guess" better than the rest of the market participants.

When there is a majority buying/selling all the shares in block without any consideration to the differences in liquidity and fundamentals between stocks it may be easier to find opportunities than when everyone buys and sells individual stocks. (But one could also say that the average stock-picker is so bad that having more active investors actually increases the gain for the talented ones rather than pressuring profits due to competition).


Index funds work because weve been in a 20 year long bull market. If the market goes sideways for a decade, or down for a decade then active investing is alot more profitable.


It's been a ~10 year bull market. 1999-2009 was basically flat.


That has never happened (as far as I know, in modern history). Of course that's not a reason it can't happen, but I feel like you owe us at least a plausible decade downturn scenario.


It has happened for specific stretches of time.

For example, if you bought near the top in 2000, you were still underwater a decade later.


https://www.marketwatch.com/story/john-bogle-has-a-warning-f...

> Bogle pointed out that as indexing increases to a certain point, it opens opportunities for active investors to exploit inefficiencies in the pricing of some stocks. But past that point, wherever it might be — somewhere beyond 75%, in his view — the market could become a dangerous place.


I’m know I’m a dummy when it comes to economics, and an investor in index funds because of that.

But it strikes me that index funds are parasitical in a way and depend on price signals from active investors.

Some people say that it’s ok, the situation is self-correcting.

But what if the smart active money is active in places we can’t see in the public markets? Again, I’m a dummy, but I believe a lot of investment is happening privately these days.


Private investment is absolutely where the money is these days. Look at how the media claims an IPO that doesn't pop 30% or more on day one is a "failure." No one who actually contributed to that company's value benefits from that pop, and best case scenario is really to either have the IPO be flat or even go down a little. But those elite that buy their way to the front of the line demand to have that 30% pop for contributing nothing.

I'll just point out there's the argument about market makers and underwriting and blah blah blah. If that's such an issue just price that into the underwriter fees to begin with. No reason the public markets should lose out on a 30% gain to people that didn't actually take a risk and invest early in the company, and only intend to hold the stock for 8 hours at most.


You are not a dummy. Over the past 30 years, index funds have outperformed active management, especially when you consider the fees.

You are a ”dummy” in the sense you don’t have perfect information awareness on every possibly tailwind or headwind that could impact a particular stock. But everyone is a dummy in that sense.


> But everyone is a dummy in that sense.

And even honest people well versed in economics will tell you that they are too.


> What's the limit - 30% 40%, 50%, 60%?

I'd wager at least 90 percent. Passive investing is generally designed to track active investor activity without effort, so it shouldn't add much inertia to the system. If Dave thinks IBM is overvalued and Under Armor is overvalued, the act of buying and selling will shift those numbers, and the index investors, instead of taking the opposite trade and undoing that flow of information, hold their portfolio.

IMO, the real challenge is active investors competing for access to that 10 percent of active invested money. There's no shortage of people happy to manage money under the 'heads I win, tails you lose' fee structure, and one hopes that the same people fighting over a smaller pool of cash would (more strongly than status quo) favor people who can actually produce results.


>but it's obvious to anyone used to thinking in terms of systems that index funds can't work after a certain amount of the money poured into the system is managed by index funds.

If you have one trillion dollars invested, and the entire exchange volume is based on me and my friend trading a single share back and forth, everything will still work. It doesn't matter how much you own, because my friend and I are going to want a fair price for that one share in any case. There's no practical limit to how passive things can get before a problem kicks up, as long as a few hedge funds stay in.


> entire exchange volume is based on me and my friend trading a single share back and forth, everything will still work

You are talking like such a market is extremely liquid (there are enough shares for everyone). I think when a third person enters such a market your example breaks apart. Now you have one person who constantly wants to buy a stock but is unable to do so. Because you and your friend trade at a fair price and index fund does nothing. So he have to buy at an unfair price and rises his bid until index funds kicks in the game.


> it's obvious to anyone used to thinking in terms of systems that index funds can't work after a certain amount of the money poured into the system is managed by index funds

How so? I can certainly see "lost opportunities" where good stocks are undervalued just because they aren't in the index funds, but I don't understand why you think index funds can't work. Can you elaborate?


Really? I've never been as hung over as the day after I was drinking vodka with my polish counterpart software developers. Granted, they were still living in Poland, so not really immigrants, but they were probably amongst the 5-10% highest paid people in the country.

(Swede here.)


I can tell from experience that this may be due to them graduating from a technical university in Poland.

Young men[0] around here still bond via drinking lethal amounts of vodka. I know of at least one death caused by this.

[0] The gender imbalance in CS is so severe that I remember bragging to my friends how I was not only acquainted with both of the female students in our year at the Warsaw University of Technology, but also both of the ones that started studying CS in 2010 at the University of Warsaw.


On my first day in a new house, I was greeted with a mug of vodka by my (long time settled in Ireland) Polish housemate.

Yeah, that was a tough one to swallow, literally.


Perhaps it differs from country to country, most Eastern European immigrants I know are Jewish people who immigrated in the 90s. But most of the more recent ones are also not really drinkers


It was about the Hong Kong police using a new technology (likely "DNA paint") to mark protestors, likely for future prosecution


DNA isn't mentioned anywhere in the linked story. It only says blue dye and there is nothing new or technological about that.


Note: This story was just artificially moved from the top of page one to the bottom of page two.


Molotov cocktails are generally really dangerous, yes. Have there been any documented uses of Molotov cocktails in these protests?


Yes https://www.gettyimages.hk/detail/新聞照片/protester-holds-a-mol...

However they seem to restrain from throwing it to the police directly. There is only one report of anyone injured by any one of these, by CCTV.


The protesters are smart. I hope they capture (or trace the manufacturer and counterfeit a batch) some of this dye and spray it onto local CCP-oriented politicians and police officers.

(My assumption is that the CCP/HK police used a chinese-made clone of something like https://en.wikipedia.org/wiki/SelectaDNA rather than just a random blue paint.)


This is a great idea. Knowing HK, they already have a knockoff cooked up. Water balloons filled with this + balloon launchers would work well to fight back.


Yikes! spraying people with random DNA fragments?

That sounds like a great way to end up with viri/cancer. Plus a lot of the chemicals used in manipulating DNA are extremely carcinogenic.


Do you have any support for the claims you just made? Synthetically made DNA is not going to give people viruses or cancer when physically applied. It’s hard for me to imagine it doing much of anything even if inhaled/consumed. Similarly, the replication and purification of synthetic DNA is trivially easy to purify of any contaminants.


Reddit has been quite dysfunctional for me the past hour or so.


Really? I feel like that's been going on for about 5 years or so now


Yes unable to login, noticed it right aware as the bright white glare blasted through what would otherwise be night mode.


Yes. Popular loaded, but my feed doesn’t. Just did, but was short.


Same. I thought it was my WiFi at first.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: