Saturday, January 22, 2005

Better Benchmark?

Interesting twist on indexing in a Barron's article this morning by Lawrence Strauss.

"Bob Arnott and Research Affiliates, (RA) of which Arnott is chairman, has developed a series of indexes whose criteria include trailing five-year average operating income, trailing average sales, trailing average gross dividends and total employment. The Fundamental 1000 Composite, as his signature index is known, reflects sales, cash flow, book value and dividends (if applicable). At least 400 of the companies in the S&P 500 overlap with the new index, Arnott estimates....For his part, Arnott maintains that the new index "captures the average company, while the cap-weighted index captures the average stock. It's not the fundamental index that is skewed. It's the cap-weighted indexes that skew investors with a bias toward high-multiple stocks and companies with high growth expectations, which may or may not be realized."

RA's premise, based on a superficial review of their 39 page research paper, is that cap market weighted indicies are suboptimal because overvalued stocks are overweighted and undervalued stocks are underweighted. Their solution is to create in essence a "fair value" index by using a composite of company fundamental factors to drive stock selection and ranking. Certainly arguable whether this represents "fair value" though the resulting composite portfolio index, according to the research, showed a statistically significant positive excess return over the S&P500. The composite portfolio index has a distinct value bias and may have a smaller size bias ( i.e. mid cap value). It just so happens that historical evidence shows value and smaller capitalization biases have provided better historical returns so the composite portfolio index outperformance may well be attributable to size and style factors as well as or in lieu of market capitalization (ie pricing inefficiencies).

While Bob is an industry leading thinker and there is a serious research base to this idea, there is little, as a practical matter, that separates this "index" from a quantitatively managed portfolio strategy? The reference benchmark for this strategy is still the S&P500.
  • "Research Affiliates has landed one institutional account so far -- $100 million from the South Dakota Retirement System, and the firm is in negotiations with other institutional investors. "We thought it would outperform traditional capital-weighted indexes over time," explains Matt Clark, state investment officer in South Dakota. "We thought the idea would attract other investors, which might give a little boost to early adopters."
    The retirement system pays an annual management fee of 0.18%, plus 20% of the outperformance of the new index versus the S&P 500."

.18% is more than fair for a quantitative enhanced index strategy and a 20% slice of performance is even better considering there may be systematic upside bias built into the portfolio. Seems like a nice deal.....but as is often the case.... very hard for an investor to figure out if they're paying for beta or alpha?


Post a Comment

Links to this post:

Create a Link

<< Home