Grateful For Bad News

Speculators were blamed for rising oil prices a few months back, but not for recent falling oil prices.  Short-selling speculators were recently blamed for falling stock prices, and actually banned for a few weeks, but no one proposed banning buying speculators two years ago when stocks were rising.  Now Steven Pearlstein of the Post wants to close financial markets for a week:

What we need to do is to stop making things worse by continuing to over-rely on financial markets and financial institutions that have proven to be incapable of performing their core missions: getting capital to where it needs to go and pricing that capital in a way that reflects the risks and underlying economic values. We have to stop digging. Another week like this one, and there won’t be much left to rescue. 

To begin, the markets could use a timeout just about now, something that lasts longer than a weekend and gives policymakers around the world the chance to get a good nice sleep and evaluate their options without feeling like they have to respond to every movement flashing across their Bloomberg screens.  It would allow some time for passions to cool and for real investors to regain control of markets now dominated by the computerized short-term trading strategies of hedge funds and hot-shot money managers desperate to recoup some of their losses.

I can’t imagine Pearlstein suggesting closing newspapers for a week, or banning them from printing bad finance news for a few weeks.  So Pearlstein doesn’t get it:  financial markets are news institutions, just like newspapers!  The fact that newspapers report a lot less news on this crisis on weekends shows that most crisis news now comes via financial markets.  Don’t blame the messenger for telling you bad news; blame those who caused the bad news, and who keep you from learning sooner.

Aside from times when firms issue stock or buy it back, stock trades do not change a firm’s total capital; they just gives us news about its future prospects.  Sure some of of these stock "reporters" can have incentives to mislead us, but newspaper reporters can also have incentives to mislead us.  Systems for detecting and punishing misleading reporters are far stronger and more effective in financial markets than in newspapers.

Yes both kinds of reporters are human, and can succumb to fear, herd mentalities, and other human biases.  But I see no authority we could trust to ban or overrule news on this basis – better to let those who see our mistakes early be rewarded for saying so within our news institutions. 

Yes financial prices do often effect more than just finance traders, but that is almost entirely because financial prices contain valuable news, and we have set up our institutions to respond to that news in many ways.  If some of these responses are poorly designed, then let’s fix them, but don’t blame the news itself. 

To the contrary, we should want even more similar news institutions to inform us on topics, like public policy, where current financial markets do not speak clearly.  That is the prediction market vision which I’ve been pursuing for two decades. 

It turns out our banking system was in bad shape, and now we are finally learning just how bad.  Instead of sticking our head in the sand to block bad news, we should be grateful to those who have finally told us, be eager to learn more quickly, and be angry with those who kept us from learning sooner.  But don’t ban news, bad or good; we need news now, more than ever. 

GD Star Rating
loading...
Tagged as: , ,
Trackback URL:
  • Will Pearson

    I have read, but do not know the accuracy of, that some institutions are being forced to sell lots of stocks at cut prices in order to raise capital for other debts.

    This does not mean they have information about the stocks being bad, they just need the money. Other people think someone has information they don’t about stocks being bad and so sell before things get worse. Causing a cascade.

    A stock price is a very small bandwidth way of communicating news, and people selling may not be trying to communicate anything about that stock.

    Not that I think that stock trading should be stopped now. Just that I don’t rate it as a source of news. Stock markets aren’t about news. They are about moving money around. Sometimes you get news out of it, other times you don’t.

  • http://hanson.gmu.edu Robin Hanson

    Will, news can be misleading, but still be news. Newspaper articles can also suffer from random errors, cascades, and people with agendas. When speculators know that trades are due to liquidity needs that don’t signal deeper news, they don’t let the price change much.

  • http://occludedsun.wordpress.com Caledonian

    Robin Hanson: Among the psychological aspects of this sort of issue that I believe you are neglecting are: a) the need to find a person to blame for uncontrollable bad things, and b) the need to find someone else to blame for the bad things we are ourselves partially or completely responsible for.

