Search
× Search
Tuesday, November 19, 2024

Archived Discussions

Recent member discussions

The Algorithmic Traders' Association prides itself on providing a forum for the publication and dissemination of its members' white papers, research, reflections, works in progress, and other contributions. Please Note that archive searches and some of our members' publications are reserved for members only, so please log in or sign up to gain the most from our members' contributions.

Pure Sector Returns

photo

 Greg Kapoustin, Principal at AlphaBetaWorks; Senior Analyst at Burlingame Asset Management, LLC

 Tuesday, July 22, 2014

Has anyone used orthogonalization or other techniques to filter market noise out of sector performance, generating pure industry-specific returns? What were the results? The following discusses the benefits of these techniques for risk management and portfolio construction:


Print

5 comments on article "Pure Sector Returns"

photo

 Mihai-Andrei M., Junior Quant Analyst at MarketScience

 Thursday, July 24, 2014



Hello Greg, take a look at discriminative clustering and merging methods for fuzzy clustering.


photo

 Joe Leonard, Trader at Cambridge Investment Research, Inc.

 Thursday, July 24, 2014



Any suggestions to gain experience and/or knowledge in market noise and Orthogonalization? Hungry for knowledge. Thank you


photo

 Mark Leeds, quantitative analyst/statistical consultant

 Friday, July 25, 2014



Hi: I have no actual experience trying to generate sector/.industry returns using factor models but Eric Zivot's S+Finmetrics book ( 2003 or 2004 ) has some really nice examples of how to to do just that in S+ using the S+Finmetrics package. If you don't have the book and are interested, my email address is markleeds2@gmail.com. The book might even be on the internet somewhere by now since it's 10 years old.

Obviously, it's been a while and I forget the details but I bet the examples in there can now be done in R pretty easily. Also, Eric is coming out with a new book about financial econometrics and R but I'm not sure when. If it's anything close to the quality of his S+Finmetrics book, then I highly recommend it.


photo

 Mark Leeds, quantitative analyst/statistical consultant

 Friday, July 25, 2014



below is the book by zivot and wang that I was referring to earlier. the material related to this discussion is in chapter 15. IIRC it doesn't deal with the issue someone on this list raised about interpretation difficulty. But there are good references at the end of each chapter also if you want to take things further.



Like I said, it's really great book aside from the fact that everything should be re-worked in R. In fact, I don't even know if TIBCO ( which used to be Insightful at the time the book was coming out ) still sells S+Finmetrics anymore. It was sold as an add-on if you were an S+/Insightful customer and I was back in the early 2000's.



http://faculty.washington.edu/ezivot/econ589/manual.pdf


photo

 Greg Kapoustin, Principal at AlphaBetaWorks; Senior Analyst at Burlingame Asset Management, LLC

 Friday, July 25, 2014



There is also a good basic financial example of factor and principal component analyses in the “R Cookbook” (https://www.inkling.com/read/r-cookbook-paul-teetor-1st/chapter-13/recipe-13-9)



The paper, “Measuring Portfolio Diversification Based on Optimized Uncorrelated Factors” (http://symmys.com/node/599) is a good discussion of Minimal Torsion. It is in the context of cleanly measuring portfolio diversification, but the approach can be adapted to other tasks.

Please login or register to post comments.

TRADING FUTURES AND OPTIONS INVOLVES SUBSTANTIAL RISK OF LOSS AND IS NOT SUITABLE FOR ALL INVESTORS
Terms Of UsePrivacy StatementCopyright 2018 Algorithmic Traders Association