Stock movement, standard deviation and psychology

range ? what range ? The Bands are moved with the MOVING average by construction so it is no more a range it just an EXTRAPOLATION that follows the price haha ! It was supposedly based "scientifically" on student test which gives the prob of 95% (for 2 standard deviation) but in student test the mean is fixed it is not moved at each increment so that the "scientific" adjective is rather a bullshit (there is another statistical flaw about this student test interpretation but it's too technical I won't talk about it here). I don't mean that BB is not useful but it is not as rigorously based as advanced.

And what do you mean by psychology ? If one buys a chicken in a supermarket do you include this in your psychological reason case ? Because if your definition of psychology cause is large enough it can englobe all activities for sure so that it will become a TOE (Theory Of Everything hee hee ) but is it still useful since it will be too much fuzzy ?

Last but not least, if psychological cause has any justification, it is in the case when the bands are "exploded" and so behave out of "normal" random law - see again Mandelbrot's article about 10 standard deviations - so it is funny that you pretend that things are psychological because price is contained within bands since if things behave smoothly in market there wouldn't any scientific research about the reason behind hystery of market and advance the hypothesis that market COULD (because it's not demonstrated) be driven by the psychology of crowd.

Quote from jbtrader23:

Bollinger Bands are simply one tool of measuring standard deviation. You could use a % away from the 10 day moving average for example and calculate standard deviation from that. You could calculate standard deviation over the 200 day moving average. There are literally infinite possibilities.

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.

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. Wouldn't random walk theory suggest that stocks could go up for 30 days in a row for example. After all its a 50-50 random chance stocks can go up or down for a day. If you toss a penny enough times, statistically, you'll get 30 straight heads eventually. Even with a hundred years of stock data, I dont think stocks have ever gone up or down for 30 days in a row for example. I believe record up days in a row for the DOW is somewhere around 10 or 12.

I think stocks charts are simply a graph of mass crowd pyschology, nothing more. And throughout human history, even if you took a chart of dutch tulips in 16th century Holland (I'd love to see a chart of that by the way), I think you'd find the same patterns that you'd find in modern day markets. Even though many novice traders think "it's different this time" (i.e. we're smarter than the dutch tulip traders, we're more informed than the participants in the crash of '29 etc), we still behave exactly the same way they did.

The chart of Japan overlapped with the S&P is amazing. Even two completely seperate events (the Crash of '87 and the crash of '98), appear very similar on the chart. Japan was the first major market in the world to eclipse it's previous '87 high and thus added fuel to the fire for the bubble believers. Likewise in '98. The NASDAQ shot over its previous high and was off to the races.
More...
 
BTW see http://www.elitetrader.com/vb/showthread.php?s=&threadid=25036&perpage=6&pagenumber=1
"Macro-Evolutionary Theory is filled with holes "

"there should be something else and this something can involve a more deterministic process than pure randomness of Evolution and if one believe Wolfram http://www.forbes.com/asap/2000/1127/162_7.html


"Biologists," he says, "have never been able to really explain how things get
made, how they develop, and where complicated forms come from. This is my answer.
" He points at the shell, "This mollusk is essentially running a biological
software program. That program appears to be very complex. But once you
understand it, it's actually very simple."
"
<font color=RED>"I've come to believe," says Wolfram, "that natural
selection is not all that important." </FONT>


It's very similar to Stock Market problematic in fact where official theory pretends that Randomness of multiple agents which compete among them - Evolutionary theory - dictates Market Behavior whereas I affirm, through my model that it is an illusion, that there is a deterministic process that is not due to the multiple agents (I mean the MAJORITY of course that there are SOME agents that makes the market behaves like it behaves but it is not those officially theory focus upon). And to explain that I use in fact a genetic metaphore see Plectics: "The study of simplicity and complexity"


and

See http://www.elitetrader.com/vb/showthread.php?s=&postid=378686#post378686
--------------------------------------------------------------------------------
Quote from maxpi:

Macro evolution is widely believed simply because it is taught from childhood in the schools.
--------------------------------------------------------------------------------

Like psychology of crowd in stock market that is widely spread and accepted by the vast majority which is just the counterpart of Evolution Theory applied to Stock Market (that's why there are so many researchs based on agents modelling and they just can't prove much more than herd behavior but it is qualitative and not quantitative enough to make prediction and they will NEVER be able to do prediction because it is impossible that agents can accord themselves by "magic" except if thought transmission is admitted and at the speed of light at least hee hee) ... and that I deny with my model
 
Quote from jbtrader23:

A post mentioned pyschology as a possible holy grail in terms of market timing and trading.

