Is the stock market efficient?

[Dissipate]

I've talked to friends, relatives and other people about the stock market over the years and a lot of the time we usually come to the conclusion that goes something like this: "Yeah, you can make decent money in the stock market if you spend a lot of time researching it." Well quite to my surprise I recently discovered that some well qualified individuals believe that this is actually not true. They are proponents of something called the Efficient Market Hypothesis.

This is an explanation of the hypothesis:

The Efficient Market Hypothesis formally dates from the 1964 dissertation of Eugene Fama, the work of Nobel price winning economist Paul Samuelson, and others in the 1960s. Its pedigree, however, goes back much earlier, to a dissertation in 1900 by Louis Bachelier, a sutdent of the great French mathematician Henri Poincare. The hypothesis maintains that at any given time, stock prices reflect all relevent information about the stock. In Fama's words: "In an efficient market, competition among the many intelligent participants lead to a situation where, at any point in time, actual prices of individual securities already reflect the effects of information based both on events that have already occurred and on events which, as of now, the market expects to take place in the future.

This is from the book: A Mathematician Plays the Stock Market by John Allen Paulos

There is a corrolary to this hypothesis that Paulos goes on to talk about. The corrolary is that the stock market is actually random, and any attempt to determine future stock prices through a method called technical analysis will be futile. The reason why is that if the hypothesis is true then only new information about a particular company can change that company's stock value. New information about that company is of course unknown right up until the time it is released where upon investors will very quickly digest the new information and the stock price will then reflect the new information, so on and so forth.

It turns out that this hypothesis creates a kind of paradox. I will briefly describe it, basically if investors believe the market is efficient they will stop using technical analysis and other methods to try to predict its future price. When they stop doing this they actually cause the market to stop being efficient because the stock price no longer takes into account past information. On the contrary if they believe the market is inefficient they will continue to employ techniques such as technical analysis to try to determine what it will do in the future, which in turn causes the stock price to include past information about the stock and then making the market efficient.

Well in reality it turns out that there is no shortage of investors who believe the stock market is inefficient. There are many stock market analysts and people who spend countless hours cramming data into their computers and trying to model all information pertinent to the stocks they are invested in. I say let them live in their fantasy world, and for heaven's sake don't tell them they are shooting themselves in the foot so to speak. Ooops, too late. There was a book published in the '60s called A Random Walk Down Wall Street by Burton G. Malkiel that propagates the Efficient Market Hypothesis and provides evidence for it. Basically the author's advice is to not invest in individual stocks but rather to invest in index funds that are tied to an entire market index.

I'm quite bewildered and I don't really know what to think of all this, but I will say it kind of makes me glad I am not invested in stocks. I will soon start investing in ETFs that track a particular index, however. If the market is in fact efficient it makes me wonder why there are still so many people out there who continue to employ research methods trying to beat it. Oh well, I guess its kind of like wondering why people gamble.

 

[gherald]

You will not find many people claiming the stock market is efficient in a mathematical sense.

I think that what makes it efficient (and I use this term loosely) is that it allows for human greed.

As for this "Efficient Market Hypothesis", I do not understand what basis it has on reality. At a glance it seems like it might be a nice ideal, but good luck applying it to the Real World (tm).

Index funds are a viable option for someone who does not have the time/inclination to research the market. This is not news.

The bottom line when buying selecting stocks individually is to be "better than average." at doing your research. So if you think you are smarter (luckier?) than 50% of the investors out there, then statistically you have a chance. Otherwise you're just setting yourself up to loose money and will be better of with an index fund.

Mutual funds are another option, but there's a lot of controversy about their management so I'm not inclined to trust them.

 

[f95toli]

I think it is well known that you can not beat the index in the long run, of course you can make money by buying the "right" stock at the right time but in the long run it will average out since you will also lose money once in a while.
AFAIK there are only two exceptions to this rule.

*You know something that no one else on the market knows. This is insider knowledge and is
illegal.

