Subscribe to List View Past IssuesRSS translate   facebook facebook Like 0 Comment 0Share twitter
Email not displaying correctly? View it in your browser.


Investment technology for the 21st century.
Beta release of version 1.04

A beta of the next version of Portfolio Probe has been released and is available for testing.

The two biggest changes are that the idea of "risk fraction" has been expanded, and constraints on linear combinations of risk fractions is possible.

More risk fractions

The original risk fraction (added in version 1.03) was the fraction of the portfolio variance accounted for by each asset.  So these add to 1.

A new possibility is the amount of variance each asset accounts for.  So these add to the portfolio variance.

These two possibilities are also available for the variance relative to the marginal contribution to the benchmark.

Further information is in the blog post http://www.portfolioprobe.com/2011/10/10/generalizing-risk-fractions/

Linear constraints on risk fractions

The most important change is the possibility of linear constraints on risk fractions

Traditionally sector constraints, country constraints and their ilk are on the weights within each category.  But we really think about the constraints in terms of risk, not weight.  We really want to constrain risk.  This new functionality allows that.

Some more on this is in http://www.portfolioprobe.com/2011/10/13/linear-constraints-with-risk-fractions/  

Risk parity


One of the possibilities with the risk fraction technology is to create risk parity portfolios.  This concept is discussed in the blog post http://www.portfolioprobe.com/2011/10/31/risk-parity/.

Random input software testing

Testing is an important part of creating software.  Standard tests create a problem and then see if the correct answer is found.  This is necessary, but not necessarily sufficient to get high quality software.

Another possibility is to create random problems and see if trouble appears.  This allows a much wider set of test cases than the standard test method, which is very labor intensive. 

The beta of version 1.04 of Portfolio Probe benefited from this approach.  The official release of 1.04 will benefit even more from it.

My presentation on this topic at useR!2011 is at http://www.burns-stat.com/pages/Present/random_input_test_annotated.pdf 

From the blog


In case you haven't been following the Portfolio Probe blog, here are a few of the more substantive posts:

On testing a particular prediction (amazingly old news by now): http://www.portfolioprobe.com/2011/07/11/testing-an-sp-500-prediction/

A plea for us to broaden our idea of "market": http://www.portfolioprobe.com/2011/08/01/the-benchmark-gambit/

Related to that is: http://www.portfolioprobe.com/2011/08/09/the-indices-understate-the-carnage/

There were a few posts exploring portfolios with beta equal 1: http://www.portfolioprobe.com/tag/beta-equal-1/

An important post, I think, on performance measurement was: http://www.portfolioprobe.com/2011/09/19/finding-good-active-managers/

The predictability of kurtosis and skewness was explored in: http://www.portfolioprobe.com/2011/10/03/predictability-of-kurtosis-and-skewness-in-sp-constituents/

You are on this list because you have either signed up via the Portfolio Probe website or you have indicated via a request form that you would like to be added to it.

Unsubscribe patrick@burns-stat.com from this list.

Our mailing address is:
Burns Statistics
4-b Jodrell Road
London, England E3 2LA

Add us to your address book

Copyright (C) 2011 Burns Statistics All rights reserved.

Forward this email to a friend
Update your profile
Email Marketing Powered by MailChimp