Talk:Efficient-market hypothesis

Terrible Analogy

The meteorological analogy is a very bad one and just serves to confuse people who do not have a strong grasp of the markets. Even if every single person in the world tomorrow says it will rain in London tomorrow, that has NO EFFECT on whether it will actually rain or not. However, if every single person thinks Google is worth US$2, guess what happens to the price of the Google? It goes straight to US$2. Regardless of the argument put forth in the following paragraph, about whether the market can be 'right', there is no denying the fact that the market can be influenced by individual's beliefs, while the weather most certainly cannot. I don't see the benefit of this analogy at all and think it should be removed.] -B —Preceding unsigned comment added by 58.152.140.7 (talk) 13:39, 15 November 2008 (UTC)

Change to definition

King Brosby, I don't know if your change here [1] made the definition of EHM more accurate, but it sure made it more confusing. I'd like to fix it but I'm not sure how - can you try? Otherwise I'll do my best. Axlrosen 21:16, 3 Oct 2003 (UTC)

Other things that need fixing:

• We need to explain that the EMH only applies to "efficient" markets, which we need to define.

• "Later work ... found that there were no significant dependences in price changes suggesting that the UK stock market was weak-form efficient." I have no idea what this is trying to say. Axlrosen 21:33, 3 Oct 2003 (UTC)

Thank you for your information. It is very useful to my study. Is it possible that I can ask you a question? for the Low P/E stocks, they tend to have positive return over the long run, do u think that it is consistent with the strong violation of the strong form of efficient market profolio?

Informational efficiency

Some observations on the definition: I would say that informational efficiency only makes sence as an equilibrium concept, i.e. characterising prices (and possibly allocations). This would imply that we are should not talk abour efficient markets, but of informational efficient prices (and possibly allocations). A definition would be something like: A set of prices is informational efficient with regards to a certain set of information if it would constitute an equilibrium even when that set of information was made available to all agents.

I would think that we would then not need a definition of informational efficient markets. But if we were to have such a definition it could be something like: A certain organisation of markets is informational efficient with regards to a certain set of information if all possible equilibria in that organisation of markets (for all combinations of agent characteristics, i.e. preferences, endowments, etc) would be informational efficient in the above sence. --Olejasz 20:40, 23 November 2005 (UTC)

EMH is controversial

The Efficient market hypothesis (and it has always been just a hypothesis) is highly controversial, especially after the stockmarket runup in the late 1990s. There is a significant amount of research that shows that markets vary in their efficiency, and that this depends on market structure and organization. Grossman and Stiglitz in their landmark academic paper "The “Impossibility of Informationally Efficient Capital Markets”" show that liquidity would be zero in efficient markets (Stiglitz is a Nobel prize winner, so is not easily discounted). Successful investors like Warren Buffet catagorically reject technical analysis in favor of fundamental analysis. Though there may still be supporters of the efficient market hypothesis, I would suggest that this hypothesis can no longer be considered mainstream. (Chris Westland, Nov. 2005)

EMH has great validity in weak and semi-strong forms, but is limited by two false assumptions: 1) EMH assumes that new information is assimilated by the markets virtually instantaneously, and 2) it assumes absolute liquidity.

When new information comes into play, there is a time lag while it is being processed. During this time delay, uncertainty about investment risk is relatively high. Overall investment risk includes all forms, including opportunity cost risk. Also, uncertainty about risk is itself risk. These facts mean that risk is volatile, often causing markets to behave in ways that are misconstrued as evidence of irrationality, by Keynesians and other EMH critics. Uncertainty itself seems to be misconstrued as evidence of irrationality, when it in fact should be viewed simply as an inherent inefficiency in the market. (An example of true "irrationality" is to say that because markets aren't perfectly efficient, government should intervene. Despite their imperfections, markets are far more efficient and rational than government at processing information and valuing assets.)

Liquidity issues can be triggered by any number of events, such as a natural disaster or a market slide brought on by new information and the time delay during which the market is assimilating some new information. In the case in which new information triggers it, it goes like this: 1) New information becomes available that will lower the equilibrium price. 2) As the price falls into the target price range of new buyers, these buyers must first liquidate other assets and free up the cash. 3) However, while assimilating the new information, prospective buyers become more "conservative" because of the increased risk associated with the information itself and with the uncertainty. Buyers lower their bid price and the assets become temporarily less liquid. 4) Some asset holders in need of liquidity are compelled to sell. This need for liquidity may even be exacerbated by the falling prices and the impaired liquidity, pushing some sellers into a "motivated" selling mode which can easily be misconstrued as "irrational". 5) Motivated selling coupled with reluctant buying cause a slide until the markets have assmimilated the information, prospective buyers have reset their bid price, AND these willing buyers have converted other assets to cash sufficient to buy the surplus from the motivated sellers. All of this creates the appearance of "irrationality". Kelly J Bailey (talk) 23:34, 3 October 2009 (UTC)97.116.24.66 (talk) 23:32, 3 October 2009 (UTC)