*You are managing for example a pension fund which is large enough the affect the whole market, by clever trading you can make to market go up/down in a way you can decide. This is quite common on small markets.

A good exemple of the second exception is Gerge Soros, he managed a big american pension fund that speculated in currencies. He made a lot of money in 1991 (I think, I am not sure of the year) by trading in small currencies, the end result was that the interest rates in a few small
countries sky-rocketed
(where I lived it was more than 100% for a couple of days and a large part of our national debt is actually due to the this event ,a lot of money was spent trying, in vain, to support the national currency).

 

[Dissipate]

Originally posted by: gherald
You will not find many people claiming the stock market is efficient in a mathematical sense.

I think that what makes it efficient (and I use this term loosely) is that it allows for human greed.

As for this "Efficient Market Hypothesis", I do not understand what basis it has on reality. At a glance it seems like it might be a nice ideal, but good luck applying it to the Real World (tm).

Index funds are a viable option for someone who does not have the time/inclination to research the market. This is not news.

The bottom line when buying selecting stocks individually is to be "better than average." at doing your research. So if you think you are smarter (luckier?) than 50% of the investors out there, then statistically you have a chance. Otherwise you're just setting yourself up to loose money and will be better of with an index fund.

Mutual funds are another option, but there's a lot of controversy about their management so I'm not inclined to trust them.

Well John Allen Paulous, a mathematician says in his book that he does believe in some form of the Efficient Market Hypothesis, so I guess there is some mathematical foundation for his beliefs.

Remember, this is a hypothesis only. I don't think any sane person will claim they know exactly what makes the market tick. However, if you want to know more about it and you want to see the mathematical arguments for this hypothesis I suggest you pick up the 8th edition of the book "A Random Walk on Wall Street".

Furthermore, I'm not sure if it is a matter of how smart you are. Paulos in his book argues that some strings of events are so complicated that it would require an ENORMOUS amount of processing power to make sense of them. Without this ability the events are virtually random. Near the end of his book he describes this concept and how it applies to the stock market. The idea being that no person or computer has the ability to really make sense of the outcomes on the market with current technology. Also, consider the fact that increase in computers and computing power is essentially levelling the playing field among investors. Since they can now use their computers to do all their technical analysis for them there is less brain power required. Plus, the more accurate people's predictions are about the stock market the more efficient it becomes. Its a self defeating proposition to study the stock market unless no one else is doing it, which they are of course.

 

[rimshaker]

There are some important points to remember:

- the stock market is not an 'entity' as most people think of it. It's just people. Think of it as LOTS of people in a HUGE global flea market.

- what makes a market possible is company stock (the goods in the flea market). Therefore, it's all about supply and demand when you boil eveything down to the bone.

To be successful trading for a living, you have to boil it down even further. When you're using technical analysis to look at charts, what you're really doing is studying people's past emotions. And as mortal human beings, emotions are never forgotten (especially when it's money related, true?). It's all about the ability to sense and read people... whether it be a single trader on the other end of a trade, a small group of people, or the large masses.

This is why trading is the hardest profession on the planet, not in terms of manual labor per se.

 

[djNickb]

To be successful at the market you must develop a system for trading and stick to that system religiously, the moment you allow emotion to dictate your actions in the market, you will lose.

 

[rimshaker]

Originally posted by: djNickb
To be successful at the market you must develop a system for trading and stick to that system religiously, the moment you allow emotion to dictate your actions in the market, you will lose.

In general, yes. Just as everyone is different, each person has their own system that suits them best. You can't be religious about it though, there has to be flexibility because market conditions change. You shouldn't trade soley on emotions but it doesn't mean get rid of it. Like I said, everyone's only human and it's how you interpret and perceive emotions (including your own) that gives successful traders the advantage.

