Why do equal-weighted portfolios tend to outperform capitalization-weighted portfolios? Is this tendency related to the size effect? In the May 2008 update of their paper entitled “The Effect of Value Estimation Errors On Portfolio Growth Rates”, Robert Ferguson, Dean Leistikow, Joel Rentzler and Susana Yu examine how value estimation (stock valuation) errors affect long-term returns for several portfolio weighting methods. Based on simple assumptions and general statistical analysis, they conclude that:
- Stocks with high (low) market capitalizations tend to have positive (negative) valuation errors, such that weighting by capitalization weights by both value and value estimation error.
- The tendency of value estimation errors to revert to zero (correcting stock mispricings) therefore produces, on average, a benefit for (penalty against) the future returns of small-capitalization (large-capitalization) stocks.
- This effect causes a drag on returns for capitalization-weighted portfolios compared to portfolios with weightings less connected to estimation error.
- In other words, portfolios that overweight small stocks relative to capitalization weights (such as equal-weighted portfolios) tend to suffer less mispricing drag and therefore have higher growth rates than capitalization-weighted portfolios.
In summary, statistical analysis of stock pricing noise explains both a size effect and an advantage of equal-weighted portfolios over capitalization-weighted portfolios.
