Edhec-Risk
Features
Indexes and Benchmarking - February 23, 2016

Diversified or Concentrated Factor Tilts?

In a new research paper published in the latest issue of the Journal of Portfolio Management, entitled "Diversified or Concentrated Factor Tilts?," EDHEC-Risk Institute and ERI Scientific Beta researchers have highlighted the limitations of purely factor-driven approaches that aim to concentrate portfolios in a small number of stocks that are highly exposed to one or more risk factors, in order to obtain, over the long term, the best possible return associated with these risk factors. Since it neglects diversification of specific risk, this factor concentration approach exposes the investor to high idiosyncratic volatility and ultimately delivers risk-adjusted performance that is inferior to that of well-diversified factor or multi-factor indices.

Using 40 years of U.S. data, the research published compares the risk and performance characteristics of capitalisation- vs. equal-weighted factor-tilted portfolios constructed from broad (half-universe) and narrow (top two deciles) stock selections across six popular factors (Mid cap, Value, positive Momentum, Low Volatility, Low Investment and High Profitability). Equally-weighted portfolios deliver higher returns and risk-adjusted returns and have higher probabilities of outperforming the broad market. While moving from a broad selection to a narrow selection produces higher gross returns, it also increases volatility and tracking error, resulting in at best marginal gains in risk-adjusted performance before taking into account the costs of severely heightened turnover and reduced liquidity associated with narrower selections. In the end, the benefits of (naively) diversifying factor-tilted portfolios based on the broad selection far outweigh those of shifting to a narrow selection while remaining cap-weighted. Doing so produces much better performance and risk-adjusted performance in the short and the long term while only marginally impacting turnover. Over the whole period, diversifying the broad selections of factor-tilted stocks improves performance by 15.4% and produces 20% and 100% gains in Sharpe and information ratios (with the Sharpe ratio reaching 1.5 times that of the broad market). Shifting to narrow selections while maintaining cap-weighting boosts performance by only 8% and improves the risk-adjusted performance ratios by 5.4% and 9%. Ultimately, being “smart” in the area of smart beta investing means not only being diversified between factors, but also diversifying away the specific risk of the indices that represent these factors, while avoiding their overconcentration.

These new results confirm previous work on the risks of a purely factor-driven approach in smart beta. In a previous paper entitled “The Limitations of Factor Investing: Impact of the Volkswagen Scandal on Concentrated versus Diversified Factor Indices,” EDHEC researchers had shown that in their concern to maximise factor exposures, multi-factor index providers favoured concentration of indices to the detriment of their diversification, and that on the occasion of the Volkswagen scandal, and despite their multi-beta or multi-factor features, many indices were highly exposed to the specific risk of that manufacturer, and, more broadly, of the automobile industry.

They observed for example that the J.P. Morgan Europe Multi-Factor index was very strongly exposed to the risk of Volkswagen AG stock, as was the MSCI Europe Diversified Multi Factor index. As such, these indices respectively contained almost 1.5 and more than 2 times more Volkswagen AG stock than the Stoxx Europe 600, and almost 10 times and 16 times more Volkswagen AG stock than a well-diversified multi-smart-factor index.


URL for this document:
http://www.edhec-risk.com/features/RISKArticle.2016-02-23.1108

Hyperlinks in this document:
(1) http://docs.scientificbeta.com/Library/External/Academic_Papers/JPM_Winter_2016_Diversified_or_Concentrated_Factor_Tilts
(2) http://www.edhec-risk.com/latest_news/Alternative%20Investments/RISKArticle.2005-01-11.4331/attachments/ERI%20Working%20Paper%20Limitations%20of%20Factor%20Investing%20Oct%2015.pdf