 

[Witling]

On slight correction to f95toli's statement that, "*You know something that no one else on the market knows. This is insider knowledge and is illegal." It's the editor in me that makes me do it. Insider knowledge is not illegal per se. Trading on insider knowledge is. Obviousl F95 was contemplating trades based on insider knowledge.

 

[ReiAyanami]

the market is not efficient, it goes from being overbought to being oversold in days

think Nasdaq 5000 an example

 

[Dissipate]

Originally posted by: ReiAyanami
the market is not efficient, it goes from being overbought to being oversold in days

think Nasdaq 5000 an example

The hypothesis accounts for bubbles.

 

[f95toli]

A while back there was a TV-show that wanted to check how good a few well-known stock brokers really were.
The experiment was done in the following way: A goldfish swam around in an aquarium for a few weeks and the reporters traded according to where the fish spent most of its time during a particular day (as far as I remember they divided the aquarium into several zones, each zone representing a company).
The result was that the fish acttually beat more than half of the brokers, simply because the goldfish portfolio followed index.
The amazing thing was that some of the brokers were really bad, they were actually much worse than what you would get by rolling a dice, it was almost as if they were trying to loose money.

 

[Davegod]

Efficient Market Hypothesis is a pure academic hypothesis and does not hold true in the real world. It holds partly true, but partly isnt good enough to allow for the extreme core assumption - that "stock prices reflect all relevent information about the stock", which implies that not only is all relevant information available, not only is it available to all easily and cost free, but that everybody relevant knows and understands it. Fundamental flaw, but enough of it holds true to enough of an extent that the theory is still useful.

In theory, over the long term every invenstor should make the same returns as the index. Clearly on average and in total this is true, but the theory goes that no individual should be able to consistently outperform the market without insider information. Clearly this is not true. The easiest way to disprove the theory is that all information can be interpreted in different ways by different people - as a Accounting "major" I'd say financial info can be presented and interpreted in whatever way you or the CFO chooses.

 

[dkozloski]

On the one hand you have people that invest on the basis of market research and mathematical dynamics. On the other you find people that ignore this altogether and buy on the basis of human psychology. The advice I have gotten from a very rich man was to invest in what you perceive to be a very good consumer product. A good example is Victor Kiam. He moved on the realization he was getting a good shave with his Remington razor and others were likely to do the same. I know other people that made investments based on the usefulness and viability of a product rather than on how it's stock was likely to perform. George Foreman has sold millions of his grills because it is a product that really works and the public knows it. Some investors attack the problem mathematically like a Las Vegas gambler with a winning scheme, others start from the other end by throwing darts at the financial page of the New York Times. Strangely enough they are both about as likely to succeed.

 

[Dissipate]

Originally posted by: Davegod
Efficient Market Hypothesis is a pure academic hypothesis and does not hold true in the real world. It holds partly true, but partly isnt good enough to allow for the extreme core assumption - that "stock prices reflect all relevent information about the stock", which implies that not only is all relevant information available, not only is it available to all easily and cost free, but that everybody relevant knows and understands it. Fundamental flaw, but enough of it holds true to enough of an extent that the theory is still useful.

In theory, over the long term every invenstor should make the same returns as the index. Clearly on average and in total this is true, but the theory goes that no individual should be able to consistently outperform the market without insider information. Clearly this is not true. The easiest way to disprove the theory is that all information can be interpreted in different ways by different people - as a Accounting "major" I'd say financial info can be presented and interpreted in whatever way you or the CFO chooses.

Well, there happens to be weak forms of the hypothesis and strong forms. The weak form which a lot of economists accept is that all relevent public information about a stock is included in its price.

You are right, the accounting info of a firm can be interepreted in any number of ways but we have to remember that there are professionals out there who are going to figure out wtf is going on so help them god. There are droves of people always searching for that thing called an "undervalued stock". In their belief that the market is inefficient they are constantly searching for a stock that other investors have passed up and or overlooked. Does such a stock exist? Possibly, but it will only exist because someone has overlooked it. With all the thousands of "professional" investors out there all searching for an "undervalued stock" you have to wonder if such a thing even exists.