    If you continue to act and think as though the things people say and do are predicated upon rational arguments and coherent reasoning, understanding will forever prove elusive. People do not act out of rational motives, and rational arguments will not induce them to cease acting irrationally.

  • http://yudkowsky.net/ Eliezer Yudkowsky

    I’m not sure if I can endorse this as a statement true of the is-universe rather than the should-universe. Imagine a world where most people held stocks for the dividends – owned stocks because they wanted them for themselves, and not in the hope of selling it to anyone else. Even in this world, stocks would tend to end up efficiently priced, because any large concentration of wealth would have a motive to hire specialists to find the most attractively priced stocks in terms of future dividends. In this world, the markets report the news.

    But what we actually have is people owning stocks in order to sell them, Keynesian beauty contests, loans with stocks as collateral, highly leveraged nodes where a vanishing capital price means that lenders will be unwilling to roll over the remaining debt. And above all, many many nodes – especially intermediary nodes – that borrow short to minimize interest paid and lend long to maximize interest received, which includes (in disguised form) hedge funds that allow redemptions.

    These organizations rely on their reputation – what stock prices report – to prevent market runs on their short borrowing, which creates self-fulfilling prophecies. And if the market is driven by Keynesian beauty contest speculators who are holding the securities in hopes of selling them to someone else, you can get huge cascades…

    …and it does seem to me that with the right kind of bankruptcy laws, what ought to happen in this case is that a lot of speculators and shareholders are ruined, and others wind up with their bonds converted to stocks (long lending revealed as long), but all the actual makers-and-servicers just keep on ticking; while sensible people increase their asset allocation to stocks because dividends just got cheaper. But somehow or other we don’t seem to live in that world – or we’re being taxed because people believe we don’t live in that world, another kind of self-fulfilling prophecy.

    The speculative news-reporting part of the economy does not seem properly coupled to the making-and-selling-actual-stuff part of the economy. It’s possible that the right amendment to the laws governing lending and bankruptcy would fix this instantly, without the need for trillion-dollar bailouts – that we could just go through a Gargantuan string of systemic bankruptcies that enormously reallocated paper wealth while the factories just kept on ticking – but I don’t claim to know what those laws would look like, and it seems more likely that in our real world, all this is just a dream.

  • Grant

    Will,

    Firms selling stock can try to sell it at any price they wish. If the stock is trading at $5 per share, they could try and sell for $10, couldn’t they? The reason no one buys at this price is that financial markets clearly communicate to buyers that they can purchase the exact same shares for $5 from another seller. From where I’m sitting, this seems like a form of “news”.

    I would love a post on where we should turn when markets fail to produce good communication (i.e., when they bubble). Conventional wisdom seems to be that when markets have obviously failed to do something like price houses correctly, we should turn to political mechanisms instead. However, this presupposes that politics can both accurately detect bubbles, and know a more “proper” price during a bubble. Given how insane prices can be during bubbles this seems like it should be pretty easy to do, yet I don’t think its ever happened? Could other market mechanisms be employed to burst bubbles? It doesn’t seem to me that prediction markets would be necissary given that we are dealing with prices of tangible assets?

  • Grant

    Eliezer,

    Keynesian beauty contests are inherently short-term phenomena, so I’ve always wondered what might happen if exchanges had the freedom to restrict or regulate short-term trading (under current SEC regulations I don’t believe they can do this). This would be conceptually similar to giving banks the freedom (of contract obviously, as depositors would have to agree to it) to suspend withdrawals during runs.

    I believe reducing short-term trading is the motivation behind Berkshire Hathaway having never split their share price.

    It seems to me that you’re observation that the “speculative news-reporting part of the economy does not seem properly coupled to the making-and-selling-actual-stuff part of the economy” doesn’t hold in the long-run, unless political responses to short-term problems produce long-term consequences (e.g., the Great Depression). It seems to me that the Internet bubble might be a good example of this.

  • Jef Allbright

    Eliezer: “…but I don’t claim to know what those laws would look like, and it seems more likely that in our real world, all this is just a dream.”

