|
PureBytes Links
Trading Reference Links
|
Bob Fulks recently posted his EL version of the T3 Moving Average
function. Because it uses xAverage at its core, I was inspired to
see if it could be adapted to use my exponential moving standard
deviation instead. I hoped to create a standard deviation that has
greater smoothness for the same lag.
It worked great! The code is at http://unicorn.us.com/trading/el.html
The only caveat I can see (and this is true for the original
exponential moving standard deviation) is that your input data
must be approximately gaussian. Therefore, the exponential moving
standard deviation of closing prices will NOT approximate the
traditional standard deviation, but if you use C-C[1] as the input
data, or even True Range (which have a bell-curve distribution), the
exponential and traditional versions will track each other quite
well.
The advantage to exponential standard deviations is that they are
well-suited for time series analysis, especially time series with
short memories (like markets), because old data values aren't
weighted equally with new data.
--
,|___ Alex Matulich -- alex@xxxxxxxxxxxxxx
// +__> Director of Research and Development
// \
// __) Unicorn Research Corporation -- http://unicorn.us.com
|