|
PureBytes Links
Trading Reference Links
|
RT'ers I forgot to cc the group in my response to
tim maratz below on his query re "what is standard
deviation". Especially since I don't think I ended
up making a clear, non-mathematical, explanation
and so I asked if other RTers could help out.
Regards,
Stan Rubenstein
----------
> From: Stan Rubenstein <stansan@xxxxxxx>
> To: timaratz@xxxxxxxxx
> Subject: Re: standard deviations
> Date: Wednesday, July 15, 1998 5:33 PM
>
> Let me (Stan Rubenstein) try to answer this.
> Stock prices go up and down as we all know.
> Some times they go up more and down more
> than other times.
> If we ignore trends the average up and down
> might be close to zero but if we quantify the daily
> up and down, call them deviations, taking the square
> of these deviations mathematically gets rid of the +(up)
> or - (down) signs, then we can add these up and down
> excursions and get a summary quantity. The summary
> calculation involves taking the square root so we get
> rid of the squaring effect we introduced.
> This resultant quantity, for the window of price data
> that we are interested it, represents an average or
> standard deviation of the stock prices about its trend.
> Now we can compare such average deviation to
> other periods of time, or to other stocks, or indices.
> Additionally there is a theory that these deviations
> have a pattern, called the normal distribution, which
> represents probabilities of the stock price reaching
> the higher or lower (de-trended) levels of deviation.
> This theory says that a price level that is inside the
> range of price calculated as 1 standard deviation above
> the trend and below the trend is more likely to happen
> than price levels greater than such a 1 standard deviation
> range. But the theory says that if the price deviations have
> a normal probability shape than 95% of all price excursions
> will show up in a price range that is twice the standard deviation
> (for our stock, in our window). These standard deviations
> are calculated for the moving window or lookback period so
> each day's price excursion is a basis for calculating a daily
> standard deviation. Bollinger said that a channel can be
> created above and below the trending stock that represents
> the volatility of the stock and we use standard deviation to
> represent price volatility.
>
> Whew I thought this would be simple when I started writing
> the explanation but on read-back I realize that the mathematics
> are easier to understand. But maybe another RT'er can
> do a better job.
> Good luck, Stan Rubenstein
>
> ----------
> From: Peter Timaratz <timaratz@xxxxxxxxx>
> To: RealTraders Discussion Group <realtraders@xxxxxxxxxxxxxx>
> Subject: Re: standard deviations
> Date: Wednesday, July 15, 1998 11:04 AM
>
> I won't attempt to explain what a standard deviation is but I will say
that
> I think it is important for a trader to have an understanding of basic
> statistics. That doesn't mean you should use it but you should understand
> the principles.
>
> Two good books available at most libraries are 'How to Think About
> Statistics' by John Phillip and 'Statistics Without Tears' by Derek
> Rowntree.
>
> -----Original Message-----
> From: bullcom <bullcom@xxxxxxxx>
> To: RealTraders Discussion Group <realtraders@xxxxxxxxxxxxxx>
> Date: Tuesday, July 14, 1998 11:34 PM
> Subject: standard deviations
>
>
> Hey RTīs,
>
> Peter was asking if 2 instead of three standard deviations wouldnīt be
> enough.
> Could you please let me know, what a sd is?
>
> Thanks - Ulrich
>
>
>
>
>
> ----------
>
|