Stock market a random walk?

**Gerenuk** · Dec26-09, 09:16 AM

My impression from looking through the internet is, that it is a well-accepted opinion that stock market calculations are based on the random walk hypothesis.

A friend of mine says otherwise. Could you provide some opinions or rather references to prove one view or the other?

**mugaliens** · Dec26-09, 12:24 PM

It's based on the "efficient market" hypothesis. A lot of people believe this holds true. However, if it were true, a stock's value would only change either proportional to the market as a whole, the market segment, or the company itself.

Instead, we find stocks rising and falling all the time for no apparent reason. I say "apparent" because they are rising and falling for some reason!

That reason is rapidly changing demand for that company's stock. The problem is that the demand variance far exceeds intrinsic value changes in the market, market segment, or the company.

The $64 Trilion Question is why?

The fluctuations beyond value changes appear random, but they are most certainly not random, nor are they pseudo-random.

There are three primary factors affecting the market as a whole, the market segment, and the company in particular:

1. Market (and market segment) optimism/pessimism. Market optimism in the 1990s, particularly in dot-coms, was rampant. As a result, it overinflated stock prices in the dot-com sector, which tended to bleed over and overinflate prices in other sectors. Throughout most of this century's first decade, and more sharply, from about 1.5 years ago to 6 months ago, pessimism has depressed the market as a whole, but things are back on the upswing in recent months.

The savvy investor can adjusting a stock's price as compared to the overall market and discern a truer idea of the stock's value.

2. News. Stock prices rise on good news and drop on bad news. The problem is, by the time the news hits the streets, prices have already adjusted. Sometimes, however, price fluctuations overreact or underreact to the news.

The savvy investor is well-tuned to the market and the company, and can discern whether or not the reaction is over/under, and thereby capitalize on the differences.

3. Market trends. This is simply figuring out what most people will want before most people figure it out for themselves. Apples i-Everything seems to have caught on with the masses.

Then there's always The Unknown Factor: Who'd have thought Jobs would come back to work for Apple and more than quadruple it's value over the last 5 years? That's 43% a year! Yes, he was largely responsible for i-X's success, but the question is: Would iPhone and iEverything else exist if he hadn't?

It's nearly impossible to guess the 5 w's of an X-Factor before it happens. Even after it happens, it's still difficult to correctly guage its effects. However, when it became known that Jobs was back on the job, Apple's stock soared.

Savvy investors dumped loads of capital into AAPL the moment it was known that Jobs had returned, and they're still reaping the benefits, today!

**whome9** · Dec26-09, 04:28 PM

Originally Posted by Gerenuk

A friend of mine says otherwise. Could you provide some opinions or rather references to prove one view or the other?

Do a hypothesis test for randomness to the .05 level of significance. There is certainly enough data available.

http://en.wikipedia.org/wiki/Randomness_tests

**Gerenuk** · Dec27-09, 01:25 AM

Originally Posted by mugaliens

It's based on the "efficient market" hypothesis. A lot of people believe this holds true.

My friend and I, we are good at science but don't know much about economy details. Can you provide a reference that supports the efficient market hypothesis. I mean a line saying that a Noble winner or the majority of economists or whoever believes in it

Originally Posted by mugaliens

Instead, we find stocks rising and falling all the time for no apparent reason. I say "apparent" because they are rising and falling for some reason!

Of course, one doesn't even have to explain this in a complicating way, but one could even simply say the market depends on what people do and they do it for a reason.

But that is not the question. Assuming that a mathematician is only given the history of the stock price and no further real world knowledge: will he be able to extract a non-zero trend for the future and make a profit?

**Astronuc** · Dec27-09, 08:05 AM

Originally Posted by Gerenuk

My friend and I, we are good at science but don't know much about economy details. Can you provide a reference that supports the efficient market hypothesis. I mean a line saying that a Noble winner or the majority of economists or whoever believes in it

Please refer to - http://www.investopedia.com/articles/02/101502.asp
http://www.princeton.edu/~ceps/worki.../91malkiel.pdf

and one can search Google with "Efficient Market Hypothesis" (EMH)

But that is not the question. Assuming that a mathematician is only given the history of the stock price and no further real world knowledge: will he be able to extract a non-zero trend for the future and make a profit?

