2
$\begingroup$

I have Bloomberg Data PX_LAST for the MSCI world (MXWO Index). I also have Bloomberg Data PX_LAST for all subindices for the MSCI world. My goal is to compute from the subindices the return of the index. I tried log and percentage returns and aggregating based on these formulas. I weighted them like that, $$\textrm{weight Subindex}_i=\frac{\textrm{PX_LAST}_i}{\sum\textrm{PX_LAST}_j} $$ enter image description here

Unfortunately, the results do not match the returns of the index itself.

Where can I find information on how the MSCI world index is aggregated from its subindices?

Does anybody know how they are aggregated?

I found this link https://www.msci.com/index-methodology but wasn't lucky finding any info there either.

Improve this question
$\endgroup$

1 Answer 1

Reset to default
2
$\begingroup$

The document you're looking for is MSCI Index Calculation Methodology. At a very high level, MSCI indices are market-cap weighted. A crude aggregation would require that you weight the component returns by their market cap as a percentage of total market cap. For example, US currently has a 59% weighting in the index (where 59% = US market cap / world market cap). Doing this should get you close, but if you need exact values, you need to follow the formulas in the document precisely.

Improve this answer
$\endgroup$
3
  • $\begingroup$ where can i get the market cap data? Thanks for the answer. $\endgroup$
    – PalimPalim
    Jun 22, 2017 at 19:55
  • $\begingroup$ It's not free. You can potentially use data from world-exchanges.org/home as a proxy. On the other hand, may I ask why you want to do this, given aggregate indices are readily available? $\endgroup$
    – Helin
    Jun 22, 2017 at 21:46
  • $\begingroup$ I would like to use it to benchmark different VaR models for the MSCI world. I would like to also add multivariate models as well as univariate ones. For the multivariates I need to know how the returns are aggregated. $\endgroup$
    – PalimPalim
    Jun 23, 2017 at 9:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.