Swedroe: More Reasons To Diversify Factors

Factors work better in combination.

Since the publication in 1992 of Eugene Fama and Kenneth French’s paper “The Cross-Section of Expected Stock Returns,” the traditional way to think about diversification has been to view portfolios as a collection of asset classes. However, we now have a nontraditional way to think about diversification.

Specifically, we can view portfolios as a collection of diversifying factors. Support for such factor-based investing strategies is provided by Antti Ilmanen and Jared Kizer in their 2012 paper, “The Death of Diversification Has Been Greatly Exaggerated.” Their work, which won the prestigious Bernstein Fabozzi/Jacobs Levy Award for the best paper of the year, made the case that factor diversification has been more effective at reducing portfolio volatility and market directionality than asset class diversification.

Six Factors

Mehdi Alighanbari, Raman Aylur Subramanian and Padmakar Kulkarni, authors of the September 2014 MSCI Research Insight paper “Factor Indexes in Perspective: Insights from 40 Years of Data,” contributed to the literature with their review of the performance of factors over the period November 1975 through March 2014.

The authors provide new insights into the behavior of factor indexes over various time periods. Their paper identified six factors, “each of which have been empirically tested in years of academic research and for which there are solid explanations on why they have historically provided risk premia.” The factors they studied are: value, low size, low volatility, high dividend yield, quality and momentum. 

Following is a summary of their findings:

• Each of the six factors achieved higher risk-adjusted performance than the benchmark MSCI World Index.
• Some factor indexes achieved higher performance with lower risk (for example, indexes for quality, high dividend yield, risk weighted and minimum volatility). By design, the minimum-volatility index demonstrated the lowest volatility over this period—about 20% lower than its parent index.
• All factors experience periods of underperformance. For instance, as the authors write, “the Quality index experienced long but slow underperformance from 1980 to 1988. Similarly, the Risk Weighted and Equal Weighted (low size) indexes experienced shorter but much steeper underperformance during the dot‐com bubble of the 1990s.”
• Demonstrating the benefits of diversifying across factors, the authors noted that the factor indexes “have reacted differently and sometimes in a complete opposite direction during certain sub‐periods. For example, the Momentum factor index has been negatively correlated to most other factor indexes, particularly in the run‐up to the dot‐com bubble’s March 2000 peak.”
• Combining factor indexes over the sample period, the authors write, “would have generated long‐term outperformance while smoothing short‐term volatility and mitigating the issue of lengthy or steep underperformance.”
• All factors persistently delivered outperformance, with the likelihood of outperformance increasing as the horizon increased. Overall, the highest odds of outperformance occurred with a balanced, equal-weighted portfolio, demonstrating the importance of diversification and discipline. Of the six factors, the highest odds of outperformance, regardless of horizon, were for momentum. The lowest odds of outperformance, not surprisingly (as it should be expected to underperform during rising markets, which are more prevalent), were for minimum volatility.
• The equal-weighted factor portfolio showed no diminishment in performance over the 40-year sample. In fact, the authors noted, “one can argue that outperformance has been more pronounced over the past decade or so, which is interesting given recent inflows into factor investing.”
• They write: “Correlations among factor index returns showed a large increase starting in the mid‐1980s, peaking around 1990. Since then, correlations have gradually declined, reaching zero around 2009. Correlations between factor index returns remained low during the recent financial crisis. These low correlations suggest that combining multiple factors may have potential diversification effects.”

Factors In Different Environments

In their related April 2014 MSCI Research Insight paper, “Index Performance in Changing Economic Environments: A Macroeconomic Framework,” authors Abhishek Gupta, Altaf Kassam, Raghu Suryanarayanan and Katalin Varga examined the performance of factors across different economic regimes. Their study also covered the same period, November 1975 through March 2014.

Following is a summary of their findings:

• The equal-weighted and value-weighted factors have been the most pro-cyclical, and the minimum-volatility and quality factors have been the most defensive.
• The momentum, value-weighted and equal-weighted factor indexes tend to exhibit the highest active returns after a large positive shock to trend growth in developed-market economies. In other words, they are the most sensitive to economic growth.
• Minimum-volatility, high dividend yield and quality tend to outperform following a large negative shock. Said another way, they are “growth hedging.”
• The momentum and equal-weighted factor indexes, the authors note, “performed best in high growth regimes, and Minimum Volatility showed the largest outperformance in the presence of low growth and high inflation. Defensive sectors have outperformed best during periods of low economic growth and low inflation.”
• They write: “When growth has remained positive or strengthened then the Equal Weighted Factor Index has historically outperformed and Minimum Volatility underperformed. On the other hand, in an environment of slowing growth an investment in Quality has generally outperformed, and when coupled with low inflation then Minimum Volatility has outperformed.”

Summary

These findings highlight the benefits of diversifying across factors, especially considering there doesn’t seem to be much, if any, evidence supporting the view that economists can forecast returns any better than what is implied by stock market prices, and, by definition, “shocks” are not forecastable.

Taken together, these two papers provide support for the strategy of diversifying across factors. They also demonstrate the need to have a strong belief system that will allow investors to stay disciplined (meaning, invested) through the inevitable long periods of underperformance that each factor experiences because its relative performance is dependent on time-varying economic regimes.

Larry Swedroe is the director of research for The BAM Alliance, a community of more than 140 independent registered investment advisors throughout the country.