I think it is the holy grail! The beauty about studying mass crowd pyschology in the markets is that it has never changed and it never will. Studying traingle formations, breakouts, technical indicators, etc does not provide IMO as deep of an understanding of the markets as studying pyschology.
More...

jb
It seems to me there is two themes in your post.

First one, how price stays inside 2 or 3 SD most of the time is nothing to marvel about. It's like we were surpirsed the moving average rises when the price rised - of course it will because new higher price is part of the formula. Same with BBands. It's in the formula to chase and capture the price back into the SD range.

The second theme is crowd psychology. I also try to explain price movement in terms of psychology and it is very helpful. However, even this isn't going to predict the price so I wouldn't call it the holy grail. Consider a breakout to the upside, the euphoria of joining the ride, only to be killed by sellers. Or runing stops. These all can be analysed in terms of psychology and emotions of participants but these studies can only be done in restrospect, or at best, in present.
 
Now don't misunderstand : when Wolfram says "that natural selection is not all that important" he doesn't say that natural selection doesn't exist, for it is well known and demonstrated scientifically that it exists, for example mutations of genes exist indeed and it is a RANDOM mechanism since it occurs by errors of copying some ADN or ARN pieces, sometimes it can lead to something useful for the new organism but more often it leads to its death since the probability of obtaining something useful is very low if Nature uses a random law mechanism :D. It's somehow the same thing when I say that psychology is not primary cause, I don't deny that psychology exists in stock market and can play a role, I pretend that it is only secondary comparatively with what I have discovered as other underlying mechanism. That is to say you can use psychology to help you analysing the market but It will always stay fuzzy even if quantified by model like the one of Sornette or others it stays stochastic approach with loose prediction inherent to this type of model.


Quote from harrytrader:

BTW see http://www.elitetrader.com/vb/showthread.php?s=&threadid=25036&perpage=6&pagenumber=1
"Macro-Evolutionary Theory is filled with holes "

"there should be something else and this something can involve a more deterministic process than pure randomness of Evolution and if one believe Wolfram http://www.forbes.com/asap/2000/1127/162_7.html


"Biologists," he says, "have never been able to really explain how things get
made, how they develop, and where complicated forms come from. This is my answer.
" He points at the shell, "This mollusk is essentially running a biological
software program. That program appears to be very complex. But once you
understand it, it's actually very simple."
"
<font color=RED>"I've come to believe," says Wolfram, "that natural
selection is not all that important." </FONT>


It's very similar to Stock Market problematic in fact where official theory pretends that Randomness of multiple agents which compete among them - Evolutionary theory - dictates Market Behavior whereas I affirm, through my model that it is an illusion, that there is a deterministic process that is not due to the multiple agents (I mean the MAJORITY of course that there are SOME agents that makes the market behaves like it behaves but it is not those officially theory focus upon). And to explain that I use in fact a genetic metaphore see Plectics: "The study of simplicity and complexity"


and

See http://www.elitetrader.com/vb/showthread.php?s=&postid=378686#post378686
--------------------------------------------------------------------------------
Quote from maxpi:

Macro evolution is widely believed simply because it is taught from childhood in the schools.
--------------------------------------------------------------------------------

Like psychology of crowd in stock market that is widely spread and accepted by the vast majority which is just the counterpart of Evolution Theory applied to Stock Market (that's why there are so many researchs based on agents modelling and they just can't prove much more than herd behavior but it is qualitative and not quantitative enough to make prediction and they will NEVER be able to do prediction because it is impossible that agents can accord themselves by "magic" except if thought transmission is admitted and at the speed of light at least hee hee) ... and that I deny with my model
More...
 
