I wasn't applying your metaphor in the samr context - just borrowing your rather effective imagery. And yes, of course, ultimately BB's can't exist without price, but BB's measure a specific manifestation of price - volatility - and 95% of the time price is trapped between the BB walls because the BB's adapt to the tendency of price to change. Let's back up to the original question - ( paraphrasing ) "What is the psychological explanation for the tendency for price to remain within the BB limits ( 2 Standard Deviations )?" . Answer: The question is based on an erroneous presupposition: price does not stay within these boundaries for any direct psychological reason ( other than that the trading population doesn't consistently make insane trades ) ; price stays within the 2-SD bands because the bands are reactive to trading - the BB's go with the flow and serve only to indicate volatility. During market crashes the trading population does trade insanely - then we see price move significantly outside the BB's.
I, for one, am thorougly convinced that the market is not a random walk, far from it. There are definately limits to psychological/emotional reactions by human beings, these limits are seen over and over again in the markets. Check this out: http://www.safehaven.com/showarticle.cfm?id=1130. Don't concentrate on the predictions for the future as much as the parallels, notice that even though the bubbles in japan and the us occured 10 years apart, the way the market reacted (and both the time between each reaction and the length of each reaction in %/$ move) were strikingly similar. Here's another chart http://www.safehaven.com/html/sornette/1131_a.htm comparing post-crash nikkei and s&p, the correlation of the moves is striking. It is my oppinion that this firmly proves that the market is NOT a random walk, it is a representation of mass psychology, and only that. An ocean and thousands of miles (not to mention 10 years) seperate 1990 Nikkei and 2000 S&P and their aftermaths yet the way they moved after their peaks was almost identical. I conclude by saying that there must exist a measure that will be able to gauge market psychology and forecast turning points in price and in time, what it looks like is yet to be seen, but it could be the holy grail
Rodden, I believe your logic flawed... You say that price does not stay within the BB's for any direct psychological reason but then you state 'other than that the trading population doesnt consistenly make insane trades'. There you go, are you not contradacting yourself? What constitutes an insane trade? Maybe unknowingly, psychologically, trades far outside of the 2 standard deviation area? What keeps the population from not making insane trades? Their psychology and emotions of course.
You're making a circular argument, that price stays within the Bands because the Bands remain near to but outside the range of the price. The intent of the original question was to explore why prices stay within a certain distance of a mean. The BBs are tangential. You say the question, as you state it, is based on an erroneous presupposition, but you provide no evidence that the presupposition is erroneous. If the thread-starter was wondering solely about BBs, of course, without regard to psychology, then that's another matter. As for "during market crashes the trading population does trade insanely - then we see price move significantly outside the BB's", price moves outside the BBs all the time. And the BBs follow.
Actually, there is, and has been for quite some time: the relationship of volume and price and of price to itself. Robert Rhea went into all of this over a hundred years ago, then Richard Wyckoff, more recently Victor Sperandeo. However, after indicators became so popular in the fifties and sixties, most people left all that behind. Now they will find almost any excuse to avoid looking at a naked price chart.
This is really basic statistical knowledge: cusum chart http://science.ntu.ac.uk/rsscse/ts/bts/thomas/text.html don't need psychological cause to explain that points stay without boundary because it is just RANDOM LAW (in the example above there is no psychology involved). And in fact BB cheats because it moves the mean and so the bands, whereas in classical statistical inference, the mean should stay fixed.
"Price does not stay within the BB's for any direct psychological reason other than that the trading population doesn't consistently make insane trades." There's a contradiction here? Were we seriously considering the probability that the trading population consistently makes insane trades? We're begging the psychological question here. My statement was conditional, not contradictory.
Actually, you're addressing two different questions at the same time: the psychological reasons, if any, why price tends to stay a certain distance from a mean, and the reasons why price tends to stay within BBs with a certain setting. Until you distinguish between these two questions, you're not going to get very far.
The width of the bands tracks volatility. But, the volitility bands are centered on a simple moving average of price. If price goes up, the band center slides up. The bands do track price via their center. "Look ma, I can capture 95% of the range as long as I contstantly move the center of my bands into a location so as to capture 95% of the range!"
I don't see where all this is going... You could have BB's that use all data available, they would encompass approximately 95% of the data used. The reason why BB's use only recent data is because of the dynamic nature of the stock market price data, btw, what's your point about the bands being created around a simple moving average? Think back to high school statistics, in any mathematical model, a value's standard deviation is calculated from it's distance from the MEAN. The job of the bands is to encompass price, if price goes up the bands are supposed to go up. I think the point of this entire thread though is to point out that price tends to stop and reverse as it approaches and touches the BB's, this has nothing to do with the fact that they stretch and move in relation to the mean of the last few price values.