My personal theory

I have a theory that all form of EMH are valid but they are proportionally relevant in regard to the totality of that information within the market, based on the weighted distribution among asset class for example Domestic Large-Cap Growth Style i.e. S&P 500 will demonstrate strong-form of EMH, but if I go into Emerging micro-cap value style - I would expect weak-form of EMH. I have been talking to my finance professors via email correspondence here is a Transcript After reading the transcript, I would like to hear any suggestion and comment that you might have. Paul.Paquette -here's a comment: your link doesn't work and nobody cares about your theory. —Preceding unsigned comment added by 98.185.232.114 (talk) 01:55, 10 December 2010 (UTC)

Bubbles

Sorry if I'm wrong, but I'm pretty sure that market participants have to be RATIONAL, profit maximisers (Fama 1965 Random Walks in Stock Market Prices p76 second paragraph) therefore bubbles are not an example of an efficient market! it's actually the exact opposit (see also the efficient market theory in wikipedia: http://en.wikipedia.org/wiki/Efficient_market_theory where it says the exact opposit as what's written here). 203.217.87.5 00:07, 19 April 2006 (UTC)Emmanuel

Efficient Market theory text (to be merged)

This is the text from Efficient Market Theory which obviously should be merged in this article. Much of it is overlap, but please look to see where some of this can go!

Efficient market theory is a field of economics which seeks to explain the workings of capital markets such as the stock market. According to University of Chicago economist Eugene Fama, the price of a stock reflects a balanced rational assessment of its true underlying value (i.e., rational expectations); its price will have fully and accurately discounted (taken account of) all available information (news).

The theory assumes several things including (1) perfect information, (2) instantaneous receipt of news, and (3) a marketplace with many small participants (rather than one or more large ones with the power to influence prices). The theory also assumes that (4) news arises randomly in the future (otherwise the non-randomness would be analysed, forecast and incorporated within prices already). The theory predicts that the movements of stock prices will approximate stochastic processes, and that technical analysis and statistical forecasting will most likely be fruitless.

no, no, no, no, NO! Strong form efficiency requires perfect information. Other forms assume less perfect information --Thesurveyor 01:11, 3 July 2007 (UTC)

This efficient process of price determination can be contrasted with an inefficient market in which, according to the theory, the pre-conditions for efficient pricing (perfect information, many small market participants) have not been met and prices may be determined by factors such as insider trading, institutional buying power, misinformation, panic and stock market bubbles and other collective cognitive or emotional behavioral biases.

Absence of risk-free arbitrage opportunities

I believe the key requirement for an efficient market is the absence of risk-free arbitrage opportunities. See Richard A. Brearley, Stewart C. Myers, Principles of Corporate Finance. RDSeabrook 19:12, 19 June 2006 (UTC)

a comment by someone named John McGinley

A much adored theory in the 50's, the Efficient Market Theory (EMT), has been conclusively proved never to have existed by Profs. Andrew Lo (MIT) and Craig MacKinlay in "A Non-Random Walk Down Wall St." Their proof, currently available, published circa 1988, took about two years for academics to finally accept. I feel the EMT theory should not be promulgated on your site, except as a dinosaur. The Random Walk we will deal with next.

John McGinley,CMT, Past Director of the Market Market Technicans Assn.

I have met with and discussed this with Andrew at great length. A shame it is not as well known as the proven fact that is. EFM does not exist, and never has.

Your references to Malkin's book are surpflulous and irrelevant. The entire book, given the recent reseach, is not accurate, attestable, credible. Knowing what we know now, it never should have been printed.

I hope this will be published in place of the innacurate info.

John McGinley, CMT, past board member Market Techicians Assn.

a rejoinder -- how very intriguing. Lo and McKinley's work is, indeed, well regarded in academic circles. However, to use it as a total refutation of the EMH is... well... let's be kind and say "shortsighted". Even Fama has of late recognized certain shortcomings in the EMH. Nonetheless, we teach it because it helps us understand how the market works. No one hypothesis captures the entire market, but the EMH has certainly withstood the test of time. As an example, Sharpe's CAPM was "refuted" by the Roll Critique. Nonetheless, we still use the CAPM, we still teach it in nearly every Finance or Investments text, and the majority of investors still use CAPM or some variant to determine required rates of return. It's a very useful model, even if the theoretical underpinnings aren't as solid as once thought to be. --Thesurveyor 01:10, 3 July 2007 (UTC)

Rational and Rational Expectations

For this article, we need to draw a distinction between rationality and rational expectations. It correctly says people do not need to be rational (thinking based on reason, not emotion or habit, which leads to action that achieves goals) for the EMH to work. However, they do need rational expectations (updates their predictions with new information and are not systematically wrong). I will make these changes now. --David Youngberg 23:38, 12 August 2006 (UTC)

My internet clunked out right after I saved the above message, denying me from editing the page. It just came back and I changed the page as I said but I forgot to sign in again when I did. The edits by 68.100.77.213 are me. David Youngberg 03:38, 13 August 2006 (UTC)

A more rigorous definition

By saying that all that is needed for weak-form EMH to hold is that "investors' reactions be random and follow a normal distribution pattern" are we saying that impulses to the system are stochastic and the overall response function is ergodic (or weak-sense stationary, at least)? If so, the mathematical basis of this stuff could probably be enhanced. Just a thought. —The preceding unsigned comment was added by 200.122.159.125 (talk) 03:06, 28 December 2006 (UTC).

