Although the core ideas and research behind low-volatility investing have been in academia for several decades,1 only in the aftermath of the global financial crisis did this category of investment products begin to gain popularity. With increased interest among finance practitioners, investors and academics, recent years have witnessed a flurry of low-volatility themed indexes and corresponding exchange-traded funds hitting the marketplace. The coming of age of low volatility has largely coincided with renewed focus on the inefficiencies of traditional capitalization-weighted portfolios and how they can be exploited by notions of "alternative/smart/strategic" beta. The timing is not coincidental; uncertainty in markets and a recent history of macro-driven bouts of risk-on/risk-off has created an increasingly low-dimensional risk landscape where risk-based strategies offer an attractive way to lower volatility, reduce correlation and perhaps beat the market.
Choosing a suitable low-volatility product, however, can be a daunting task. Apart from the sheer number of choices, different providers employ different methodologies to capture the systematic sources of return common to low-volatility stocks. Approaches range from simple rules-based selection screens to those involving more mathematical tools. It would appear that "low volatility" means different things to different people. This study attempts to differentiate the various methods of creating low-volatility portfolios and to understand the risk and return implications of each.
Low Volatility As A Source Of Return
The effectiveness of low-volatility strategies appears to challenge classical financial theory, which asserts that higher return must be accompanied by higher risk. Without embarking on a thorough review of the growing body of literature attempting to explain the foundation for low-volatility returns, one can loosely categorize hypotheses surrounding the "low-volatility anomaly" as generally drawing from one or more of the following claims:
- Delegated portfolio management suggests that even though money managers may be aware of returns in low-volatility names, they are unwilling to buy such stocks at the cost of deviating from their benchmark mandate, typically a capitalization-weighted core equity index (Brennan and Li , Baker and Wurgler , Blitz and van Vliet P. ).
- Inability to leverage heavily (or at reasonable cost) means that investors need to bias their portfolios toward higher-risk stocks in order to aim for higher potential returns, creating some degree of mispricing (Black , Black , Asness and Pedersen ).
- Portfolio managers have greater incentive to outperform in bull markets and increase AUM, therefore tending to pick higher-risk/-beta names, distorting risk/return relationships that would otherwise hold in efficient markets (Baker and Wurgler ).
Rankings-Based Risk Weighting
Risk weighting is among the earliest approaches designed to capture low volatility, assigning higher weights to stocks with lower historical risk, sometimes along with screening criteria to filter out all but the lowest-risk names. As such, "inverse risk weighting" is probably a more accurate description.2 Figures 1 and 2 provide a performance summary for such a strategy, applying risk weighting to the Russell 1000 and Russell 2000 universes once per quarter to the lowest quintile of stocks ranked by one-year realized risk.
By construction, the negative-volatility tilt guarantees that the portfolio will outperform during downmarkets. The opposite effect is true in upmarkets (albeit less strong), and combining the two effects means that from a return perspective, the efficacy of this strategy depends heavily on the prevailing market climate, as it is simply a defensive play. It comes as no surprise that products based on similar methodology had tremendous success gathering assets during the U.S. downgrade and European sovereign debt crisis in 2011.