A surprising notion is currently gaining traction on Wall Street: asset allocation and diversification are now passé. Just pick the asset classes you like and ride the market.
Really? Let’s look at this more closely.
Brinson Partners has generated significant investment value for its clients through a quantitative, disciplined approach to global asset allocation. Early in his career, Gary Brinson, one of the founders of Brinson Partners, published an important paper(1) based on empirical research that he performed with Brian Singer and Gilbert Beebower on the impact of asset allocation and the effect of active management on realized returns.
They found that more than 90 percent of the portfolio returns comes from the asset allocation, with the balance due to active management. Others have repeated his analysis, based on more recent data, and have reached the same conclusion.
The fall 2010 issue of the Journal of Wealth Management contains an article by two professors of finance from Indiana University Northwest, Bala G. Arshanapalli and William B. Nelson, with the intriguing title “Yes Virginia, Diversification Is Still a Free Lunch”(2). To see if diversification really matters, they used changing correlations to show that diversification is still the tool that enhances risk-adjusted performance. They also note, however, that economic turbulence makes it necessary to include more asset classes.
We take the view presented by Andrew Worthington in a study that he published in 2009 (3) on the extent to which Australian households have investments and the diversification of those investments. He presented a number of useful measures of diversification. He called “perfect diversification” an equally weighted portfolio of investments.
At the other end of the spectrum, he used the term “perfect concentration” for portfolios with almost all the weight given to very few assets. I like the measure he called the “Shannon entropy index” which I have modified to be on the scale of 0 percent to 100 percent. Zero percent corresponds to perfect concentration where as 100 percent is perfect diversification.
I collected the asset allocation policies for six different institutional investors which are summarized below:
The index proxies are used to simulate portfolios that are rebalanced on a quarterly basis and have corresponding returns that are the index returns. The performance of these portfolios is evaluated over three time periods, January 2000 to December 2004, January 2005 to December 2009 and the entire 10-year period. For each of the investors and time periods, I also calculate the Shannon diversification index.
The following charts summarize the results. Bottom line: the long-term record shows that good diversification leads to superior performance results. Innovative investors such as Stanford University, University of Chicago, Harvard University and Yale University benefit significantly from the inclusion of more asset classes with greater exposure across all the asset classes.
So much for fashionable notions on Wall Street.
(1) Brinson, Gary P., Brian D. Singer & Gilbert L. Beebower. (May/June 1991). Determinants of Portfolio Performance II: An Update, Financilal Analysis Journal, Vol. 47, No. 3, pp 40-48.
(2) Arshanapalli, Bala G. & William B. Nelson. (Fall 2010). Yes Virginia, Diversification Is Still A Free Lunch, Journal of Wealth Management, Vol. 13, No. 2, pp 34-40.
(3) Worthington, Andrew C. (June 17, 2009). Household Asset Portfolio Diversification: Evidence from the Household, Income and Labour Dynamics in Australia (Hilda) Survey, Griffith University - Department of Accounting, Finance and Economics.