A Random Walk Thru This Article

A few random thoughts....

Gene's dissertation wasn't on the EMH, it was on the random walk theory. Gene didn't publish on the EMH until the May, 1970 issue of the Journal of Finance, although he had discussed some of the issues in an earlier (1968) IER paper.

The EMH does not... asserts that financial markets are "informationally efficient", or that prices on traded assets, e.g., stocks, bonds, or property, already reflect all known information and therefore are unbiased. Rather, as the article goes on to make clear, EMH puts forth several different sets of conditions, "strong form", "semi-strong-form", and "weak form" efficiency. The extent to which a pricing market follows one of these forms, (or none of them, for that matter) one can use prices to infer information. --Thesurveyor 01:03, 3 July 2007 (UTC)

EMH is a falsified hypothesis

Why is it the case that this isn't asserted to be false, and listed as pseudoscience? Although economics is hardly scientific, a scientific hypothesis has to be falsifiable. Given that EMH has been falsified over and over and over and over again, then it is a falsified hypothesis. So, it is false. To have it as a criterion for scientific consideration means that it can be useful in understanding how other explanations may diverge from a falsified hypothesis, i.e. they may explain stochastic processes where EMH fails to explain observed data. To propound EMH as a valid hypothesis, after falsification, is only allowable in pseudoscience. —Preceding unsigned comment added by 71.132.135.22 (talk) 22:39, 18 February 2008 (UTC)

Newton's law of gravitation has been falsified by Einstein, too. They are still considered to be scientific, though. A valid hypothesis needs to explain something that other hypotheses can not explain. It does not need to be right all the time. Quantum mechanics is the best tested hypothesis in physics. Still, it fails when it comes to very large energies (relativity, again). Which economic theory is better than the efficient market theory and in what respect? —Preceding unsigned comment added by 84.170.93.163 (talk) 21:20, 17 November 2008 (UTC)

Let us forget binary logic. Like many things, the EMH is a half true / half false hypothesis. Now the question is: what is the bigger half ? --Pgreenfinch (talk) 21:41, 17 November 2008 (UTC)

Let's not forget binary logic. What we should forget is Karl Popper's philosophy of science, which has been largely discredited. Falsifiability is no longer taken as defining science by any serious philosopher. What we should certainly take account of in this article is that EMH is a hypothesis, and many people disagree with it. The section on 'popular reception', as it now stands seems to assume that EMH is true, and discusses why the general public is too stupid to understand it. Alboran (talk) 21:39, 3 April 2009 (UTC)

Too bad, that all serious scientists still credit Popper's philosophy. Should i explain why? Because if theory cant be falsified then it cannot predict anything. And thus if it predicts nothing why do we need it? In this case we just may use the data we acquired from experiment.

As for EMH. Yes, the strong efficiency is on one side the most effective market, on the other side it is non-scientific hypothesis. Because it states that market efficiency is strong if we got all private and public information. But hypothesis doesnt enumerate this information. So, if we stick to EMH and suddenly find out that one of our strongly efficient markets is not so efficient, then we just declare that we didnt have all the information we needed =)... Strong efficiency hypothesis is actually not better then hypothesis that efficient market is run by God =) —Preceding unsigned comment added by 85.223.172.189 (talk) 01:37, 5 January 2010 (UTC)

Flat Earth theory has been falsified, but I still use it for navigation as I drive around. I dig out my map, which is a flat representation of the local geography. I have a mental image of my surroundings which is flat. Even models with known, obvious flaws can be useful as approximations. -- 70.57.21.66 (talk) 11:35, 15 February 2011 (UTC)

Fractals

I think that the view of Mandelbrot that financial time series exhibit a multifractal pattern should be incorporated into the article. See this article: http://www.sciam.com/article.cfm?id=multifractals-explain-wall-street.

Another view on the mathematical pattern of randomness is that of Nassim Taleb. See this article: http://www.edge.org/3rd_culture/taleb04/taleb_indexx.html —Preceding unsigned comment added by Whisky brewer (talk • contribs) 13:15, 31 December 2008 (UTC)

Criticism and behavorial finance

It is clear that the first sentence of the second paragraph is incoherent - It appears like part of the sentence got deleted somewhere along the line, but I can't piece it together. Perhaps someone could fix this. Nwlaw63 (talk) 19:54, 8 January 2009 (UTC)

Criticism section needs better arguments

I really can only come at this from an approach as something of a mathematician but citing five investors who allegedly beat the market on a long-term basis isn't a very strong criticism. First of all the terms "beat" and "long-term" are pretty poorly defined. Second I would expect that any actual mathematical expression of probability (except perhaps where some variable is unbounded) would show that people consistently outperforming the market are simply unlikely not impossible.