 

[ReiAyanami]

alright, take market makers/floor specialists for example. they can supress price movement by putting a 500 block on the level II quotes and prevent anybody from pushing the price up.

then the exact opposite, a low volume stock. it only takes a few hundred shares and the "price" of the stock swings 10% in a matter of seconds (actually more like 6%, saw this friday). there are no floors so when any real investor (a single investor) tries to sell their shares, the price plummets. and when a single person is interested and wants to buy a thousand shares, it skyrockets because nobody out there has a standing limit sell order

so basically the market is vastly inefficient, market makers have to role of making thing more "efficient" but are also clouded by their profit motive.

 

[dkozloski]

It would seem to me that a really big confusion factor would be the effects of automatic trading, particularly since all the major players are using very similar software. If all the programs react to the same glitch in the same way it would seem that the thing to do is to play the programming rather than accepted market dynamics. It would also seem that the emotion driven traders have the upper hand lately because of the way the market is spasing around in response to world uncertaintanty. If everyone was the prototypical, logic grounded, thoughtful, basics grounded trader none of the reactions to terrorists, disasters, and political upheavels would matter.

 

[Hector13]

Originally posted by: Witling
On slight correction to f95toli's statement that, "*You know something that no one else on the market knows. This is insider knowledge and is illegal." It's the editor in me that makes me do it. Insider knowledge is not illegal per se. Trading on insider knowledge is. Obviousl F95 was contemplating trades based on insider knowledge.

the point is moot, as in its strictest sense, market effeciency covers insiders and predicts that even inside knowledge doesn't help.

 

[Hector13]

Originally posted by: Davegod
Efficient Market Hypothesis is a pure academic hypothesis and does not hold true in the real world. It holds partly true, but partly isnt good enough to allow for the extreme core assumption - that "stock prices reflect all relevent information about the stock", which implies that not only is all relevant information available, not only is it available to all easily and cost free, [b[but that everybody relevant knows and understands it[/b]. Fundamental flaw, but enough of it holds true to enough of an extent that the theory is still useful.

I don't think the theory doesn't require this at all. All that is required is for one investor be rational and have the resources to implement his views. Given the number and size of hedgefunds out there that can buy/short large amounts of stock (by leveraging), this is a reasonable assumption. Don't get me wrong, I am not claming that the market is fully effecient; I'm just stating that you don't need every investor to be "rational" for market effeciency.

This is the same concept underlying no-arbitrage principles for futures/options/etc. pricing. I bet you 99% of investors out there couldn't price an index future -- but you'll be hard pressed to find futures trading significantly away from their "fair value".

 

[zillafurby]

Originally posted by: Hector13

Originally posted by: Davegod
Efficient Market Hypothesis is a pure academic hypothesis and does not hold true in the real world. It holds partly true, but partly isnt good enough to allow for the extreme core assumption - that "stock prices reflect all relevent information about the stock", which implies that not only is all relevant information available, not only is it available to all easily and cost free, [b[but that everybody relevant knows and understands it[/b]. Fundamental flaw, but enough of it holds true to enough of an extent that the theory is still useful.

I don't think the theory doesn't require this at all. All that is required is for one investor be rational and have the resources to implement his views. Given the number and size of hedgefunds out there that can buy/short large amounts of stock (by leveraging), this is a reasonable assumption. Don't get me wrong, I am not claming that the market is fully effecient; I'm just stating that you don't need every investor to be "rational" for market effeciency.

