|
Fundamental Indexation and International Diversification Authors: Javier Estrada Source: Working Paper, IESE Business School, Barcelona, Spain Date: December 2006 |
The author points out that, most of the time, whether they serve for passive or active management, benchmarks are capitalisation-weighted, on the assumption that the market portfolio is mean-variance efficient. Meanwhile, academics and practitioners tend to seek strategies to outperform the market over the long term. Among them is fundamental indexation, which weights assets by fundamental measures of value, such as dividends. In this article, author Estrada investigates which capitalisation or fundamental measure, among capitalisation or fundamental measures, is the best one to weight country index funds and ETFs when building global portfolios. More specifically, he investigates whether fundamental indices are really indices; whether fundamental indexation is passive or active; and whether price-insensitive measures are a superior weighting scheme.
First, Estrada argues that fundamental indices such as those proposed by Arnott, Hsu, and Moore (2005) are not proper benchmarks in the sense that they are not representative of the returns of the average investor. Other authors (Burr 2005) also argue that they are not really indices, but “enhanced strategies calculated as indexes”, as they require periodic rebalancing. For this reason, fundamental indices also appear to be active rather than passive strategies. As a result, holding an ETF that tracks a fundamental index incurs higher turnover transaction costs and taxes than does holding an ETF tracking a cap-weighted index. According to Schoenfeld (2006), fundamental indices are really designed to generate alpha, by making active bets within their construction. The issue with these indices is that many of their providers neither highlight their active nature, nor their implicit exposure to risk factors such as value and size.
In this article, the author compares the empirical performance of a traditional cap-weighted strategy and that of a fundamental strategy by constructing two portfolios made up of 16 country indices with two different weighting systems: in the first portfolio the country indices are cap-weighted and in the second one, they are dividend-weighted. The study was performed with data covering the period from 1974 to 2005. The first portfolio was passively held, while the second one was rebalanced at the end of each year. The performance of the portfolios was not adjusted for transaction costs.
The author finds that the dividend-weighted fundamental index outperforms the cap-weighted index by the substantial margin of 1.9% a year during the sample period 1974-2005. Meanwhile, the dividend-weighted index (DWI) did not outperform the capital-weighted index (CWI) consistently over time. In addition, the substantially different performance of the DWI and the CWI follows exclusively from the substantially different weights these indices give to each country benchmark. In addition, Estrada finds that at least part of the superior performance of the DWI is a result of its exposure to the value factor and, to a lesser extent, to the size factor. These results are consistent with those found by Bernstein (2006) and Schoenfeld (2006).
In view of these results, Estrada suggests considering other possible weighting schemes that may perform even better. He finds that a simple value strategy that weights the same country benchmark by dividend yields outperforms the dividend-weighted fundamental index by 1.7% over the same period. Furthermore, with index funds and ETFs, a low-cost implementation of this value strategy is easy. He concludes that, if investors are willing to abandon the traditional cap-weighted portfolio, they can do better than adopt a fundamental strategy such as the one considered here. Instead, he suggests that taking a traditional value approach—weighting international benchmarks by dividend yields—provides substantially higher returns with an equivalent level of risk.
References
Arnott, Robert, Jason Hsu, and Philip Moore, “Fundamental Indexation”, Financial Analysts Journal, vol. 61, n°2, 2005, p. 83-99.
Bernstein, William, “Fundamental Indexing and the Three-Factor Model”, http://www.efficientfrontier.com/ef/0adhoc/fi.htm, 2006.
Burr, Barry, “Fundamental Index Concept Gets Institutional Investors’ Attention”, Pensions & Investments, vol. 33, n°19, 2005, p. 45.
Schoenfeld, Steven, “Are Alternatively Weighted Indexes Worth their Weight?” Northern Trust report, April 2006.
Related links