Knowing only the price history of a stock will not be necessarily useful to predict future trends - simply because market or economic conditions change. Companies change their focus, and some companies maintain steady business models, while others implode.

Take a look at Apple Inc (AAPL) or Oracle (ORCL) vs Exxon-Mobil (XON) or the other Dow 30 Components.

Take a look at Bear Stearns or Lehman Brothers (two of the 5 major Wall Street investment banks which collapsed in 2008).

Look into Greenlight Capital - https://www.greenlightcapital.com/ - which famously or infamously shorted Lehman Brothers, which probably helped the collapse.
Pay attention to sites like - http://dealbreaker.com/greenlight_capital/

A classic case of failure - http://www.post-gazette.com/westinghouse/prologue.asp

And one should research the demise of Penn Central - The Wreck of the Penn Central (Paperback) by Joseph R. Daughen (Author), Peter Binzen (Author)
http://www.amazon.com/Wreck-Penn-Cen.../dp/1893122085
http://en.wikipedia.org/wiki/Penn_Ce...tation_Company

It's amazing how history is repeated over and over again in the financial markets and corporate America.

**shoehorn** · Dec27-09, 01:49 PM

The efficient market hypotheses always fail spectacularly over almost any time horizon of interest. This will be grimly familiar to anyone who has spent time studying, say, the black-scholes model for derivatives pricing.

**Gerenuk** · Dec27-09, 02:28 PM

Originally Posted by shoehorn

The efficient market hypotheses always fail spectacularly over almost any time horizon of interest. This will be grimly familiar to anyone who has spent time studying, say, the black-scholes model for derivatives pricing.

That is not directly related to my question though.
The question is if anyone can predict a trend and achieve a positive expectation value of return. I'm rather looking for citations than opinions, because one can probably find both sides, but it is clear that there cannot be a positive expectation value (at least with a well-known formula) since otherwise we all would be millionaires from nothing.

**shoehorn** · Dec28-09, 09:45 AM

Well, your question is itself ill posed. It's not that you're looking for a positive expectation in order to turn a profit but an expectation that gives you a better return than a riskless bond over the same period. This is discussed in detail in plenty of finance books, Hull being a good example of such a discussion within the context of arbitrage portfolios.

**Gerenuk** · Dec28-09, 10:06 AM

Originally Posted by shoehorn

Well, your question is itself ill posed. It's not that you're looking for a positive expectation in order to turn a profit but an expectation that gives you a better return than a riskless bond over the same period.

It's inappropriate to say you know better what my question is. I said I'm looking for evidence that someone can extract a non-zero future expectation value for stock prices from just looking at the stock history.

Also I do not have access to so many books and don't have to time to read them. So what I'd appreciate most is an online reference making a statement about who believes in the efficient market hypothesis and who doesn't.

**SystemTheory** · Dec28-09, 01:20 PM

A surfer doesn't do statistics, but positions the surf board where the waves are on average about to break, and learns the skill necessary to ride waves. This is similar to the skill of a fisherman or hunter. You recognize patterns in somewhat random events, position yourself for success, and adapt to circumstances as they unfold. It is not a pure math problem!

Making Money in Stocks and Investors Business Daily are published by William J. O'Neil, a man with a track record for identify winning stocks and exploiting price waves for personal gain. Although only a few people will outperform the market at any given time, it is my experience that a skilled trader can consistently outperform a random walk portfolio.

The key to making money from price wave patterns is partial reversibility of your bets, and buy, hold, and sell rules based on fundamental and technical analysis. This is way to deal with randomness using strategic behavior. Imagine a horse race where you can take back 90% of your bet and place it on another horse at any time during the race! That's a strategy for picking winners as the race goes on ... that's the stock market!

**shoehorn** · Dec28-09, 02:18 PM

Originally Posted by Gerenuk

It's inappropriate to say you know better what my question is.

