Skip to content
All library documents

Portfolio Weighting Strategies and Covariance Estimation for Smart Beta

Article Quant Q&A · Author: Gabriel Wadih
Read the original

Summary

The document frames a thesis project applying equal weighting, minimum variance, the most diversified portfolio, and equal risk contribution to Brazilian equities represented by the IBOV index. It also raises a practical portfolio construction question: how to estimate stock weights and whether to compare covariance estimates such as the sample covariance matrix, a risk-metrics approach, shrinkage, and DCC GARCH.

The response points to educational slides and R tools, including PortfolioAnalytics and its examples, as starting points for implementing portfolio optimization. It does not provide code, explain the optimization constraints, or compare the proposed covariance methods. Nor does it resolve how many estimators are necessary. The exchange is therefore useful as a map of strategy names and software resources, but further methodological research is needed to design a defensible comparison.

Key ideas

  • The proposed smart beta comparison includes equal weighting, minimum variance, most diversified, and equal risk contribution portfolios.
  • Portfolio weights depend on the chosen construction method and its implementation.
  • Candidate covariance approaches include sample estimates, shrinkage, risk-metrics methods, and DCC GARCH.
  • R slides, PortfolioAnalytics, and its examples are suggested as implementation resources.
  • The exchange does not assess which covariance method or number of methods is appropriate.

Tags

EquitiesPortfolio constructionRisk managementStatistics
Full text
# Equall Risk Contribution and The Most Diversified Portfolio


# Equall Risk Contribution and The Most Diversified Portfolio












I am a master degree student on applied economics at Brazil and my thesis will be about smart beta strategies. I pretend to apply Equal Weigth, Mininum Variance, Most Diversified Portfolio and Equal Risk Contribution strategies on IBOV, the cap weighted index of Brazilian Financial Market.

I've tried to found some codes on matlab or R of these strategies but I didn't have success. So I would like to awnser where can I find something that would help me to estimate the weights of the stocks ? It's hard for me doing all the code, because of that I'm asking for help.

Another thing is about the covariance matrix. I want to test the results with the sample covariance matrix, risk metrics, shrinkage and maybe DCC GARCH. How about that? It's really necessary to use more than two methods?

Hope you help me!

I appreciate!

Thanks.

## Answer by Rime (score 1)

https://quant.stackexchange.com/a/16134

I have found the slides from Yollin very useful for portfolio optimization using R such as mean-variance, max-sharpe ratio portfolio etc. http://www.rinfinance.com/RinFinance2009/presentations/yollin_slides.pdf

Also, there are some packages in R for this such as $PortfolioAnalytics$ I believe : http://www.rinfinance.com/agenda/2014/workshop/RossBennett.pdf

PortfolioAnalytics Vignettes (PDF's with examples): https://r-forge.r-project.org/scm/viewvc.php/pkg/PortfolioAnalytics/vignettes/?root=returnanalytics

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.