[R-sig-finance] Outlier Threshold

[Melanie Vida]

I am trying to do a temporal analysis of the same variable "x" over a period of 2 years. I did a qqplot(x) and discovered an almost normal relationship between the variable x over 2001 and 2002.
However, there were some outliers in the latter half of the qqplot. To select an outlier threshold what is the recommended way when you have financial data with large variance?