Volatility targeting adjusts how much of a portfolio is invested so that its expected volatility stays close to a chosen level. Exposure is cut when markets become more volatile and raised when they are calmer.
How it works
Exposure = target volatility / forecast volatility, usually capped at a maximum. Forecast volatility is estimated from recent returns, for example over the past 20 to 60 days, often giving more weight to recent days.
An illustrative example
| Forecast volatility | Target 10%: exposure |
|---|---|
| 25% | 10 / 25 = 40% invested, 60% cash |
| 15% | 10 / 15 = 67% invested |
| 10% | 100% invested |
| 8% | 125% of capital, or 100% if capped |
The numbers are illustrative.
Why it is used
- Volatility clusters: periods of large moves tend to be followed by more large moves, so a forecast from recent data has some predictive value.
- It can keep risk steadier through time, which helps with position sizing and with setting expectations.
- Research on volatility-managed portfolios has found it improves risk-adjusted returns for some strategies, though not for all.
Costs and limits
- Turnover. Changing exposure as volatility moves means more trading. See portfolio turnover.
- Lag. Forecasts rely on the past, so a sudden spike can arrive before exposure is reduced.
- Whipsaw. Cutting exposure after a fall and adding it back during a rebound can lock in losses.
- It does not remove loss. It controls expected volatility, not the size of the worst outcome. See drawdown.
It is one tool among those in risk management in quant strategies. See also the Kelly criterion for a growth-based approach to sizing.
Frequently asked questions
What is volatility targeting?
It is a rule that scales a portfolio's exposure so that its expected volatility stays near a chosen target, reducing exposure when volatility rises.
How is the volatility forecast calculated?
Usually from recent returns, such as the past 20 to 60 days, often with more weight on the most recent observations.
Does volatility targeting reduce losses?
It aims to keep risk steadier, but it can still lose money, and it can cut exposure late in a fast fall or add it back during a rebound.
Why is exposure capped?
Without a cap, very low forecast volatility would call for large leverage, which adds risk if the forecast turns out to be wrong.