[R-sig-finance] correlation between two stock market indices

[kris kumar]

>>*)  You need to be careful of asynchronous data.  In this case there
>>is no problem since both are American.

More at 
http://ideas.repec.org/p/cdl/ucsdec/97-30r.html

to add to the excellent summary & add my two cents, 
to what Patrick has elaborated.   

If these indices are measured across timezones then correlation will 
vary with sampling frequency.  There is some measurement noise and some 
people have even suggested using filters.

There is a riskmetrics paper from some time ago
which gives some simple expressions and a correlation formula for 
asynchronous data.

See: Adjusting correlation from nonsynchronous data
http://www.riskmetrics.com/journals.html

Regards
Kris