Why We’re Confused About Asset Allocation
Ibbotson says it's all about misinterpreting Brinson, Hood and Beebower.
BY Bruce Grantier | July 3, 2012
On May 30, 2012 the CFA Society of Toronto held its Annual Pension Conference with the theme of “Keeping Up With Rapid Change in Pensions.” While the excellent full-day conference covered nine wide-ranging topics, including FACTA, pension governance, CIO outsourcing, bigger is better in pension management and a panel on the future direction of the various pension constructs, the following note discusses one presentation – keynote speaker, Roger Ibbotsom, Chairman & CIO, Zebra Capital Management and Professor in Practice of Finance, Yale school of Management. In his presentation, called “The Importance of Asset Allocation: A Quarter Century of Confusion,” Ibbotson explained the source of the confusion – the misinterpretation of a 1986 Brinson, Hood, Beebower (BHB) paper which found that about 90% of the variation in pension plan returns stemmed from asset mix, with the remainder explained by other factors, mainly manager selection.
As Dr. Ibbotson explained it:
“The “importance” of asset allocation continues to be one of the trickiest and more controversial subjects among financial professionals. Attempts to clarify the issues often result in more confusion. In 2011, (my) article, “The Importance of Asset Allocation,” which won the Financial Analysts Journal Graham & Dodd Award for Best Perspective article, pinpointed the sources of confusion.”
To briefly summarize Ibbotson’s arguments:
1/ The BHB paper refers to an R2 of about 90%, which only means that 90% of the variation in returns is explained by a particular variable (in this case, asset mix). It does not say anything about causation.
2/ This high R2 is due mainly due to the correlations of all equity-type asset classes and when one adjusts for this, the remaining asset mix differences are quite a bit less and roughly in the same order as value-added from other sources.
Dr. Ibbotson’s overall conclusion was that asset mix and other sources are roughly equal contributors to returns.
He went on to discuss sources of value added and how investment portfolios can be characterized as asset classes, style factors, and security selections, while separating active from passive bets. He also introduces “Liquidity” as a new investment style.
While liquidity is recognized as a factor, or risk premium, it is not recognized as a significant and distinct factor from the traditional factors – style, market cap, and momentum. Dr. Ibbotson has written a paper which suggests that liquidity ranks about equal to value in size of premium and slightly ahead of market cap. The paper also illustrates how this premium can be earned in various portfolio strategies.
1. Brinson, Gary P, Hood, L. Randolph, and Beebower, Gilbert L., “Determinants of Portfolio Performance” Financial Analysts Journal Vol. 42, No. 4 (Jul. – Aug., 1986), pp. 39-44
2. Ibbotson, Roger G., Chen, Zhiwu, and Hu, Wendy Y. (Current Draft September 2011). Electronic copy available at: http://ssrn.com/abstract=1675108, currently being reviewed by the Financial Analysts Journal.
Bruce Grantier is President, Airth Inc.