First, the most important thing anyone needs to know about statistics: Swedish professor Hans Rosling confirms what we statisticians have known for years: “Statistics is now the sexiest subject around.”
When the Data Struts Its Stuff
Journalist Jeff Sommer explains the “risk on, risk off” paradigm that has become the dominant investment strategy over the past few years. The risk measure discussed in the article corresponds with the class concepts “volatility” or “total risk,” measured as the standard deviation of historical returns (a time-series measure).
Line Dancing With the Markets
HSBC Research provides a more rigorous explanation of the “risk on, risk off” paradigm. in recent years, the correlations between the returns of most major asset classes have increased closer to +1.0 -- both during the 2007-08 bear market and the 2009-2011 bull market. This presents a problem for portfolio managers because the benefits of diversification depend on asset correlations being significantly less than +1.0. Higher return correlations lead to higher portfolio volatility, so the risk-reduction benefits from diversification are reduced when asset correlations increase.
“Risk On, Risk Off” -- How a Paradigm is Born
In a series of papers published in 2010, Larry Gorman, Steve Sapra and Rob Weigand show that cross-sectional (rather than time-series) volatility is a more relevant measure of risk for active portfolio managers. The empirical paper was published in the Journal of Investing:
The Cross-Sectional Dispersion of Stock Returns, Alpha, and the Information Ratio
The theoretical paper was published in Investment Management and Financial Innovations:
The Role of Cross-Sectional Dispersion in Active Portfolio Management
Russell Investments now publishes a series of cross-volatility indexes based on the Gorman-Sapra-Weigand studies:
http://www.parametricportfolio.com/crossvol/research
