Low volatility has been one of the most in vogue strategies during the past decade, with market participants still cognizant of the drawdowns that occurred during the financial crisis. At S&P DJI, two of the most common strategies are applied to the S&P 500® universe to capture the low volatility anomaly—which is the observation that over the long term, less-volatile stocks have outperformed more-volatile stocks on a risk-adjusted basis. Both the S&P 500 Low Volatility Index and the S&P 500 Minimum Volatility Index have historically taken advantage of this anomaly, but the portfolio construction approaches for these indices are quite different. Exhibit 1 gives an overview of the methodology differences:
The S&P 500 Low Volatility Index employs a rankings-based approach, where stocks in the S&P 500 are sorted by the past one-year volatility of returns, and the 100 stocks with the lowest volatility are selected for index inclusion. The index does not consider other constraints in portfolio construction (e.g., sector concentration or turnover) and instead simply selects the least volatile stocks. The S&P 500 Minimum Volatility Index employs what could be considered a more sophisticated approach, using an optimizer and risk model to gain exposure to the price volatility factor, while also controlling for things such as unintended exposure to other factors, active sector weights versus the benchmark, and rebalance turnover.
A simple way to see that the two methodologies can lead to meaningfully different portfolios is to look at the constituent overlap, or how many stocks are constituents of both indices. At year-end 2016, the S&P 500 Minimum Volatility Index had 96 constituents; just 44 of those were also constituents of the S&P 500 Low Volatility Index—less than a 50% overlap. The differences in portfolio composition lead to deviations in sector composition, return attribution, and factor exposures, all of which are discussed in our recently released paper, Inside Low Volatility Indices.