Some of the psychological research is interesting but it somewhat strikes me as "criticism du jour" in that, again "short term" and "long term" aren't well defined and that suppose that attempting to take advantage of certain cognitive biases does allow you to make some money but it seems to leave out the possibility that a significant percentage of people could attempt this and negate its effect.

Similar things about bubbles. Again I would expect these to be "rare" not "impossible".

While the section itself needs work, there have been too many papers to mention that prove that EMH belongs in the same place as Flat Earth and the sun orbits the earth. Yes, the argument that a few people beat the indices is silly, as the law of large numbers of course predicts that, but there is no doubt among people who are willing to think on their own, rather than listen to dogma, that EMH is dead. Sposer (talk) 14:53, 27 January 2009 (UTC)

I'm not discounting the possibility that EMH has been falsified (Although I haven't done enough reading to find a formulation that is actually experimentally falsifiable let alone experimental evidence to that effect) if the argument that a few people beat the indices is bad then it is just as bad to keep those in the article as it is to keep the article neutral about EMH if it has been proven false. Considering that someone like yourself who is critical of the EMH finds said arguments as ludicrous as I do makes a good case to remove them. Even if not everyone (here) can agree on the state of the EMH. Also in the interest of Wikipedia being a high-quality source of information. I think it's a good idea to keep popular fallacies out. For example this kind of poor argument shows up online and in that new book by Cooper 'Origin of Financial Crises'. If there are really that many papers disproving the EMH then we are really doing people a disservice by not mentioning them and pulling relevant information out of them and into the article. —Preceding unsigned comment added by 205.211.168.53 (talk) 22:17, 27 January 2009 (UTC)

I've removed the silly paragraph identifying people that have beaten the market. Will look at adding other academic papers later. Sposer (talk) 22:27, 27 January 2009 (UTC)

I'm reading the 'non-random walk' papers now by Lo and the "Macs". Nothing mind blowing so far. I'm kind of excited to see what the arguments are. At the back of my mind I'm still bothered by the poor definition of terms in the EMH like "long-term" which make it hard to falsify. So far virtually all criticism I've read so far seems to mistake anecdotal evidence for counter-example. That is to say that most of the critics seem to think that the EMH is written like it's a logical absolute (therefore only needing a single counter-example to disprove) however from what I've read - even here in the Wiki sounds more like a probabilistic statement. Which of course would require a criticism that is also probabilistic. So far only three possible counter examples come to mind: i) Something like "Any formulation of the EMH requires event X to be bounded at odds R but empirical evidence shows that event X occurs much more frequently than R" ergo no reasonable formulation of EMH can be true. This seems like it's going to be hard since these are large datasets. The other possibility ii) would be to explaining a strategy where you can beat the market indefinitely even when everyone knows how to do so - which seems logically invalid. iii) Is coming up with some internal inconsistency which might be the best path to attempt a disproof. Also people don't seem to differentiate between a strategy and a result. For example if I was granted by the Almighty the ability to perfectly predict stock values, assuming I don't sell or otherwise disclose my picks then even though I am consistently and forever making money off the market I would state that this is still not a violation of most of the forms of the EMH since it is difficult for this information to become what "everyone knows". —Preceding unsigned comment added by 206.248.171.144 (talk) 05:25, 28 January 2009 (UTC)

To further my questions I offer the opinion of J. Farmer [2] a critic of the EMH who co-authored a paper with Lo (one of the authors of non-random) "...the fact that the EMH, by itself, is not a well posed and empirically refutable hypothesis" this is starting to parallel my opinion that it's ignorant to call the EMH falsified. —Preceding unsigned comment added by 206.248.171.144 (talk) 05:39, 28 January 2009 (UTC)

The article probably needs more mention of the fact that the hypothesis has come under a great deal of recent attack from both scholarly sources and news sources.Nwlaw63 (talk) 01:18, 5 January 2010 (UTC)

Unable to locate reference

I am unable to locate the Dreman paper referred to in the sentence, "According to Dreman, in a 1992 paper, low P/E stocks have greater returns. In this paper he also refuted the assertion by Ray Ball that these higher returns could be attributed to higher beta,[17] whose research had been accepted by efficient market theorists as explaining the anomaly[18]:151"

Can anyone?

Normxxx (talk) 02:00, 26 February 2009 (UTC)

EMH assumes random walk based on stable Paretian distribution

Fama paper "The behaviour of stock markets prices" (1965)says random walk based on stable Paretian distribution NOT normal distribution, so incorrect references on the page ...