    Take your reasoning one level further, map it onto the “paradox” of superrationality, and the question reduces to something like “is there a rational basis for individual action promoting, in principle (meaning in the limit, with zero knowledge of any *specific* expected utility), a good greater than themselves?”

    Certain insects societies already seem to get this. For reasons related to particular evolutionary trajectories, humans mostly don’t. Closer to home, it’s a little like the greater adaptability of gorilla groups in comparison with the higher individual intelligence of chimpanzees.

  • http://www.bthomson.com Brandon Thomson

    Will: “…Other people think someone has information they don’t about stocks being bad and so sell before things get worse. Causing a cascade…”

    It seems almost obvious that not all investors choose to buy or sell stocks based on good information about the specific stocks, but rather on emotions, unrelated media events, and a host of other things that a proper Bayesian would probably ignore or at least weight less heavily. A lot casual investors seem to have mistaken notions like “stocks have been dropping rapidly recently, so they can only continue to drop” or “stocks are hot right now, so they can only continue to rise”. I know anecdotal reports are pretty worthless, but I certainly know more than one person who has sold a lot of equities recently out of fear and certainty that prices can only continue to fall rather than a solid analysis of the probability that stock prices will never rise again (or even before their time horizon for actually needing the money). I would think this behavior contributes to the current equities price declines in the market.

    All other things being equal, do stock prices tend to be lower than they should be based on fundamental information following events that cause people to be fearful, and higher than they should be following periods of good performance that cause people to be reckless? Warren Buffet reminds us to be greedy when others are fearful and fearful when others are greedy…

  • Will Pearson

    Warren Buffet reminds us to be greedy when others are fearful and fearful when others are greedy…

    I’ve been thinking what is the rational thing to do when you think the market as a whole is acting irrationally. It is not to immediately correct that, it is to wait until the highest or lowest possibly point. It exacerbates the irrationality.

    You don’t want to sell immediately when the market over values the company, you want to wait until just before the market crashes, similarly you don’t want to buy until the price bottoms out, to get the best bargain.

    You are playing chicken with other rational investors, who can wait the longest to buy or sell and signal to the rest of the irrational investors the real price of the company.

    Hmm reading this brief abstract, it looks like the agents that trade on the fundamentals should be evolutionarily stable in the long run.

    Perhaps due to the churn of new inexperienced investors coming into the system, the chartist strategy is still useful.

  • http://pdf23ds.net pdf23ds

    I think a central point here is that whether markets function well or not depends critically on human rationality. (Less so as automated trading becomes more advanced, but computers dominating trading volume is a long way off.) If humans are irrational in a way that leads to specific vulnerabilities to bubbles and such, there’s probably not a good way to prevent all such irregularities. We might be able to patch holes here and there, but there will always be another failure condition.

  • Julian Morrison

    I’m no banker, but… it seems to me that any asset, the primary value of which is expected buy-in by third parties (and not any intrinsic earning potential), is a pyramid scheme. You can pretty it up and tangle it in fancy talk, but in the end the first guys get rich and the last guys get screwed, and then the whole thing blows up.

    So why is selling them legal?

  • http://pdf23ds.net pdf23ds

    Well, selling stock has to be legal sometimes, or else a 50% owner of a company couldn’t decide to let a potential co-owner buy in and share management responsibility.

  • Tom

    Will Pearson:

    >Hmm reading this brief abstract, it looks like the agents that trade on the fundamentals should be evolutionarily stable in the long run.

    This may be a silly question, if so I apologise, but I must ask: what are “the fundamentals”? Where can I find out about them?

  • Will Pearson

    Wikipedia is your friend.

    I am by no means an expert myself.

  • http://hanson.gmu.edu Robin Hanson

    Brandon, newspaper articles are also often based on emotions and unrelated events, rather than good info. Any news system will be.

    pdf, this isn’t a post on the goodness of markets in general, but specifically on speculative trading as a news system.

