2 Messages

Digest #3082

Computing total return correlations by "andrei radulescu-banu" iubica2

Re: Computing total return correlations by "Randy Harmelink" rharmelink

Messages

1a

Computing total return correlations

Sun Jun 1, 2014 6:06 am (PDT) . Posted by:

"andrei radulescu-banu" iubica2

I'm trying to compute ETF and fund total return correlations, and found
this page at Morningstar with monthly and quarterly returns. The 'freq'
can be 'm' for months or 'q' for quarters, the years 'y' is configurable as
well.

http://performance.morningstar.com/Performance/fund/historical-returns.action?&t=SPY&align=m&y=15&freq=m

Looking at smf-elements-4.txt, it has are entries for parsing quarterly
returns from a similar page (
http://quicktake.morningstar.com/fundnet/printreport.aspx?symbol=SPY).

But is this the right data to use for computing modern portfolio theory
correlations? It seems to me dividends and capital gains are included in
these monthly/quarterly returns, which is what I need, but Morningstar is
not very specific on this point.

Are there other, simpler APIs for computing correlations on total returns?

Thanks,
Andrei

--
==================================
Andrei Radulescu-Banu
86 Cedar St, Lexington MA
617.216.8509 (m), 781.862.5854 (h)
lex-wiki.org, lex4lang.org,
andrei4schools.com, bitdribble.com
==================================

Reply to sender . Reply to group . Reply via Web Post . All Messages (2) . Top ^

1b

Re: Computing total return correlations

Sun Jun 1, 2014 8:42 am (PDT) . Posted by:

"Randy Harmelink" rharmelink

I have no answers. Anyone else?

On Sun, Jun 1, 2014 at 6:06 AM, andrei radulescu-banu bitdribble@... wrote:

>
> I'm trying to compute ETF and fund total return correlations, and found
> this page at Morningstar with monthly and quarterly returns. The 'freq'
> can be 'm' for months or 'q' for quarters, the years 'y' is configurable as
> well.
>
>
> http://performance.morningstar.com/Performance/fund/historical-returns.action?&t=SPY&align=m&y=15&freq=m
>
> Looking at smf-elements-4.txt, it has are entries for parsing quarterly
> returns from a similar page (
> http://quicktake.morningstar.com/fundnet/printreport.aspx?symbol=SPY).
>
> But is this the right data to use for computing modern portfolio theory
> correlations? It seems to me dividends and capital gains are included in
> these monthly/quarterly returns, which is what I need, but Morningstar is
> not very specific on this point.
>
> Are there other, simpler APIs for computing correlations on total returns?
>