Popular Reception

This section suffers from both awkward writing and a lot of unsourced opinions. I've tried to clean it up a little bit on both counts. Nwlaw63 (talk) 22:42, 8 May 2009 (UTC)

I'm revisiting the awkward writing and unsourced opinions in this section, if no one objects. Some serious clippers need to be taken to this article, starting here.Nwlaw63 (talk) 05:36, 15 January 2010 (UTC)

All right, I'm taking the axe here. I'm starting with moving or deleting parts of this section that have absolutely nothing to do with the theory's popular reception - in other words, most of the section. Chime in if you have suggestions or disagree, of course, but I think it's clear that 99% of this section (and parts of other sections) are not written the way an encyclopedia article should be written - in addition to the off-topic inclusions, there are unsourced opinions and references to people like Jim Kramer or Peter Lynch that aren't really related to the Efficient Market Hypothesis. Nwlaw63 (talk) 20:30, 28 January 2010 (UTC)

Awkwardness of the Historical Section

Although some improvements have been made, the article still seems to suffer from some awkward writing, and an awkward order of the sections. In the historical section, Firth's research is discussed in terms of being semi-strong-form efficient before that term is explained in the theoretical section below. Perhaps some shuffling of the sections is in order. And that awful sentence leading the last paragraph of the historical section has to go - I'll delete it soon if no one else objects. Nwlaw63 (talk) 17:53, 27 May 2009 (UTC)

Criticism and Recent Events

EMH has taken a severe beating from economists recently, with many of them blaming belief in efficient markets for the current financial crisis. You would never know this from reading the article - I have added a statement in the lede to reflect this. I have started to add corresponding material in the criticism section as well, which I will expand to appropriately reflect current sentiment. Nwlaw63 (talk) 17:39, 12 June 2009 (UTC)

This has remained in the lede uncontested for almost a year until an editor simply removed all of the sourced material with no discussion. I have re-inserted the removed material, pending any other consensus regarding the materials. Given how much the hypothesis has been in the news in regards to the economic crisis, it seems entirely appropriate to mention this in the first paragraph - heaven forbid we should make the theory's oft-mentioned relevance to real-world events into a key part of the lede. I've read the lede policies, and I'm fairly certain this material is important enough to have a place in the article. I'd be curious as to the opinions of others here. Nwlaw63 (talk) 21:13, 9 April 2010 (UTC)

The fact that it has remained there has no bearing on whether or not it should remain there. I removed it with no discussion, because the opinion of a single author clearly does not belong in the lead of the article. I'm not saying that it shouldn't be in the article at all, but just not in the lead. The lead should provide a brief overview of the topic, and should be concise. This person's opinion is not critical to providing an understanding of the topic, and should be included later on in the article, if at all. -- Jrtayloriv (talk) 22:34, 9 April 2010 (UTC)

I'm certainly not wedded to the specific quote removed, but criticism of the theory related to the financial crisis certainly isn't the opinion of one writer - with a little checking, you can find 20 citations for such a view, though the three currently there are probably enough of a representative sample. Many economists are rethinking the theory in light of recent events, and this well-sourced view, which is clearly not a minority view, certainly deserves a place in the lede. That being said, I am satisfied with its current mention at the end of the section. Nwlaw63 (talk) 19:07, 12 April 2010 (UTC)

Statoman's recent addition to the criticism section, while being something I would agree with personally, isn't sourced, and is written from a personal rather an encyclopedic perspective. If no one objects, I will delete or merge into the rest of the article if I don't find reliable sourcing for it. Nwlaw63 (talk) 20:19, 20 December 2011 (UTC)

Survivorship bias

Has anyone scrutinised the data given in the pretty scatter graph of the P/E ratio to see if it is influenced by survivorship bias? If you select a number of companies and go forward 10 or 20 years from year X, then many companies will fail, be taken over, or otherwise disapear. If however you select a group of companies and go back 10 or 20 years, then they will not have failed etc because they are the survivors - you are ignoring the failed companies. 78.146.3.82 (talk) 20:07, 3 September 2009 (UTC)

This question is particularly relevant for this particular graph, because some fraction of the low P/E companies are priced that way because it is widely known that future earnings will decline. In other words, some of these low P/E companies may well fail, and thus unusually subject to survivorship bias. Furthermore, beta analysis that is based on historical betas will not catch this. Eweinber (talk) 05:23, 22 January 2011 (UTC)

Criticism in the beginning

IMO you can't just put a criticism by a "renown financial journalist", that's not credible enough for the main section. It must be removed or changed to a criticism by an academic, and with less agitated wording.