    Eliezer: “What we actually have is people owning stocks in order to sell them, Keynesian beauty contests, loans with stocks as collateral, highly leveraged nodes where a vanishing capital price means that lenders will be unwilling to roll over the remaining debt. And … nodes …. that borrow short to minimize interest paid and lend long … which creates self-fulfilling prophecies. And if the market is driven by Keynesian beauty contest speculators who are holding the securities in hopes of selling them to someone else, you can get huge cascades.”

    When the existence of a business is tied to news of expectations of its profitability, bad news can kill a business. In general, when other systems are tied into news systems you can have self-fulfilling and self-defeating prophecies. When the success of systems are tied to perceptions of their popularity, you create beauty contests. These are failings of other systems, not news systems. More other systems are tied into financial market news systems than other news systems because they are more reliable and informative news systems. If other new systems were better, they would be move obviously connected to problems too.

  • http://thesaifhouse.wordpress.com saifedean

    This actually reminds me of something I’d been thinking about for a while: the stock market should be open 24/7.

    This is good for two reasons, both of which are a natural outcome of holding the opposite view to the one that Pearlstein expresses in this piece:

    1- It would allow markets to communicate prices more efficiently and at all times. It doesn’t make sense that if a company suffers a massive loss after 5pm on Friday that its price continue to be inaccurately reflected until Monday. Opposite to what Pearlstein would believe, allowing this news to be communicated earlier is better.

    2- It helps people realize how complex markets are, and how trivial their knowledge is. Thanks to humans’ need for sleep, no one will be able to constantly follow and make sense of the market, and people would realize that they cannot understand it, and that having central planners in charge of it would not work.

    So we’d get better market communication of knowledge, and less human overestimation of knowledge.

  • http://entitledtoanopinion.wordpress.com TGGP

    greater adaptability of gorilla groups in comparison with the higher individual intelligence of chimpanzees
    I hadn’t heard of that. But I’ve got a good quote from E. O. Wilson on the eusociality of insects here.

    I think a central point here is that whether markets function well or not depends critically on human rationality.
    The folks at Catallarchy disagree. Sheer idiots could produce equivalent outcomes.

  • Thomas

    Oil prices always fall before an election when a sitting republican is president. Check the stats.

  • http://pdf23ds.net pdf23ds

    TGGP, I’m not going to read that post–but does the author take into account the possibility of systematic, rather than random, bias in rationality?

  • radish

    In theory, Robin, you’re absolutely right (plus I doubt it would help this particular problem to close the markets anyway). However, you could use the same argument to oppose medication for pain. Or preventive anticonvulsants. Or antipsychotics. Those also censor useful signals in the strictest sense, but avoiding them as a matter of unyielding principle is a bit silly.

    Pain and seizures and delusion are “news” in the same sense in which you’re using the term. The problem is that they’re not necessarily “new” news and the side effects associated with allowing them to propagate can’t be controlled without blocking the cascade at its origin.

  • Mikko

    We have computer systems that automatically work on the market price, thus creating a huge automatic negative feedback loop. This has nothing to do with the value of the stock.

    Additionally, from earlier post: “In situations in which a person is not in control, they’re more likely to spot patterns where none exist, see illusions, and believe in conspiracy theories.”

  • Grant

    radish,

    One can censor their pain signals, but its even easier to ignore a market. Plenty of businesses remain private in order to avoid the problems associated with being traded on a public exchange. No one is forced to participate in market exchanges or look at market prices (if they were, it wouldn’t be a market). Admittedly, the regs associated with public companies make it very difficult to revert to private status.

    When pain signals are not longer useful, we discard them. The same should be true of market signals, but how can we know when they cease to be useful? We have to have some other signal to compare them to, and a mechanism to decide between the two.

  • http://hanson.gmu.edu Robin Hanson

    radish, would you even consider censoring newspapers on the medication-blocking-pain analogy?

  • http://profile.typekey.com/Unbathed/ Unbathed

    Is noise news?

  • http://silasx.blogspot.com Silas

    Jef_Allbright: Certain insects societies already seem to [act for the benefit of a greater cause]. For reasons related to particular evolutionary trajectories, humans mostly don’t.