This is the same concept underlying no-arbitrage principles for futures/options/etc. pricing. I bet you 99% of investors out there couldn't price an index future -- but you'll be hard pressed to find futures trading significantly away from their "fair value".

hector on market efficiency

the truth is the market exhibits mean reversion.
that is the more something becomes apparrent the more participants recognise it and evaluate it accurately.
take future earnings, the closer they get to the earnings date the better more and more people can estmate them, also the more predictable they are the more accurately people can estimate them. therefore the closer the investment's price should be price to its intrinsic value.
think of the barn door analogy, some people are good shots in any weather, some are hopeless e.g. mutul fund managers and your average private investor. the closer you get to the barn door and the better the weather the more and more people hit it, even if they arent looking at it when they fire!

 

[Hector13]

Originally posted by: zillafurby
hector on market efficiency

the truth is the market exhibits mean reversion.
that is the more something becomes apparrent the more participants recognise it and evaluate it accurately.
take future earnings, the closer they get to the earnings date the better more and more people can estmate them, also the more predictable they are the more accurately people can estimate them. therefore the closer the investment's price should be price to its intrinsic value.
think of the barn door analogy, some people are good shots in any weather, some are hopeless e.g. mutul fund managers and your average private investor. the closer you get to the barn door and the better the weather the more and more people hit it, even if they arent looking at it when they fire!

i honestly didn't understand any of your post... what does mean-reversion have to do earnings predictability? Are you saying companies' earnings tend to mean revert (ie, a positive earnings surprise last quarter will lead to an earnings disappointment this quarter)? If so, what does the time to the next fiscal quarter end have anything to do with predictions?

 

[zillafurby]

Originally posted by: Hector13

Originally posted by: zillafurby
hector on market efficiency

the truth is the market exhibits mean reversion.
that is the more something becomes apparrent the more participants recognise it and evaluate it accurately.
take future earnings, the closer they get to the earnings date the better more and more people can estmate them, also the more predictable they are the more accurately people can estimate them. therefore the closer the investment's price should be price to its intrinsic value.
think of the barn door analogy, some people are good shots in any weather, some are hopeless e.g. mutul fund managers and your average private investor. the closer you get to the barn door and the better the weather the more and more people hit it, even if they arent looking at it when they fire!

i honestly didn't understand any of your post... what does mean-reversion have to do earnings predictability? Are you saying companies' earnings tend to mean revert (ie, a positive earnings surprise last quarter will lead to an earnings disappointment this quarter)? If so, what does the time to the next fiscal quarter end have anything to do with predictions?

mean reversion: tendencies for prices to trend towards a mean over time. eg commodity prices tend to oscilate around a broadly 'economic' price.

also lets say that company x's earnings are:

04: 10c
05: 12c
06: 16c
07: 20c
08: 15c
09:05c
10:31c
11c:35c

firstly you have an average growth rate - despite good and bad years there will be a mean growth rate over the cycle. this leads to an intrinsic value. which you try and estimate in advance.

secondly, due to investors in-ability to either see further than a few quarters, see beyond bad quarters, see through bad management, (which is to be replaced), and other reasons, the market price is likely to oscilate around the intrinsic value. soro calls this process reflexivity.

either way whether the market price is over or under the Intrinsic value the market price will tend to mean revert to/ around the intrinsic val over time.

 

[Hector13]

Originally posted by: zillafurby
mean reversion: tendencies for prices to trend towards a mean over time. eg commodity prices tend to oscilate around a broadly 'economic' price.

also lets say that company x's earnings are:

04: 10c
05: 12c
06: 16c
07: 20c
08: 15c
09:05c
10:31c
11c:35c

firstly you have an average growth rate - despite good and bad years there will be a mean growth rate over the cycle. this leads to an intrinsic value. which you try and estimate in advance.

secondly, due to investors in-ability to either see further than a few quarters, see beyond bad quarters, see through bad management, (which is to be replaced), and other reasons, the market price is likely to oscilate around the intrinsic value. soro calls this process reflexivity.

either way whether the market price is over or under the Intrinsic value the market price will tend to mean revert to/ around the intrinsic val over time.

I have no problem with the concept of mean reversion -- I just don't see how it applies to market effeciency (at least not in your argument). If anything, what you are saying supports market effeciency.