Agree. This article seems backwards. The first section refers to a criticism by a journalist, the last section "Popular reception" refers to a criticism by an academic. We need to get the idea across that the EMH is a fundamental concept still used daily by finance academics and practictioners (almost none of whom believe all versions are 100% true). But we also need to report that there is academic criticism and enormous popular disbelief Expertofsome (talk) 23:19, 3 October 2009 (UTC)

Externalities

Once factored in, externalities render the hypothesis dead and done. 24.36.78.185 (talk) 03:06, 4 December 2009 (UTC)

Excellent article from The Economist about the EMH

This article is a WP:RS for this topic and should be integrated in the text: "Whither the efficient-markets hypothesis?" Jul 16th 2009, from The Economist print edition[3] MaxPont (talk) 23:53, 11 December 2009 (UTC)

Elliott Wave mention

Much as I love Elliott Wave, and even wrote a book on it, I do not think it deserves a paragraph, or even mention, in the EMH article. Elliott Wave is a subset of Technical Analysis, and Behavioral Economics. The market psychology noted in the paragraph is equally relevant to the better known Dow Theory. Although I agree that EMH is 100% invalid, much of the article is already spent refuting it. Certainly, mention of work of Behavorial Economists and technical analysts in general makes sense, but a paragraph on a subset of these larger methods does not really make sense here. Sposer (talk) 18:05, 20 January 2010 (UTC)

Markets are efficient if and only if P = NP

Apparently, it's been proven that the weak form of the EMH is equivalent to P=NP. http://arxiv.org/pdf/1002.2284v1

I prove that if markets are weak-form efficient, meaning current prices fully reflect all information available in past prices, then P = NP, meaning every computational problem whose solution can be verified in polynomial time can also be solved in polynomial time. I also prove the converse by showing how we can “program” the market to solve NP-complete problems. Since P probably does not equal NP, markets are probably not efficient.

A.Prock 05:05, 6 March 2010 (UTC)

It's not clear why the references to the P=NP paper were removed. If there's anyone has a problem with them, please discuss them here.A.Prock 05:27, 11 March 2010 (UTC)

The paper hasn't been published (see WP:SELFPUB). Anyway, a claim this preliminary and outstanding certainly doesn't belong in the lead, and may not even belong in the article, and that's coming from someone who does not believe in even the weak-form efficient market theory. As far as the Algorithmn Design book you added, please provide the page number and a quote supporting the sentence. I looked at the Betting on permutations paper, and I'm not seeing it discuss the issue in question. II | (t - c) 06:05, 11 March 2010 (UTC)

It's in chapter 8 of Algorithm Design. Regardless, it's fairly well established, so I'll dig out another half dozen references or so. That make take a little time, but not too long. I'm not sure what you're missing about the Betting on Permutations paper, it's exactly what the paper is about. I haven't reviewed the original paper yet, but here is another example [4] It's also worth noting that Herbert Simon won his 1978 Nobel Prize for work related to the infeasibility of optimal pricing. I'll look for a cite on that as well. A.Prock 07:21, 11 March 2010 (UTC)

None of the easily accessible sources you've provided so far except the arxiv paper directly connect the P=NP problem to the efficient markets hypothesis. I'll give you a couple days to find sources, but if you don't provide direct quotes I will be reverting. Also, use ~~~~ to sign. Please do sign. II | (t - c) 16:51, 11 March 2010 (UTC)

Sorry about signing, dunno what happened there. To clarify your objection, once the Maymin paper is reviewed and published, including it won't be a problem, correct? Here is a relevant quote from Betting on Permutations

Pair betting, described in Section 3.3, allows traders to bet on the final ranking of any two candidates, for example “candidate D will defeat candidate R”. In Section 5, we show that optimal matching of (divisible or indivisible) pair bets is NP-hard, via a reduction from the unweighted minimum feedback arc set problem.

A.Prock 17:26, 11 March 2010 (UTC)

Notice there's no mention of the efficient markets hypothesis? You're doing original research by saying this paper says the same thing as the arxiv paper. As far as the signature, sorry, I guess you're signing but you disabled the wikilinking of your username (in the my preferences menu). Not sure why you did that; it makes it harder for people to get to your talk page and see how much experience you've had in Wikipedia. II | (t - c) 17:37, 11 March 2010 (UTC)

I'll be happy to pick apart the paper when I've got more time. I take it you'll be fine including this information when the Maymin paper is published? A.Prock 19:03, 11 March 2010 (UTC)

Certainly including in the body in that case, although probably not in the lead. I would also perhaps be OK with including it in the body if a notably published person comments on the situation, even if that person's statements aren't particularly published (e.g. Fama, Shiller). Philip Maymin is not known for his work in this area (he received his PhD in 2007). Would you remove this stuff from the article until you can find a direct connection in a published article or a more prominent author? II | (t - c) 01:41, 12 March 2010 (UTC)

I've reverted the lede for now. I'm going to spend some time looking for more refs, and I'll work on the body to improve it. A.Prock (talk) 02:42, 12 March 2010 (UTC)

The paper in question is of a very poor quality, weather it will be accepted for print and where is questionable. For instance, it says that market is efficient if P=NP, where from the text is clear that it would only be potentialy efficient if this were the case, since efficiency of a market is a statement about actual state, not what would be had all the rules been searched for and used. Also, the definition of a rule is quite problematic - the accidental rules according to this model are easily discovered, and they have nothing to do with efficiency of a market (patterns can be case specific and this sort of "rules" mean nothing, they always exist since there are too many rules of this type, as many as there are market curves in fact, grows in double exponential). The more interesting implication (efficiency of a market implies P=NP) is dubious to say the least, it uses market model that does not even exist, and uses efficiency in the sense of using an offer/oportunity, not in the sense of market price dependence. That is stretching the definition about market efficiency a bit too far. He could as well say: If X offers 10$ for solving particular instance of a NP problem, then if market were efficient, than surely if a solution exists, someone would have taken the oportunity to earn 10$. This is plain ridiculous! Moreover, even if market were somehow able to solve instances of a NP problems, that does not imply P=NP, since this may be done by means other than polynomial time algorithms (say, by using Quantum computers, and it is of course not known if QP=P).