    Are you kidding? Have you even seen how selflessly my liver cells act? And that’s not even the heart! 😉

  • floccina

    Also it is insufficient to say that markets preformed poorly must also show that some other method does better, I do not see evidence of that especially when you examine regulation and Government.

  • Will Pearson

    Also it is insufficient to say that markets preformed poorly must also show that some other method does better, I do not see evidence of that especially when you examine regulation and Government.

    We can say with hindsight after bubbles and the like that the markets weren’t doing well at providing news. This could inspire us to find *better* as yet unknown ways of getting news. Perhaps this will entail an altered market rules, or minimal training in financial analysis or human biases which cause inaccuracies in markets before being allowed to trade in that market.

  • http://www.lebleu.info LeBleu

    Robin,

    If markets are supposed to be news institutions, why did it take them so long to communicate that over-leveraging on CDSes and depending on ever escalating housing prices were bad things? It’s clear that some investors (e.g. Warren Buffet) knew this already years ago. Why wasn’t there some way for these investors to profit from that knowledge in a way that brought bank and insurance stocks down sooner and more gradually?

  • http://hanson.gmu.edu Robin Hanson

    Will, shall we also consider alternating newspaper rules?

    LeBleu, what other news institution did a better job?

  • Will Pearson

    Sure, why not? Or rather I would create different institutions with different rules (news/bloggers/markets ) and see which gave the most informative news.

  • radish

    radish, would you even consider censoring newspapers on the medication-blocking-pain analogy?

    In practice, no. But only because I agree wholeheartedly with this part:

    But I see no authority we could trust to ban or overrule news on this basis…

    I assume that it applies to me as well as to anyone, and look forward to ruling the world, so I can confirm my own benevolence…

    However, for purposes of the argument I’m making, it doesn’t matter whether somebody makes that decision as supreme ruler or the market makes that “decision” by not buying newspapers. I was simply pointing out that your disagreement with Pearlstein is a disagreement about cost, not about principle. It’s true that he might not “get it” at all, but he might also “get it” perfectly well and still disagree with you.

    This is because there is no cut-and-dried, a priori argument for (e.g.) keeping markets (or newspapers) open when the information flowing from them is redundant or highly compressible. The argument for keeping the markets open can be made, but it cannot be made a priori. It can only be based on cost and made in context.

    The question I’m proposing is basically the one Grant asks:

    The same should be true of market signals, but how can we know when they cease to be useful?

    And my answer is that we can’t really know that. Not only that we can’t know as members of a polity, but that the polity itself can’t “know” even if we stipulate that it has agency. The polity (or market) can only guess, and decide, just as an individual guesses that the signal of the pounding headache is safe to treat as noise because it’s important to get some work done, and therefore decides to take some aspirin. In the case of an individual there’s a presumption of unitary agency that conceals a lot of messy neurological and philosophical questions. In the case of a polity, even the question of who “we” are is open to debate, and the decision isn’t a conscious one in any practical sense, so the guts of thye mechanism are visible in all their glory. But the underlying problem, information theory-wise, is essentially the same.

    Even though you can (mathematically) distinguish between noise per se and signal per se there is no a priori way to predict the “relevance” of a given signal. The only way to assess the relevance of a signal is to extract it from background noise and actually process it.

    But the cost of processing can easily exceed the value of the signal itself, particularly if it’s repetitive and compressible. Absent unlimited processing power there is always going to be more information than you can process, and you deal with this by selective attention. By dropping certain signals that you know are signals, without processing them. This is a fundamental problem of information processing, and it matters long before politics or economics enter the picture.

  • http://rhollerith.com/blog Richard Hollerith

    Banking and finance should be boring and effective. If they become the subject of electrifying news stories and all-night emergency meetings high officials, something is wrong.

  • michael e sullivan


    I’m no banker, but… it seems to me that any asset, the primary value of which is expected buy-in by third parties (and not any intrinsic earning potential), is a pyramid scheme. You can pretty it up and tangle it in fancy talk, but in the end the first guys get rich and the last guys get screwed, and then the whole thing blows up.

    So why is selling them legal?