Maybe you could be a little more clear in your posts (or maybe I am just too dense after a weekend of work)... are you supporting the market effeciency theor or saying it's wrong (or neither?).

 

[zillafurby]

Originally posted by: Hector13

Originally posted by: zillafurby
mean reversion: tendencies for prices to trend towards a mean over time. eg commodity prices tend to oscilate around a broadly 'economic' price.

also lets say that company x's earnings are:

04: 10c
05: 12c
06: 16c
07: 20c
08: 15c
09:05c
10:31c
11c:35c

firstly you have an average growth rate - despite good and bad years there will be a mean growth rate over the cycle. this leads to an intrinsic value. which you try and estimate in advance.

secondly, due to investors in-ability to either see further than a few quarters, see beyond bad quarters, see through bad management, (which is to be replaced), and other reasons, the market price is likely to oscilate around the intrinsic value. soro calls this process reflexivity.

either way whether the market price is over or under the Intrinsic value the market price will tend to mean revert to/ around the intrinsic val over time.

I have no problem with the concept of mean reversion -- I just don't see how it applies to market effeciency (at least not in your argument). If anything, what you are saying supports market effeciency.

Maybe you could be a little more clear in your posts (or maybe I am just too dense after a weekend of work)... are you supporting the market effeciency theor or saying it's wrong (or neither?).

i think price searcher is the key, participants search for the best investments and find them based on skill. therefore the market is competetively priced, but not correctly priced.

 

[glugglug]

The efficient market hypothesis is not quite correct. The paid market analysts and banks rating the various stocks tend to completely lack any sense, yet they pretty much control the stock prices. The trick is to predict their hairbrained analysis. That's not as impossible as it sounds. There are a lot of predictable events, for example the dates of earnings reports are known. If you are familiar with the products a company makes and the current market for them, news about yield situations, etc., you can make educated guesses as to their quarterly profits and see if it matches analyst consensus, buying before the reports if the analysts are being too conservative (and the stock price doesn't seem to already have adjusted; sometimes it does despite the estimates). I find that for tech companies, you can generally get a better idea of where the company is headed reading boards such as this, and tech news sites like Inquirer and the Register than the analysts have.

If you think about the profit margin changes on a company's products, its pretty easy to beat the market actually. For example the recent rise of AMD stock with their change to the server market segment was totally predictable. Memory stocks are all about to go up because of the price increases with the current "shortage", especially Micron and Infineon who are benefitting from the ability to fix prices from the current tarriff situation.

 

[Dissipate]

Originally posted by: glugglug
The efficient market hypothesis is not quite correct. The paid market analysts and banks rating the various stocks tend to completely lack any sense, yet they pretty much control the stock prices. The trick is to predict their hairbrained analysis. That's not as impossible as it sounds. There are a lot of predictable events, for example the dates of earnings reports are known. If you are familiar with the products a company makes and the current market for them, news about yield situations, etc., you can make educated guesses as to their quarterly profits and see if it matches analyst consensus, buying before the reports if the analysts are being too conservative (and the stock price doesn't seem to already have adjusted; sometimes it does despite the estimates). I find that for tech companies, you can generally get a better idea of where the company is headed reading boards such as this, and tech news sites like Inquirer and the Register than the analysts have.

If you think about the profit margin changes on a company's products, its pretty easy to beat the market actually. For example the recent rise of AMD stock with their change to the server market segment was totally predictable. Memory stocks are all about to go up because of the price increases with the current "shortage", especially Micron and Infineon who are benefitting from the ability to fix prices from the current tarriff situation.

I don't buy it. You say that you should look at tech companies and try to predict their earnings and if their future earnings are going to be more than they analysts predict then the stock will go up? How could you predict the earnings better than the pros? I don't think you will be able to just by searching Internet chat rooms and message boards. Analysts aren't dumb, they probably surf those boards more than anyone else.