The other citations also report similar findings. I'll replace the text for now while I work on citing more sources. A.Prock (talk) 16:25, 12 March 2010 (UTC)

They do not. I have checked them both. They are certainly better written, but they make no such outrageous claims. They discuss particular problems related to computational complexity of assesing certain types of bets, but do not pertain to efficiency of the market hypothesis. The connection, certainly not made by them, is OR to say the very least, and bad OR since the thing is wrong and can only appear to those who do not know what they are talking about (yet it seems they think they do). —Preceding unsigned comment added by 213.198.212.127 (talk) 18:24, 12 March 2010 (UTC)

I don't think you understand the paper or the EMH. The EMH says that the market asses the bet, assigning the correct price based on all information. See below for a reference which should clear things up. A.Prock (talk) 18:45, 12 March 2010 (UTC)

Yeah right. Didnt it occur to you that it is your understanding that is seriously lacking. Possibility of efficient assesing in polynomial time and actual assesing that takes place on the market are two diferent things. Is that so difficult for your "advanced" understanding? And as I said, in principle idealized market would not be limited to polynomial time machines. As appears from completely different reasonings market as it is now is not efficient, but that does not imply NP!=P. And even the other side of the equivalence is far from proven: market efficiency does not require P=NP, for there might be other ways of real time computing (ever heard of quantum computers?), and noone is discussing EMH implications in the papers you quote, discounting the disasterous sensationalism of the peer review unpublished piece of speculation by Maymin. The best that can be said is that the papers give some evidence against EMH, but that is certainly far from the "equivalence" of EMH and P=NP problem. Your inability to gasp subtle (and not so subtle) distinctions of this sort is appaling. —Preceding unsigned comment added by 213.198.212.127 (talk) 08:09, 13 March 2010 (UTC)

I'm not sure what you're going on about. There are plenty of sources which state the result. I'll refer you (again) to the Betting on Permutations paper, as well as the chapter in Algorithmic Game Theory. When I've got time, I'll write up a more comprehensive description.A.Prock (talk) 16:29, 15 March 2010 (UTC)

I suspect that part of the problem here is that some editors do not understand the predictions markets, the matching problem, and how they relate to financial markets and the EMH. I'm not sure how best to educate them on this, so I'll try to find some less technical sources that might clarify things. A.Prock (talk) 18:17, 12 March 2010 (UTC)

I realize it's a bit advanced, but a good technical review is available in Chapter 26 of "Algorithmic Game Theory", by Nisan, Roughgarden, Tardos, Vazirani. A.Prock (talk) 18:21, 12 March 2010 (UTC)

The Betting on permutations paper does not have "efficient markets" in it. I haven't looked at the other sources, but considering this source, you need to provide a direct quote from these other sources before they're going to be allowed in. II | (t - c) 18:55, 12 March 2010 (UTC)

Yes, those exact words are not used. You have to understand the definition of the EMH when you read the paper. As I said before, when I get a chance I'll get to that. A.Prock (talk) 19:10, 12 March 2010 (UTC)

The Wikipedia policies of original research do not allow us to draw conclusions from sources which the sources don't make explicitly, especially conclusions which require as thought and expertise as this. II | (t - c) 19:24, 12 March 2010 (UTC)

Certainly, but they do allow us to use the definition of words and concepts. A.Prock (talk) 19:36, 12 March 2010 (UTC)

Synthesis

Aprock, in regards to your reorganization of this section [5], you're still engaging in WP:SYNTH. See the definition:

Do not combine material from multiple sources to reach or imply a conclusion not explicitly stated by any of the sources. If one reliable source says A, and another reliable source says B, do not join A and B together to imply a conclusion C that is not mentioned by either of the sources. This would be a synthesis of published material to advance a new position, which is original research.