    Because it’s turtles all the way down.

    Just try to come up with an asset, *any* asset, that at it’s core isn’t based on what people will pay/trade you for it (or something it is based on).

    That’s the fundamental definition of value.

    “fundamentals” of securities are still generally fiat money numbers which could become worthless in the event of hyperinflation. Earnings which can evaporate under bad business conditions. And what are bad business conditions? Merely conditions where what people will pay for your service has fallen. It’s all ultimately about whether the asset is worth something, there is no ontological market value.

  • David J. Balan

    All that self-fulfilling prophecy stuff could, in principle, be a good reason for shutting down financial markets for a short time (I”m not saying that is the case here). Robin seems to suggest that believing this requires one also to believe that it would be good to shut down newspapers. This does not follow: we have a lot of deep reasons why the government shouldn’t be allowed to shut down newspapers, even during an emergency, that don’t apply to shutting down the financial markets.

  • http://hanson.gmu.edu Robin Hanson

    David, that would have been an opportune time to mention the “lot of deep reasons.”

  • frelkins

    @David

    “All that self-fulfilling prophecy stuff could, in principle, be a good reason for shutting down financial markets for a short time.”

    No. Historically, shutting the market seems to have only made things worse when it re-opened. For an example we all remember, take the market shutdown after 9-11. There the market closed for a shocking week, supposedly to prevent panic. Yet when the market re-opened, the panic happened anyway: the market dove 7.13% on Sept. 17, 2001.

    Further in this case since the situation is/was global, wouldn’t all major markets have had to shut down at the same time to prevent “panic shift?” Such a co-ordinated shut-down perhaps would make sense only in the context of, say, a united G8 announcement on a unified restructuring or unified regulations.

    Extraordinary measures like closing the market only makes people think things must be much worse than they know, because the market carries so much information. When people suddenly are cut off from information, that’s what’s paralyzing and frightening. Without information people can’t plan, and when they can’t plan, they feel like they don’t have control – that’s what causes fear.

  • David J. Balan

    Robin, the “deep reasons” are merely the standard (but good) ones. A core principle of liberal society is that individual self-expression is both a fundamental right and a key element of limiting governmental abuse. So we’ve put up a high fence around that principle, making it very hard for the government to get away with saying “we need to suppress free speech for a little while, just until things settle down.” And rightly so, even though it may occasionally help a bad situation, we forbid it anyway because we see how easily it could be abused and made into an excuse for plain old governmental control of information. No such core principle is at stake in the operation of financial markets.

  • http://hanson.gmu.edu Robin Hanson

    David, you’ve only talked about deep reasons for “free speech”. In order to explain why we should ban speculative trading but still allow newspapers, we need reasons why the first is not, but the second is, “free speech.” After all, both are primarily news institutions. Why protect some forms of news but not others?

  • David J. Balan

    I think this is just a case where the obvious answer is the right one. There are many examples where real tyranny has advanced itself by doing things like placing restrictions on what you’re allowed to write in a newspaper, so we are (rightly) particularly sensitive to that in a way that we’re not to other things.

  • http://hanson.gmu.edu Robin Hanson

    David, tyrants have and continue to limit the topics on which we can bet. And the fact that Hitler and Stalin did not ban web posts seems a poor reason to be unconcerned about web post bans now.

  • David J. Balan

    I don’t think we’re being careful enough with the word “tyranny.” We would both probably agree that being able to write what you want in a newspaper is a basic right, that it would take a very compelling public interest to justify supressing that right, and that suppressing it without such a compelling interest would in-and-of-itself constitute tyranny. You (I think) would say the same about being able to bet how you want. I would disagree (though not completely), but let’s grant the point for argument’s sake. Then there is the futher issue that suppression of free speech is extra bad because it is particularly dangerous, because it can bring about Hitler/ Stalin style tyranny, which is much worse than the suppression of free speech itself. My claim is that supression of betting rights is not correspondingly dangerous, and therefore does not deserve the same level of protection.

    I don’t get the thing about the web posts. That clearly is just a contemporary version of a newspaper.