CFA Institute Journal Review 18 July 2019 Volume 49 Issue 7
The Impact of Volatility Targeting (Digest summary)
Journal of Portfolio Management
Volatility scaling targets a constant volatility exposure by leveraging and deleveraging the portfolio according to time-varying expected volatility. This strategy effectively manages volatility and mean shortfall risk across several asset classes but improves the Sharpe ratio only for equity and credit portfolios.
What Is the Investment Issue?
Volatility varies over time but exhibits phases of persistence. A period of high volatility tends to continue into the near future. Investors can design and manage portfolios to address this clustering of volatility by constructing portfolios to target a constant level of volatility exposure rather than a constant notional exposure. Volatility targeting involves leveraging exposure during low-volatility periods and scaling back exposure during high-volatility periods, which will change the portfolio’s risk and return characteristics. Volatility scaling may reduce the probability of extreme losses by limiting the tail risk of extreme returns.
How Did the Authors Conduct This Research?
The authors use daily value-weighted returns of firms listed on the NYSE, AMEX, and NASDAQ, along with 10 industry equity portfolios. The equity return dataset spans 1927–2017. Daily US Treasury returns (computed from yields) are from the Federal Reserve website and begin in 1962. S&P 500 Index and Treasury futures are used to estimate intraday volatility with five-minute intervals beginning in 1988. The authors use daily futures and forward returns for 50 liquid assets, including various commodities, currencies, equity indexes, bonds, and interest rate contracts. Credit returns, net of the Treasury curve component of returns, are estimated using Bank of America Merrill Lynch US Corporate Master Total Return Index hedged with Treasuries.
The authors estimate volatility using the standard deviation of daily returns, with exponentially decaying weights. Given an annualized volatility target of 10%, volatility-scaled returns are inversely proportional to the estimated volatility, which is known 24 hours ahead of time. To allow for comparison with a non-volatility targeted strategy, unscaled returns are adjusted by a constant to produce a 10% ex post realized volatility.
The authors examine scaling at the asset level as well as the portfolio level by looking at portfolios holding different mixes of equities, bonds, and other assets.
What Are the Findings and Implications for Investors and Investment Professionals?
Why did volatility scaling improve the Sharpe ratio for such risky assets as equities and credit but not for less risky asset classes? Leverage and momentum effects may help explain these findings. Negative equity returns cause firms’ leverage to increase as measured by their debt-to-equity ratios. Higher debt/equity implies greater volatility, which will decrease position sizes as a result of volatility scaling. Consequently, volatility scaling introduces some time-series momentum as position sizes adjust in the direction of past returns. For less risky assets, this correlation is negative, which indicates mean reversion.
An important factor in this study is that equity returns do not increase to compensate for higher expected volatility, so reducing exposure during periods of heightened volatility has less effect on returns than the volatility itself, improving the portfolio’s risk-return characteristics (Sharpe ratio).
Volatility scaling makes overall volatility significantly more stable, as measured by volatility of volatility, and reduces the tail risk of extreme returns, as captured by mean shortfall in the left tail and mean exceedance in the right tail. These improvements to risk are consistent across almost all asset types.
The ability to identify risks that do not compensate expected returns is of value to all investors. The ability to limit extreme losses associated with left-tail distribution is of particular value to risk parity investors and those with high sensitivity to mean shortfall risks.
About the Author(s)
Karl Strauss, CFA, is at St. Bonaventure University.