Now, I'll admit that the last paper by Maymin is not original research, but that paper is currently unreliable. The first two need to be taken out for synthesis reasons, and the third needs to be taken out because it's unreliable. Let's leave this out while the academics put the polishing finishes on it. II | (t - c) 16:52, 15 March 2010 (UTC)

I did no such thing. If you think I did, please provide the specific quotes you think are original research or synthesis and I'll fix them. A.Prock (talk) 16:57, 15 March 2010 (UTC)

The first part of your new section is synthesis:

The problem of algorithmically constructing prices which reflect all available information has been studied extensively in the field of computer science<ref>{{cite book | year = 2005 | author = Kleinberg, Jon; Tardos, Eva | title = Algorithm Design | publisher = Addison Wesley | isbn = 0321295358 }}</ref><ref>{{cite boot | year = 2007 | author = Nisan, Roughgarden, Tardos, Vazirani | title = Algorithmic Game Theory | publisher = Cambridge | isbn = 0521872820}}. For example, the complexity of finding the optimal match in a pair betting market has been shown to be NP-hard.<ref>{{cite journal | author = Chen, Y; Fortnow, L; Nikolova, E; Pennock, D | title = Betting on permutations | journal = Proceedings of the 8th ACM conference on Electronic commerce | volume = 8 | pages = 326 - 335 | year = 2007 | ISBN = ISBN:978-1-59593-653-0}}</ref>. Likewise, it can be proven that the weak form of the efficient market hypothesis is equivalent to the statement that [[P versus NP problem|P = NP]]<ref>{{cite web|url=http://arxiv4.library.cornell.edu/abs/1002.2284 |title=Philip Maymin, NYU-Polytechnic Institute, Markets are efficient if and only if P = NP.}}</ref>

So you're essentially tossing in unrelated facts, and then saying (based on your own understanding) that these are "likewise" to Maymin's results. This is clear original research. Maymin does not cite these sources. If Maymin did cite these sources, or you were to use only Maymin's sources, then this would reduce to just an unreliable source, and it would still need to be removed. II | (t - c) 17:05, 15 March 2010 (UTC)

Could you clarify this. Is it the word "Likewise" that is the problem? The main sentence of the paragraph is that the problem is studied. But I'll use a different word if that helps. A.Prock (talk) 17:10, 15 March 2010 (UTC)

Could you please read WP:SYNTH carefully and slowly? The current section says this:

The problem of algorithmically constructing prices which reflect all available information has been studied extensively in the field of computer science[19][20]. For example, the complexity of finding the arbitrage opportunities in pair betting markets has been shown to be NP-hard.

This section does not say anything about EMH because the sources do not, but it's still synthesis because it is implicitly saying that this is relevant to EMH, which the sources do not say. I'm taking it out. We have to wait until someone publishes the argument. While it seems clear to you, there are still plenty of people who believe in semi-strong EMH - including Nobel Laureates in Economics - so it's not as simple as 2 + 2 = 4. II | (t - c) 05:16, 26 March 2010 (UTC)

I think you'll have to do a better job of describing how that is synth. From the article: "The weak version of EMH supposes that prices ... reflect all past publicly available information." Including a discussion of algorithmic research which describes how to construct prices that reflect all information is not synth. I'll be reverting in a day or so unless you can be clearer about what exactly you think is synth here. A.Prock (talk) 04:13, 28 March 2010 (UTC)

• See Wikipedia:No_original_research/Noticeboard#Synthesis.2Foriginal_research_at_Efficient_markets_hypothesis - opened an WP:OR/N discussion on the issue. II | (t - c) 05:20, 1 April 2010 (UTC)

R

what does "unstable information" mean?

"Defenders of the EMH caution that conflating market stability with the EMH is unwarranted; when publicly available information is unstable, the market can be just as unstable"

??? —Preceding unsigned comment added by Itzkin (talk • contribs) 13:00, 6 June 2010 (UTC)

Recent edits attempting to add Buffet quote

Although, IMO, believing in EMH is akin to believing in the tooth fairy, the example given does not in any way, shape or form, disprove EMH. Buffet's arguments are that the markets are not rational, if I am reading the quote correctly, but that is not enough. However, as the article states elsewhere: "Note that it is not required that the agents be rational. EMH allows that when faced with new information, some investors may overreact and some may underreact. All that is required by the EMH is that investors' reactions be random and follow a normal distribution pattern so that the net effect on market prices cannot be reliably exploited to make an abnormal profit, especially when considering transaction costs (including commissions and spreads)." Sposer (talk) 16:00, 9 June 2011 (UTC)

Who? When?

"Investors and researchers have disputed the efficient-market hypothesis both empirically and theoretically". Should this not read "Some..." rather than the implied "All..."? Researchers and possibly investors set up the EMH in the first place. 92.23.38.244 (talk) 19:46, 21 June 2011 (UTC)

EMH and Keynes

This sentence: "The financial crisis has led Richard Posner, a prominent judge, University of Chicago law professor, and innovator in the field of Law and Economics, to back away from the hypothesis and express some degree of belief in Keynesian economics." should be changed or removed. Keynesian economics has very little to do with EMH, and is certainly is not some sort of competing theory, as this sentence implies. — Preceding unsigned comment added by Kaweron (talk • contribs) 03:08, 2 July 2011 (UTC)

The article states that the crisis led Posner to question both EMH and the Chicago School and move towards favoring Keynsian theory. Maybe the sentence could use some wordsmithing, but it is accurate in regards to the article. It does not say EMH and Keynes are the same. It just questions EMH and Supply Side economics. Sposer (talk) 19:26, 2 July 2011 (UTC)