# Autocorrelation of daily time-series using acf and pacf plots

I have a time series dataset with daily entries over the past 20 years.

I would like to see whether the series is autocorrelated.

To do so I have done the following>

dataset=read.csv(file="https://dl.dropbox.com/u/22681355/dataset.csv", head=TRUE)
par(mfrow=c(2,1))
dataset=diff(dataset)
acf(dataset)
pacf(dataset)


From the plots it seems that there is no autocorrelation, however, I'm not sure whether the acf and pacf plots are suitable for my dataset.

What is a suitable test for stock data ?

-
May I suggest you do a little research on our site? These questions have appeared many times here and there are many examples of solutions. Start with the time-series tag and narrow your search from there. If you don't find what you're looking for, then I invite you to return here and edit your question to make it less general and more focused. –  whuber Oct 26 '12 at 17:22
@whuber, I have edited the question and have already used google and the stackexchange search bar to find relevant entries but I couldn't. –  dbr Oct 26 '12 at 17:26
You mean like google.com/search?q=acf+site%3Astats.stackexchange.com? –  whuber Oct 26 '12 at 18:05
yes, I have not found any sources on whether acf and pacf plots are suitable for assessing autocorrelation for stock data. –  dbr Oct 26 '12 at 18:25
Just add "stock" to the Google search. –  whuber Oct 26 '12 at 19:03
add comment