CAS In The News


Getting the Drift
Growth, Value Benchmarks: Is the Tail Wagging the Dog?

Edited by Robin Goldwyn Blumenthal
Barron's, March 1, 2004
 

Remember how value managers grumbled during the dot.com boom that they were being unfairly punished for sticking to a value discipline and underperforming against growth-infested benchmarks?

Turns out that wasn't just sour grapes. A study last week in the Journal of Investment Consulting concludes there is an underlying "flaw" in the structure of some indexes that forces all stocks within their universe into either growth or value subsets, regardless of whether they fit those profiles. And this can "contaminate" the performance of either subset.

Authors Dorla Evans, a finance professor at the University of Alabama in Huntsville, and Kenneth Scislaw, head of consultant Scislaw Capital, say dividing the indexes by equal market capitalization muddies the waters for "pure" managers who are measured against the benchmarks.

Though managers had complained of the problem, "there's a difference between intuitively knowing and seeing it in black and white," Evans says. Relative-value benchmarks emerged, but Scislaw says they include growth-type stocks and aren't "a reflection of a legitimate, well recognized methodology, but seem to have created a methodology."

The authors looked at several indexes, including the Standard & Poor's Barra index and the Wilshire Large Cap and Target Series indexes, and concluded that the Target, because it classifies stocks based on an independent definition of growth and value, better differentiates the two.

Barra had no comment. David Blitzer, chairman of S&P's index committee, notes his company has the S&P Citigroup U.S. Broad Market Index,  which divides growth and value by a series of measures, for active managers. For passive managers, and others, he says the S&P Barra works very well indeed. Steven Foresti, a managing director of Wilshire, says characterising indexes as have "a mathematical flaw almost suggests there were errors made in the algorithms or algebra, rather than it being a consequence of having these neutral stocks" around the middle of the continuum. "That is the exact reason we have two series."