Diversified or Concentrated Factor Tilts?

Abstract

The authors compare two approaches to single-factor index design: concentrated and diversified indices. From a conceptual perspective, the authors emphasize several issues with highly concentrated portfolios. Concentration in a few stocks reflects high confidence in the precision of the link between expected returns and factor exposure, whereas expected returns are notoriously difficult to estimate precisely. Moreover, the empirical asset-pricing literature emphasizes the need to construct broad portfolios that are not unduly influenced by a small number of stocks. The authors’ empirical analysis compares broader and more narrow stock selections, as well as two different weighting schemes, equal-weighting and cap-weighting. Their results show that concentrated factor-tilted portfolios come with problems. Trying to improve a cap-weighted factor-tilted portfolio’s performance by selecting fewer stocks that are most strongly tilted to the factor does not have any effect on risk-adjusted performance. With concentration, returns and risk increase. However, concentration leads to problems such as higher turnover, high idiosyncratic risk, and longer times to trade. Conversely, achieving concentration through a move to equal-weighing leads to higher Sharpe and information ratios, with only marginally higher turnover levels.

TOPICS: Analysis of individual factors/risk premia, factor-based models, portfolio construction