Volatility Targeting - Holding Risk Constant, Letting Exposure Vary

By EC Assets Research Team, Portfolio Research · Published 20 May 2026 · Updated 15 October 2026

Volatility Targeting: Volatility targeting scales exposure inversely to forecast volatility - up in calm markets, down in turbulent ones - to keep a portfolio's realized volatility near a chosen target. Because volatility clusters and is forecastable, it tends to smooth returns and, most reliably, to reduce drawdowns by de-risking into turbulence.

What Volatility Targeting Is

Volatility targeting is a rule for sizing exposure: scale the portfolio up when markets are calm and down when they are turbulent, so that its realised volatility stays close to a chosen target. A fixed allocation holds capital constant and lets risk swing with the market; a volatility-targeted allocation holds risk constant and lets capital swing.

It is one of the most widely used overlays in systematic investing, underlying risk parity, many managed futures programmes and a range of multi-asset and risk-premia strategies.

How It Works

The exposure for the next period is the target divided by the volatility forecast, usually with a cap on leverage:

$$w_t = \min\left(\frac{\sigma_{\text{target}}}{\hat\sigma_t},\ \text{cap}\right)$$

The rule relies on volatility clustering, the tendency of high-volatility days to follow high-volatility days, which Engle (1982) formalised in ARCH models. Because volatility is persistent, recent volatility is a usable forecast of near-term risk, while returns themselves are close to unpredictable. For equities, high volatility has also tended to coincide with poor returns, so reducing exposure when volatility rises avoids part of the worst losses.

Moreira and Muir (2017) showed that scaling factor portfolios by the inverse of recent variance raised Sharpe ratios across many factors. Harvey, Hoyle, Korgaonkar, Rattray, Sargaison and Van Hemert (2018) found that volatility targeting reliably reduces extreme returns for risk assets, though it improves Sharpe ratios mainly for equities and credit. Scaling by the inverse of variance, as Moreira and Muir do, reacts more aggressively than scaling by the inverse of volatility: a doubling of volatility quarters exposure instead of halving it.

The forecast can be trailing realised volatility, an exponentially weighted average, a GARCH model or implied volatility; the Rule of 16 converts any of them between annual and daily terms.

Worked Example

A simple test on the S&P 500 price index from January 1990 to October 2026 sets exposure each day to 15 percent divided by the realised volatility of the previous 21 trading days, capped at 2.0. Dividends, cash returns, financing and trading costs are ignored.

Price returns only; the strategy averaged 1.2 times exposure and sat at the 2.0 cap on 9.6 percent of days.

The rule did what it promises: yearly volatility of the strategy ranged from 13.1 to 18.0 percent, while the index ranged from 6.7 to 41.0 percent, and the crashes of 2008 and 2020 cost about half as much. But its own worst drawdown, 49.2 percent from December 1999 to March 2003, came in the slow bear market of the early 2000s, when volatility stayed moderate and the rule kept exposure high. Lower volatility also reduces the gap between average and compounded returns, as the volatility drag calculator shows.

When It Applies (and Limitations)

It targets volatility, not crashes. Exposure is cut after volatility rises. A sudden gap inflicts its loss before the forecast reacts, and a slow decline with moderate volatility, as in 2000 to 2003, may not trigger much de-risking at all.

Pro-cyclical trading. The rule sells into falling markets and buys into calm, rising ones. In a V-shaped reversal it can cut exposure near the low and miss the rebound, and when many strategies follow similar rules their selling can deepen a selloff.

Out-of-sample evidence is weaker. Cederburg, O'Doherty, Wang and Yan (2020) found that most of the gains reported for volatility-managed portfolios disappear in realistic out-of-sample tests, and Liu, Tang and Zhou (2019) attribute much of the original equity result to the 2008 crisis.

The target and the cap are choices. A higher target or a looser leverage cap raises exposure in calm markets and the potential gain, but also the loss when volatility jumps from a low base, as it did in February 2018 and August 2024. Tighter settings give up some of the benefit for safety.

Costs and leverage. Frequent rescaling creates turnover, and running above full exposure in calm markets requires borrowing or derivatives, both of which cost money that the simple test above ignores.

Why It Matters for Institutional Investors

Stable risk budgets. Holding volatility near a target keeps a strategy's contribution to portfolio risk predictable, which simplifies risk budgeting across managers and asset classes.

Drawdown control. Its most dependable benefit is smaller losses in volatile crashes, which matters most for investors with leverage, redemption or solvency constraints; see maximum drawdown.

Judging track records. Much of a smooth track record can come from a volatility overlay rather than from skill. Knowing whether a manager targets volatility helps separate engineering from alpha.

Sizing logic. The same idea underlies growth-optimal sizing: the Kelly criterion scales exposure inversely with variance, so volatility targeting can be read as a practical, capped version of that principle.

References

1. Moreira, A., & Muir, T. (2017). Volatility-Managed Portfolios. Journal of Finance, 72(4), 1611-1644.
2. Harvey, C. R., Hoyle, E., Korgaonkar, R., Rattray, S., Sargaison, M., & Van Hemert, O. (2018). The Impact of Volatility Targeting. Journal of Portfolio Management, 45(1), 14-33.
3. Cederburg, S., O'Doherty, M. S., Wang, F., & Yan, X. (2020). On the Performance of Volatility-Managed Portfolios. Journal of Financial Economics, 138(1), 95-117.
4. Liu, F., Tang, X., & Zhou, G. (2019). Volatility-Managed Portfolio: Does It Really Work? Journal of Portfolio Management, 46(1), 38-51.
5. Engle, R. F. (1982). Autoregressive Conditional Heteroscedasticity with Estimates of the Variance of United Kingdom Inflation. Econometrica, 50(4), 987-1007.