Stock movement, standard deviation and psychology

[harrytrader]

And in constant dollar ?

Quote from jbtrader23:

1) Stock price distribution can't be normal b/c it's truncated at zero.
2) If you looked at stock returns instead, a lognormal will fit better - returns are positively skewed b/c of limited liability. Most you lose is 100%, the sky is the limit on the upside.
3) Kurtosis is a measure of how fat the tails are.
Cheers,


I think returns are positively skewed because of the natural tendency for the economy to rise over time. The economy rises, company profits rise. Profits rise and thus share prices rise as a result of that over the long haul (quite simplistic I admit).

 

[Cutten]

Quote from jbtrader23:


I still find it remarkable that all securities tend to stay within this 95% range though. Even in the biggest bubbles in the history of the world (Japan, Nasdaq 1999, Gold in 1980, etc), the markets never went much above 3 standard deviations.

Have you looked at past data at all or is this just a guess? Just bring up any historical price data and you will see moves that went not only 2SDs but some of 5, 6 or even 20.

My question is, at the upper extreme of 3 standard deviations, there is obviously mass crowd pyschology at work. Everyone is euphoric. Go back to when BRCM or ARBA were going up $50 a day in a near vertical ascent. Occasionally there'd be days outside the bollinger band. But it very rarely lasts. Try finding a stock chart in which prices go completely outside the band for many days at a time. How about for a few straight weeks? It never happens.

That is only because the bands adjust to large moves by expanding significantly. Your initial trade of entering on a piercing of the BB would be massively underwater, even though the price might 5 days later be within the BB.

You are mistaking 3SDs from a specific point, with 3SDs from a constantly changing point. Unfortunately we cannot change our initial entry point each day when the move keeps going further than expected. SDs must be measured from a *static* point, not a moving one.

Your observation is a tautology, with no practical implications for actual trading.

 

[jbtrader23]

I never said I entered at a specific point above BB's. You're taking what I say out of context. Yes, just because it goes outside the bollinger band doesn't mean I should automatically short it. That would be foolish. I've back tested my ideas and there are optimal points of going short when prices are outside the Bollinger Band.

Bubbles haven't run 5, 6 or 20 SD's up. The Crah of '87 and other exceptions to the rule are few and far between. It's not guess work, you can look at the data.

I'm not suprised with those who agree or disagree about my theories. It's to be expected.

 

[harrytrader]

This is the introduction which delimits the purpose of the article:

Introduction

The purpose of this article is to introduce statistical concepts useful as decision tools.
We will mostly focus on methodological aspects as what has been learned at school cannot
be applied adhoc since it is pure mathematics where premisces are often if not always supposed to be ideal
- for example independancy and normal law - whereas in real life they are not automatically fullfilled.
Moreover these methods are mostly dedicated to physical science in laboratory context where
parameters can be controlled: what if one wants to apply them in industrial context where such control is not possible ? Walter Shewart engineer and statistician at Bell Laboratories has been the famous theorician for statistical process control applied to economic business. Deming, his spiritual son, has proved
the value of his theory in practice since he is responsible for the fantastic quality progress of Japan after
wolrd war II. So one could inspire from that sucessful experience which has been in fact used also in service. We don't mean to transpose all the ISO norms about Quality Control, we prefer to use the original thought of Walter Shewart and Deming to build a specific approach to trading field from stock market modelling to trading system building.

Quote from harrytrader:

I am preparing an article for my site in 3 parts :

Part I - Understanding variation (volatility) concept in general
Part II - Applying variation concept in industrial activities - from the viewpoint of Walter Shewart (the father of Statistical Process Control and Quality Engineering)
Part III - Adapting the above framework to Stock Market

There should be some answers to your question.

 

[harrytrader]

On the origin of patterns according to legendary Richard Wyckoff : he doesn't mention that it is due to psychology of crowd ... and you know what my model (see http://www.elitetrader.com/vb/showthread.php?threadid=24706) could be viewed as a QUANTIFICATION of what he calls "the 'composite operator' theory, which stated that large pools work to manipulate the price of stocks, leaving definite footprints behind on the chart in patterns of accumulation and distribution"


https://www.lbrgroup.com/index.asp?..._Futures_patter

'Starting with a pattern' February 1999 - excerpt

Richard Wyckoff was a trader and market analyst who was active in the market
around the turn of the 20th century, about the same time as Charles Dow was
writing for the Wall Street Journal. As a trader who was curious about the
market, he arrived at a methodology that concentrated on price and volume
analysis, point and figure charting, and a comparison between related
markets and indexes. He wrote several books about the market, including the
famous 'Rollo Tape,' a book about the subject of tape reading written under
an assumed name. His writings were later compiled into a comprehensive
stock market training course, which is still offered today. Wyckoff
postulated the 'composite operator' theory, which stated that large pools
work to manipulate the price of stocks, leaving definite footprints behind
on the chart in patterns of accumulation and distribution. Wyckoff also
believed in the theory of 'cause and effect' whereby the market would build
up of supply or demand within a trading range.

 

[harrytrader]

Quote from harrytrader:

I am preparing an article for my site in 3 parts :
Part II - Applying variation concept in industrial activities - from the viewpoint of Walter Shewart (the father of Statistical Process Control and Quality Engineering)
Part III - Adapting the above framework to Stock Market

Why Walter Shewart ? Because he is a REALISTIC statistician because he has worked in REAL WORLD INDUSTRIES

"Normal distributions are not the norm."
http://www.elitetrader.com/vb/showthread.php?s=&threadid=25943


http://www.pyzdek.com/non-normal.htm
Non-Normal Distributions in the Real World

"After nearly two decades of research involving thousands of real-world manufacturing and nonmanufacturing operations, I have an announcement to make: Normal distributions are not the norm.

You can easily prove this by collecting data from live processes and evaluating it with an open mind. <font color=green>In fact, the early quality pioneers (such as Walter A. Shewhart) were fully aware of the scarcity of normally distributed data</font>. Today, the prevailing wisdom seems to say, “If it ain’t normal, something’s wrong.” That’s just not so."