[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [RT] Simple question on Bollinger Bands



PureBytes Links

Trading Reference Links

One typo on the post, the 4th line down should  have read "3 deviations is
equal to  1 - 1/3(squared) = 89%"

Having said all that I pulled a piece from Bollinger himself and he uses a
data set of 20 and feels "that it is the optimal period for most
applications".

Some people have probably felt that the Bollinger bands should contain 95%
of all price movement. If the previous post discussion is understood then
would understand that is not necessarily the case, sine the relationship
between the Mean and the Median also impact, what percentage of data falls
within the bands. Undoubtedly this is which large changes in price movement
frequently surge out side the bands, since the "Mean" and "Median" have
become "distorted" in those situations, and additional data is required for
things to fall back within a normal distribution as defined by the Pearsons
coefficient.

Bottom line no simple answer to your question until the data set is
established and its distribution measured I believe.
don ewers

----- Original Message -----
From: "Don Ewers" <dbewers@xxxxxxxxxxxxx>
To: <realtraders@xxxxxxxxxxx>
Sent: Monday, October 23, 2000 7:27 AM
Subject: Re: [RT] Simple question on Bollinger Bands


> Chebyshev's theorem, no matter what the form of the distribution, the
> proportion of measurements lying within k standard deviation of the mean
is
> at least 1 - 1/k(squared), where k is any positive number larger than one.
>
> 3 = or 1 deviations is equal  1 - 1/3(squared) = 89%
>
> If we are dealing only with bell shaped normal distributions the Empirical
> rule applies. Use of the Empirical rule requires the data set approach the
> form of a normal curve. To determine that a useful quantitative estimate
of
> skew (sk) is required an it can be provided by Pearsons coefficient of
skew.
> It expresses the deviation between the mean in terms of the standard
> deviation,
>
> sk = 3(Mean-Median)/s  s being the standard deviation.
>
> Some software packages allow the individual to measure a data set an
> determine the Pearons coefficient. Data sets with the coefficient between
+
> or - .50 are considered symmetrical.
>
> So the answer depends a bit on the data set.
>
> As shown above for "any" distribution, 3 standard deviations should
contain
> 89% however if the data set is normal, 3 standard deviations contain 99.7
%
> (2 contain 95% and 1 contains 68%).
> don ewers
>
> ----- Original Message -----
> From: "Alberto Torchio" <atorchio@xxxxxxxxx>
> To: "Realtraders" <realtraders@xxxxxxxxxxx>
> Sent: Monday, October 23, 2000 2:27 AM
> Subject: [RT] Simple question on Bollinger Bands
>
>
> > Dear Listmembers,
> >
> > I have been asked a simple question on Bollinger Bands and was unable to
> > answer...
> > Could anyone tell me the number of standard deviations allowing to
contain
> > within the bands 90% of price data?
> >
> > Alberto Torchio
> >  Torino, Italy
> >
> >
> >
> > To unsubscribe from this group, send an email to:
> > realtraders-unsubscribe@xxxxxxxxxxx
> >
> >
> >
> >
>
>
>
> To unsubscribe from this group, send an email to:
> realtraders-unsubscribe@xxxxxxxxxxx
>
>
>
>


-------------------------- eGroups Sponsor -------------------------~-~>
eLerts
It's Easy. It's Fun. Best of All, it's Free!
http://click.egroups.com/1/9699/0/_/152424/_/972307753/
---------------------------------------------------------------------_->

To unsubscribe from this group, send an email to:
realtraders-unsubscribe@xxxxxxxxxxx