I guess it's a good thing that I said no such thing then. The fact that I can point out that a question is ill-posed doesn't imply that I can think of a better question. It simply means that the way you put the question doesn't make sense.

Also I do not have
access to so many books and
don't have to time to read
them. So what I'd appreciate
most is an online reference
making a statement about
who believes in the efficient
market hypothesis and who
doesn't.

Ah, my apologies. I thought you were more interested in, y'know, actual facts as opposed to the results of an internet vote.

**Gerenuk** · Dec28-09, 08:05 PM

Originally Posted by shoehorn

I guess it's a good thing that I said no such thing then. The fact that I can point out that a question is ill-posed doesn't imply that I can think of a better question. It simply means that the way you put the question doesn't make sense.

It might not make sense to you, but other people find it less hard to understand. Also it's absolutely OK to say it's ill-posed and ask what I could have meant by the question, but it's inappropriate to claim you know better that I actually wanted to ask another question.

Originally Posted by shoehorn

Ah, my apologies. I thought you were more interested in, y'know, actual facts as opposed to the results of an internet vote.

You again haven't understood what was written. I rather need a full quote (rather than only a reference) posted on the internet with a reference given.
I find it useless if someone gives me a random unrelated reference just because he has happened to read that book.

**mugaliens** · Dec29-09, 10:53 PM

Originally Posted by Gerenuk

My friend and I, we are good at science but don't know much about economy details. Can you provide a reference that supports the efficient market hypothesis.

Efficient-market hypothesis

I mean a line saying that a Noble winner or the majority of economists or whoever believes in it

I'm not sure of any Nobel winners on tap, but the notes and references in the Wikipedia link, such as this one, provide some interesting information.

In short, EMH has been largely disproven. Investors are not rational.

Assuming that a mathematician is only given the history of the stock price and no further real world knowledge: will he be able to extract a non-zero trend for the future and make a profit?

I think a keen, dispassionate observer of human behavior would have better luck than a mathematician.

On the other hand, Warren Buffet's approach is to simply buy firms who's fundamental (financial/economic indicators) are well above their stock price. His approach assumes the market is indeed irrational, and that the EMH is bunk.

Judging by his success, I'd say he's right.

**CRGreathouse** · Dec29-09, 11:34 PM

Originally Posted by mugaliens

Judging by his success, I'd say he's right.

You know that this is a silly statement, right? Even if the strong form of the EMH was true, there would still be Buffet-like investors making Buffet-like returns.

Originally Posted by mugaliens

On the other hand, Warren Buffet's approach is to simply buy firms who's fundamental (financial/economic indicators) are well above their stock price. His approach assumes the market is indeed irrational, and that the EMH is bunk.

Not quite. Buffet's approach assumes that the strong and semi-strong forms of the EMH are false. But his approach does not require the falsity of the weak form (which essentially says that technical analysis will fail).

**Gerenuk** · Dec30-09, 05:43 AM

Originally Posted by mugaliens

Efficient-market hypothesis

I have only limited knowledge about this topic, but the connection between random walk and EMH seems dodgy.

Originally Posted by mugaliens

In short, EMH has been largely disproven. Investors are not rational.

I think that is not related to my question (directly). I wanted to know if the stock prices are a random walk. Maybe proving EMH would prove RW, but disproving EMH doesn not disprove RW!

Originally Posted by mugaliens

I think a keen, dispassionate observer of human behavior would have better luck than a mathematician.

That is also not related to my question if you mean that the observer should follow the news. I'm only talking about a mathematical analysis that can show that there is a detectable dependency of stock prices on its history such that one can make a profit out of it.

Originally Posted by mugaliens

On the other hand, Warren Buffet's approach is to simply buy firms who's fundamental (financial/economic indicators) are well above their stock price.

"Real world" information are not an issue here.

**mheslep** · Dec30-09, 09:26 AM

Originally Posted by mugaliens

...In short, EMH has been largely disproven. Investors are not rational.

The must be qualified to 'sometimes' or 'under some circumstances', otherwise you have no hope of ever finding

a keen, dispassionate observer of human behavior [...]