Haven't read Wolfram's book. (It's huge and expensive :D ) I like his ideas though. It's interesting how his ideas seem to diminish natural selection in importance and yet confirm its existence at the same time. Natural selection is diminished in importance, and yet so is "intelligent design". Intelligent design perhaps being analogous to mass psychology. However, I don't see how Wolfram's ideas regarding simple automata in nature can apply to the markets. Where is the mechanism that "uploads" the program to the markets and allows it to run?
 
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.


I think stocks charts are simply a graph of mass crowd psychology, nothing more. And throughout human history, even if you took a chart of dutch tulips in 16th century Holland (I'd love to see a chart of that by the way), I think you'd find the same patterns that you'd find in modern day markets. Even though many novice traders think "it's different this time" (i.e. we're smarter than the dutch tulip traders, we're more informed than the participants in the crash of '29 etc), we still behave exactly the same way they did.

.
More...

Re. your first point (as excerpted here): prices will stay within the BB's 95% of the time because the BB's dynamically adjust to the volatility of trading. If this is remarkable, it's in a mathematical sense - that volatility is so integrally related to price range. The psychology that drives this consistency in the volatility/price-range relationship is probably pretty basic and unremarkable. But again, maybe it isn't. Perhaps It does deserve consideration. I don't know.

Re. your second point: stock graphs are indeed "rough graphs of mass crowd psychology" in operation, but again, the psychology itself may be basic. If we consider any given stock/market/etc. as a dynamic system in which price is the moving point of equilibrium - where buyers/sellers are in balance -we should find the reasons for price volatility rooted in the statistical frequencies of various investors' perceptions of said system's level of dynamism. But this is stating the obvious.

I agree with you - psychology is what drives all markets - but how do you calibrate that psychology in a meaningful way. How about these various confidence surveys (Michigan Sentiment, etc,) - are they actually significant?

Also, of course, perceptions of the markets are continuously modified by the endless streams of markets-related information and by geopolitical events.

I think that your theory is right, but what do you do with it?

Respectfully yours....
 
What do I do with all of this?

Find the "natural limits" of the markets (i.e. the extremes) and then go the other way. It works in any market in the world, in any security at any time.

I've already done this to a certain extent. But I'm refining and constantly improving my methodologies.

I don't believe the answers are found in mainstream sentiment surveys and certaintly not University of Michigan, Conference Board, etc. Those numbers are for the general economy and population. Even the VIX, VXN, put/call ratio, etc have their limitations.

I've studied and looked at many different Jim Rogers newspaper articles and interviews (his overall contrarian approach is very similar to mine) in order to get some clues as how to measure this pyschology.

In the Market Wizards book for example, he talked about seeing hysteria in the gold market and he sold short at $675. In his words it was the "gold markets last dying gasp". He stayed with the short until over $800 and then of course picked up tremendous profits on the decline. Now, 95% of traders out there would probably have capitulated when it went over $800 a few days later. The idea of shorting was correct, but they didn't make any money. I want to be able to quantify this type of pyschology, and compare it to past historical examples.
 
In regards to surveys such as University of Michigan, there is some correlation with the market to these surveys, but not alot. At economagic.com under the St Louis Fed, you can see a 50 year chart of the survey.
 
Quote from jbtrader23:

What do I do with all of this?

Find the "natural limits" of the markets (i.e. the extremes) and then go the other way. It works in any market in the world, in any security at any time.

.
More...

I went through the $887.00 gold peak in '80, and the crash of '87.

At these extreme moments - "market limits" - one's natural impulse is to either freeze or panic. It's as though the market is a universe unto itself, unconnected to any natural law. There appear to be no "limits" except zero and infinity.

At such times, a contrarian move requires nerves of steel and/or a terrific amount of margin.

And yes, at these moments the power of psychological force ( mass hysteria ) to drive the market is terrifyingly self-evident.
 
Quote from jbtrader23:

But on average, I think its remarkable that stocks (or any traded security) stay so confined to a seemingly random number such as 2 standard deviations.
More...

I don't see how that supports your nonrandomness point. Take a completely random variable distributed Gaus/Normal. The fact that the observations will lie within say two standard deviations most of the time does NOT in any way make it seem less random :D. In other words, while the process generating the variable may be random, if it's desribed by a distribution (e.g. Normal) the expected shape of the resultant empirical distribution is far from random.
 
You must log in or register to reply here.
Back
Top