Volatility Targeting

Volatility targeting is a method of managing a portfolio in which the amount invested in risky assets is adjusted so that the portfolio's expected volatility stays close to a fixed target. When markets are calm, exposure is increased, if necessary with borrowed money. When markets become turbulent, exposure is cut. The mix of assets may stay the same; what changes is the overall scale of the position.

A conventional portfolio holds constant weights, so its risk rises and falls with the market. A volatility-targeted portfolio aims for constant risk and lets the weights vary. The technique is used by managed futures funds, risk parity strategies, insurance products with guaranteed benefits and a range of rule-based indices. It is also known as volatility scaling or volatility control.

The exposure for the coming period is:

• $w_t$ is the fraction of capital invested in the risky asset or strategy. A value above 1 means leverage.
• $\sigma_{\text{target}}$ is the desired annual volatility.
• $\hat{\sigma}_t$ is the forecast of the asset's volatility for the period.
• $w_{\max}$ is a cap on exposure.

Suppose the target is 10 percent and the cap is 1.5. If the forecast volatility of a share index is 20 percent, exposure is $10 / 20 = 0.5$: half the capital is in the index and half in cash. If the forecast falls to 8 percent, exposure is $10 / 8 = 1.25$, which requires borrowing or the use of futures. If it falls to 5 percent, the formula gives 2.0, and the cap limits exposure to 1.5.

The forecast is typically an estimate of recent realized volatility over the past one to three months, an exponentially weighted average or a model such as GARCH. The quality of the result depends on volatility forecasting.

The case rests on two observations. First, volatility is persistent, so a high reading today predicts high volatility in the near future. Second, expected returns do not appear to rise in step with volatility over short horizons. If risk goes up and expected return does not, the reward per unit of risk is lower in turbulent periods, and reducing exposure at those times improves the portfolio's Sharpe ratio.

A 2017 study by Alan Moreira and Tyler Muir reported that scaling exposure by the inverse of recent variance raised risk-adjusted returns for the equity market and for several investment factors. Later research found the benefit to be clearest for shares and credit, where volatility rises as prices fall, and weaker for bonds, currencies and commodities. Some studies have questioned how much of the gain survives outside the sample periods tested and after trading costs.

A separate and less disputed benefit is the control of extremes. Because the largest market losses tend to occur when volatility is already elevated, a targeted portfolio usually holds less at those times. Its returns have thinner tails and its maximum drawdown has generally been smaller than that of a fixed-weight portfolio with the same average risk.

• Trend-following funds. Managers running trend following programs size every position by its volatility and then scale the whole portfolio to a target.
• Risk parity. Asset classes are weighted so that each contributes equal risk, and leverage is adjusted to hold total risk steady.
• Insurance and structured products. Funds underlying guaranteed products often follow volatility-control rules, which make the cost of hedging the guarantee more predictable.
• Individual positions. Traders apply the same idea to a single position size, holding fewer units of a more volatile instrument.

The approach is related to constant proportion portfolio insurance, which also reduces risky holdings as conditions worsen, though that method responds to losses and not to volatility.

• Lagging signals. Volatility estimates are built from past returns. A sudden fall from a calm market finds the portfolio at its highest exposure, and the reduction comes only after the loss.
• Selling low, buying high. Exposure is cut after prices have dropped and restored after volatility subsides, by which time the market may have recovered. In sharp V-shaped episodes the rule locks in losses.
• Feedback effects. When many funds follow similar rules, a rise in volatility prompts simultaneous selling, which can push prices down and volatility up further. This mechanism was discussed after the market falls of February 2018 and March 2020.
• Leverage in quiet periods. Low volatility leads to high exposure, sometimes shortly before a shock.
• Turnover. Frequent rescaling generates a transaction cost at each adjustment. Managers often rebalance only when exposure drifts beyond a band.

Results are sensitive to how the rule is specified. A short estimation window reacts quickly but trades often and can overreact to a single day. A long window is stable but slow. The cap on leverage, the frequency of rebalancing and the treatment of overnight gaps all matter. Volatility targeting alters the timing and size of exposure, not the underlying sources of return, and it cannot make an unprofitable strategy profitable.